Quantification of Building
Seismic Performance
Factors
FEMA P695 / June 2009
FEMA
FEMA P695/ June 2009
Quantification of Building Seismic Performance
Factors
Prepared by
APPLIED TECHNOLOGY COUNCIL
201 Redwood Shores Parkway, Suite 240
Redwood City, California 94065
www.ATCouncil.org
Prepared for
FEDERAL EMERGENCY MANAGEMENT AGENCY
Michael Mahoney, Project Officer
Robert D. Hanson, Technical Monitor
Washington, D.C.
PROJECT MANAGEMENT COMMITTEE
Charles Kircher (Project Technical Director)
Michael Constantinou
Gregory Deierlein
James R. Harris
Jon A. Heintz (Project Manager)
William T. Holmes (Project Tech. Monitor)
John Hooper
Allan R. Porush
Christopher Rojahn (Project Executive)
WORKING GROUPS
Jason Chou
Jiannis Christovasilis
Kelly Cobeen
Stephen Cranford
Brian Dean
Andre Filiatrault
Kevin Haas
Curt Haselton
WORKING GROUPS (CONT’D)
Helmut Krawinkler
Abbie Liel
Jiro Takagi
Assawin Wanitkorkul
Farzin Zareian
PROJECT REVIEW PANEL
Maryann T. Phipps (Chair)
Amr Elnashai
S.K. Ghosh
Ramon Gilsanz*
Ronald O. Hamburger
Jack Hayes
Richard E. Klingner
Philip Line
Bonnie E. Manley
Andrei M. Reinhorn
Rafael Sabelli
*ATC Board Representative
Notice
Any opinions, findings, conclusions, or recommendations expressed in this publication do not necessarily
reflect the views of the Applied Technology Council (ATC), the Department of Homeland Security
(DHS), or the Federal Emergency Management Agency (FEMA). Additionally, neither ATC, DHS,
FEMA, nor any of their employees, makes any warranty, expressed or implied, nor assumes any legal
liability or responsibility for the accuracy, completeness, or usefulness of any information, product, or
process included in this publication. Users of information from this publication assume all liability
arising from such use.
FEMA P695 Foreword iii
Foreword
The Federal Emergency Management Agency (FEMA) has the goal of
reducing the ever-increasing cost that disasters inflict on our country.
Preventing losses before they happen by designing and building to withstand
anticipated forces from these hazards is one of the key components of
mitigation, and is the only truly effective way of reducing the cost of these
disasters.
As part of its responsibilities under the National Earthquake Hazards
Reduction Program (NEHRP), and in accordance with the National
Earthquake Hazards Reduction Act of 1977 (PL 94-125) as amended, FEMA
is charged with supporting mitigation activities necessary to improve
technical quality in the field of earthquake engineering. The primary method
of addressing this charge has been supporting the investigation of seismic
and related multi-hazard technical issues as they are identified by FEMA, the
development and publication of technical design and construction guidance
products, the dissemination of these products, and support of training and
related outreach efforts. These voluntary resource guidance products present
criteria for the design, construction, upgrade, and function of buildings
subject to earthquake ground motions in order to minimize the hazard to life
for all buildings and increase the expected performance of critical and higher
occupancy structures.
The linear design procedure contained in modern building codes is based on
the concept of converting the complicated nonlinear dynamic behavior of a
building structure under seismic loading to an equivalent linear problem.
The design process starts with the selection of a basic seismic force resisting
system for the structure. The code specifies a series of prescriptive
requirements for structures based on each such system. These prescriptive
requirements regulate configuration, size, materials of construction, detailing,
and minimum required strength and stiffness. These seismic design
performance requirements are controlled through the assignment of a series
of system response coefficients (R, Cd, O0), which represent the material
properties and design detailing of the selected system. Based on the linear
dynamic response characteristics of the structure and these response
coefficients, design lateral forces are distributed to the building’s various
structural elements using linear analysis techniques and the resulting member
iv Foreword FEMA P695
forces and structural deflections are calculated. Members are then
proportioned to have adequate capacity to resist the calculated forces in
combination with other prescribed loads to ensure that calculated
displacements do not exceed maximum specified values.
As the codes have improved over the last several decades in how they
address seismic design, one of the results was an expansion of code-approved
seismic force resisting systems, with many individual systems classified by
the type of detailing used. For each increment in detailing, response
coefficients were assigned in the code, based largely on judgment and
qualitative comparison with the known response capabilities of other
systems. The result is that today’s code includes more than 80 individual
structural systems, each with individual system response coefficients
somewhat arbitrarily assigned. Many of these recently defined structural
systems have never been subjected to significant level of earthquake ground
shaking and the potential response characteristics and ability to meet the
design performance objectives is untested and unknown.
What was needed was a standard procedural methodology where the inelastic
response characteristics and performance of typical structures designed to a
set of structural system provisions could be quantified and the adequacy of
the structural system provisions to meet the design performance objectives
verified. Such a methodology would need to directly account for the
potential variations in structure configuration of structures designed to a set
of provisions, the variation in ground motion to which these structures may
be subjected and available laboratory data on the behavioral characteristics of
structural elements.
The objective of this publication was to develop a procedure to establish
consistent and rational building system performance and response parameters
(R, Cd, O0) for the linear design methods traditionally used in current
building codes. The primary application of the procedure is for the
evaluation of structural systems for new construction with equivalent
earthquake performance. The primary design performance objective was
taken to minimize the risk of structural collapse under the seismic load of
maximum considered earthquake as specified in the current NEHRP
Recommended Provisions for New Buildings and Other Structures (FEMA
450). Although the R factor is the factor of most concern, displacements and
material detailing to achieve the implied design ductilities were also
included.
It is anticipated that this methodology will ultimately be used by the nation’s
model building codes and standards to set minimum acceptable design
FEMA P695 Foreword v
criteria for standard code-approved systems, and to provide guidance in the
selection of appropriate design criteria for other systems when linear design
methods are applied. This publication will also provide a basis for future
evaluation of the current tabulation of and limitations on code-approved
structural systems for adequacy to achieve the inherent seismic performance
objectives. This material could then potentially be used to modify or
eliminate those systems or requirements that can not reliably meet these
objectives.
FEMA wishes to express its sincere gratitude to Charlie Kircher, Project
Technical Director, and to the members of the Project Team for their efforts
in the development of this recommended methodology. The Project
Management Committee consisted of Michael Constantinou, Greg Deierlein,
Jim Harris, John Hooper, and Allan Porush. They in turn guided the Project
Working Groups, which included Andre Filiatrault, Helmut Krawinkler,
Kelly Cobeen, Curt Haselton, Abbie Liel, Jiannis Christovasilis, Jason Chou,
Stephen Cranford, Brian Dean, Kevin Haas, Jiro Takagi, Assawin
Wanitkorkul, and Farzin Zareian. The Project Review Panel consisted of
Maryann Phipps (Chair), Amr Elnashai, S.K. Ghosh, Ramon Gilsanz, Ron
Hamburger, Jack Hayes, Rich Klingner, Phil Line, Bonnie Manley, Andrei
Reinhorn, and Rafael Sabelli, and they provided technical advice and
consultation over the duration of the work. The names and affiliations of all
who contributed to this report are provided in the list of Project Participants.
Without their dedication and hard work, this project would not have been
possible. The American public who live, work and play in buildings in
seismic areas are all in their debt.
Federal Emergency Management Agency
FEMA P695 Preface vii
Preface
In September 2004 the Applied Technology Council (ATC) was awarded a
“Seismic and Multi-Hazard Technical Guidance Development and Support”
contract (HSFEHQ-04-D-0641) by the Federal Emergency Management
Agency (FEMA) to conduct a variety of tasks, including one entitled
“Quantification of Building System Performance and Response Parameters”
(ATC-63 Project). The purpose of this project was to establish and document
a recommended methodology for reliably quantifying building system
performance and response parameters for use in seismic design. These
factors include the response modification coefficient (R factor), the system
overstrength factor (..), and the deflection amplification factor (Cd),
collectively referred to as “seismic performance factors.”
Seismic performance factors are used to estimate strength and deformation
demands on systems that are designed using linear methods of analysis, but
are responding in the nonlinear range. Their values are fundamentally
critical in the specification of seismic loading. R factors were initially
introduced in the ATC-3-06 report, Tentative Provisions for the Development
of Seismic Regulations for Buildings, published in 1978, and subsequently
replaced by the NEHRP Recommended Provisions for Seismic Regulations
for New Buildings and Other Structures, published by FEMA. Original R
factors were based on judgment or on qualitative comparisons with the
known response capabilities of seismic-force-resisting systems in use at the
time. Since then, the number of systems addressed in current seismic codes
and standards has increased substantially, and their ability to meet intended
seismic performance objectives is largely unknown.
The recommended methodology described in this report is based on a review
of relevant research on nonlinear response and collapse simulation,
benchmarking studies of selected structural systems, and evaluations of
additional structural systems to verify the technical soundness and
applicability of the approach. Technical review and comment at critical
developmental stages was provided by a panel of experts, which included
representatives from the steel, concrete, masonry and wood material industry
groups. A workshop of invited experts and other interested stakeholders was
convened to receive feedback on the recommended methodology, and input
from this group was instrumental in shaping the final product.
viii Preface FEMA P695
ATC is indebted to the leadership of Charlie Kircher, Project Technical
Director, and to the members of the ATC-63 Project Team for their efforts in
the development of this recommended methodology. The Project
Management Committee, consisting of Michael Constantinou, Greg
Deierlein, Jim Harris, John Hooper, and Allan Porush monitored and guided
the technical efforts of the Project Working Groups, which included Andre
Filiatrault, Helmut Krawinkler, Kelly Cobeen, Curt Haselton, Abbie Liel,
Jiannis Christovasilis, Jason Chou, Stephen Cranford, Brian Dean, Kevin
Haas, Jiro Takagi, Assawin Wanitkorkul, and Farzin Zareian. The Project
Review Panel, consisting of Maryann Phipps (Chair), Amr Elnashai, S.K.
Ghosh, Ramon Gilsanz, Ron Hamburger, Jack Hayes, Rich Klingner, Phil
Line, Bonnie Manley, Andrei Reinhorn, and Rafael Sabelli provided
technical advice and consultation over the duration of the work. The names
and affiliations of all who contributed to this report are provided in the list of
Project Participants.
ATC also gratefully acknowledges Michael Mahoney (FEMA Project
Officer), Robert Hanson (FEMA Technical Monitor), and William Holmes
(ATC Project Technical Monitor) for their input and guidance in the
preparation of this report, Peter N. Mork and Ayse Hortacsu for ATC report
production services, and Ramon Gilsanz as ATC Board Contact.
Jon A. Heintz Christopher Rojahn
ATC Director of Projects ATC Executive Director
FEMA P695 Executive Summary ix
Executive Summary
This report describes a recommended methodology for reliably quantifying
building system performance and response parameters for use in seismic
design. The recommended methodology (referred to herein as the
Methodology) provides a rational basis for establishing global seismic
performance factors (SPFs), including the response modification coefficient
(R factor), the system overstrength factor (..), and deflection amplification
factor (Cd), of new seismic-force-resisting systems proposed for inclusion in
model building codes.
The purpose of this Methodology is to provide a rational basis for
determining building seismic performance factors that, when properly
implemented in the seismic design process, will result in equivalent safety
against collapse in an earthquake, comparable to the inherent safety against
collapse intended by current seismic codes, for buildings with different
seismic-force-resisting systems.
As developed, the following key principles outline the scope and basis of the
Methodology:
. It is applicable to new building structural systems.
. It is compatible with the NEHRP Recommended Provisions for Seismic
Regulations for New Buildings and Other Structures (FEMA, 2004a) and
ASCE/SEI 7, Minimum Design Loads for Buildings and Other
Structures, (ASCE, 2006a).
. It is consistent with a basic life safety performance objective inherent in
current seismic codes and standards.
. Earthquake hazard is based on Maximum Considered Earthquake ground
motions.
. Concepts are consistent with seismic performance factor definitions in
current seismic codes and standards.
. Safety is expressed in terms of a collapse margin ratio.
. Performance is quantified through nonlinear collapse simulation on a set
of archetype models.
x Executive Summary FEMA P695
. Uncertainty is explicitly considered in the collapse performance
evaluation.
The Methodology is intended to apply broadly to all buildings, recognizing
that this objective may not be fully achieved for certain seismic environments
and building configurations. Likewise, the Methodology has incorporated
certain simplifying assumptions deemed appropriate for reliable evaluation of
seismic performance. Key assumptions and potential limitations of the
Methodology are presented and summarized.
In the development of the Methodology, selected seismic-force-resisting
systems were evaluated to illustrate the application of the Methodology and
verify its methods. Results of these studies provide insight into the collapse
performance of buildings and appropriate values of seismic performance
factors. Observations and conclusions in terms of generic findings applicable
to all systems, and specific findings for certain types of seismic-forceresisting
systems are presented. These findings should be considered
generally representative, but not necessarily indicative of all possible trends,
given limitations in the number and types of systems evaluated.
The Methodology is recommended for use with model building codes and
resource documents to set minimum acceptable design criteria for standard
code-approved seismic-force-resisting systems, and to provide guidance in
the selection of appropriate design criteria for other systems when linear
design methods are applied. It also provides a basis for evaluation of current
code-approved systems for their ability to achieve intended seismic
performance objectives. It is possible that results of future work based on
this Methodology could be used to modify or eliminate those systems or
requirements that cannot reliably meet these objectives.
FEMA P695 Table of Contents xi
Table of Contents
Foreword ....................................................................................................... iii
Preface .......................................................................................................... vii
Executive Summary ..................................................................................... ix
List of Figures ............................................................................................. xix
List of Tables ........................................................................................... xxvii
1. Introduction ...................................................................................... 1-1
1.1 Background and Purpose .......................................................... 1-1
1.2 Scope and Basis of the Methodology ....................................... 1-2
1.2.1 Applicable to New Building Structural
Systems ........................................................................ 1-2
1.2.2 Compatible with the NEHRP Recommended
Provisions and ASCE/SEI 7 ........................................ 1-3
1.2.3 Consistent with the Life Safety Performance
Objective ..................................................................... 1-4
1.2.4 Based on Acceptably Low Probability of
Structural Collapse ...................................................... 1-4
1.2.5 Earthquake Hazard based on MCE Ground
Motions........................................................................ 1-5
1.2.6 Concepts Consistent with Current Seismic
Performance Factor Definitions .................................. 1-5
1.2.7 Safety Expressed in Terms of Collapse
Margin Ratio ............................................................... 1-9
1.2.8 Performance Quantified Through Nonlinear
Collapse Simulation on a set of Archetype
Models ......................................................................... 1-9
1.2.9 Uncertainty Considered in Performance
Evaluation .................................................................. 1-10
1.3 Content and Organization ....................................................... 1-10
2. Overview of Methodology ................................................................ 2-1
2.1 General Framework .................................................................. 2-1
2.2 Description of Process .............................................................. 2-2
2.3 Develop System Concept ......................................................... 2-2
2.4 Obtain Required Information ................................................... 2-3
2.5 Characterize Behavior .............................................................. 2-4
2.6 Develop Models ....................................................................... 2-5
2.7 Analyze Models ........................................................................ 2-6
2.8 Evaluate Performance ............................................................... 2-8
2.9 Document Results ..................................................................... 2-9
2.10 Peer Review ............................................................................ 2-10
xii Table of Contents FEMA P695
3. Required System Information ......................................................... 3-1
3.1 General ...................................................................................... 3-1
3.2 Intended Applications and Performance ................................... 3-2
3.3 Design Requirements ................................................................ 3-3
3.3.1 Basis for design Requirements .................................... 3-3
3.3.2 Application Limits and Strength Limit States ............. 3-4
3.3.3 Overstrength Design Criteria ....................................... 3-5
3.3.4 Configuration Issues .................................................... 3-5
3.3.5 Material Properties ....................................................... 3-6
3.3.6 Strength and Stiffness Requirements ........................... 3-6
3.3.7 Approximate Fundamental Period ............................... 3-8
3.4 Quality Rating for Design Requirements .................................. 3-8
3.4.1 Completeness and Robustness Characteristics ............ 3-9
3.4.2 Confidence in Design Requirements ......................... 3-10
3.5 Data from Experimental Investigation .................................... 3-10
3.5.1 Objectives of Testing Program .................................. 3-11
3.5.2 General Testing Issues ............................................... 3-12
3.5.3 Material Testing Program .......................................... 3-14
3.5.4 Component, Connection, and Assembly
Testing Program......................................................... 3-15
3.5.5 Loading History ......................................................... 3-17
3.5.6 System Testing Program ............................................ 3-18
3.6 Quality Rating of Test Data .................................................... 3-19
3.6.1 Completeness and Robustness Characteristics .......... 3-20
3.6.2 Confidence in Test Results ........................................ 3-21
4. Archetype Development ................................................................... 4-1
4.1 Development of Structural System Archetypes ........................ 4-1
4.2 Index Archetype Configurations ............................................... 4-2
4.2.1 Structural Configuration Issues ................................... 4-4
4.2.2 Seismic Behavioral Effects .......................................... 4-6
4.2.3 Load Path and Components not Designated as
Part of the Seismic-Force-Resisting System ................ 4-9
4.2.4 Overstrength Due to Non-Seismic Loading ............... 4-10
4.3 Performance Groups ............................................................... 4-10
4.3.1 Identification of Performance Groups ....................... 4-11
5. Nonlinear Model Development ........................................................ 5-1
5.1 Development of Nonlinear Models for Collapse
Simulation ................................................................................. 5-1
5.2 Index Archetype Designs .......................................................... 5-1
5.2.1 Seismic Design Methods ............................................. 5-3
5.2.2 Criteria for Seismic Design Loading ........................... 5-4
5.2.3 Transition Period, Ts .................................................... 5-6
5.2.4 Seismic Base Shear, V ................................................. 5-7
5.2.5 Fundamental Period, T ................................................. 5-8
5.2.6 Loads and Load Combinations .................................... 5-8
5.2.7 Trial Values of Seismic Performance Factors ............. 5-9
5.2.8 Performance Group Design Variations ...................... 5-10
5.3 Index Archetype Models ......................................................... 5-11
5.3.1 Index Archetype Model Idealization ......................... 5-14
5.4 Simulated Collapse Modes ..................................................... 5-16
5.5 Non-Simulated Collapse Modes ............................................. 5-20
FEMA P695 Table of Contents xiii
5.6 Characterization of Modeling Uncertainties .......................... 5-22
5.7 Quality Rating of Index Archetype Models ........................... 5-23
5.7.1 Representation of Collapse Characteristics ............... 5-24
5.7.2 Accuracy and Robustness of Models ........................ 5-25
6. Nonlinear Analysis ........................................................................... 6-1
6.1 Nonlinear Analysis Procedures ................................................ 6-1
6.1.1 Nonlinear Analysis Software ....................................... 6-2
6.2 Input Ground Motions .............................................................. 6-3
6.2.1 MCE Ground Motion Intensity ................................... 6-3
6.2.2 Ground Motion Record Sets ........................................ 6-4
6.2.3 Ground Motion Record Scaling .................................. 6-6
6.3 Nonlinear Static (Pushover) Analyses ...................................... 6-7
6.4 Nonlinear Dynamic (Response History) Analyses ................... 6-9
6.4.1 Background on Assessment of Collapse Capacity .... 6-10
6.4.2 Calculation of Median Collapse Capacity and
CMR .......................................................................... 6-12
6.4.3 Ground Motion Record Intensity and Scaling ........... 6-12
6.4.4 Energy Dissipation and Viscous Damping ................ 6-13
6.4.5 Guidelines for CMR Calculation using Three-
Dimensional Nonlinear Dynamic Analyses .............. 6-13
6.4.6 Summary of Procedure for Nonlinear Dynamic
Analysis ..................................................................... 6-14
6.5 Documentation of Analysis Results ....................................... 6-15
6.5.1 Documentation of Nonlinear Models ........................ 6-15
6.5.2 Data from Nonlinear Static Analyses ........................ 6-16
6.5.3 Data from Nonlinear Dynamic Analyses .................. 6-16
7. Performance Evaluation .................................................................. 7-1
7.1 Overview of the Performance Evaluation Process ................... 7-1
7.1.1 Performance Group Evaluation Criteria ...................... 7-3
7.1.2 Acceptable Probability of Collapse ............................. 7-4
7.2 Adjusted Collapse Margin Ratio .............................................. 7-5
7.2.1 Effect of Spectral Shape on Collapse Margin ............. 7-5
7.2.2 Spectral Shape Factors ................................................ 7-5
7.3 Total System Collapse Uncertainty .......................................... 7-7
7.3.1 Sources of Uncertainty ................................................ 7-7
7.3.2 Combining Uncertainties in Collapse
Evaluation .................................................................... 7-8
7.3.3 Effect of Uncertainty on Collapse Margin .................. 7-9
7.3.4 Total System Collapse Uncertainty ........................... 7-11
7.4 Acceptable Values of Adjusted Collapse Margin
Ratio ....................................................................................... 7-13
7.5 Evaluation of the Response Modification
Coefficient, R.......................................................................... 7-15
7.6 Evaluation of the Overstrength Factor, .O............................. 7-15
7.7 Evaluation of the Deflection Amplification Factor, Cd .......... 7-16
8. Documentation and Peer Review .................................................... 8-1
8.1 Recommended Qualifications, Expertise and
Responsibilities for a System Development Team ................... 8-1
8.1.1 System Sponsor ........................................................... 8-1
xiv Table of Contents FEMA P695
8.1.2 Testing Qualifications, Expertise and
Responsibilities ............................................................ 8-1
8.1.3 Engineering and Construction Qualifications,
Expertise and Responsibilities ..................................... 8-2
8.1.4 Analytical Qualifications, Expertise and
Responsibilities ............................................................ 8-2
8.2 Documentation of System Development and Results ............... 8-2
8.3 Peer Review Panel .................................................................... 8-3
8.3.1 Peer Review Panel Selection ....................................... 8-4
8.3.2 Peer Review Roles and Responsibilities ...................... 8-4
8.4 Submittal ................................................................................... 8-5
9. Example Applications ...................................................................... 9-1
9.1 General ...................................................................................... 9-1
9.2 Example Application - Reinforced Concrete Special
Moment Frame System ............................................................. 9-2
9.2.1 Introduction ................................................................. 9-2
9.2.2 Overview and Approach .............................................. 9-2
9.2.3 Structural System Information ..................................... 9-3
9.2.4 Identification of Reinforced Concrete Special
Moment Frame Archetype Configurations .................. 9-4
9.2.5 Nonlinear Model Development ................................. 9-10
9.2.6 Nonlinear Structural Analysis .................................... 9-13
9.2.7 Performance Evaluation ............................................. 9-17
9.2.8 Iteration: Adjustment of Design Requirements to
Meet Performance Goals ........................................... 9-21
9.2.9 Evaluation of O0 Using Final Set of Archetype
Designs ...................................................................... 9-25
9.2.10 Summary Observations .............................................. 9-25
9.3 Example Application - Reinforced Concrete Ordinary
Moment Frame System ........................................................... 9-25
9.3.1 Introduction ............................................................... 9-25
9.3.2 Overview and Approach ............................................ 9-26
9.3.3 Structural System Information ................................... 9-26
9.3.4 Identification of Reinforced Concrete Ordinary
Moment Frame Archetype Configurations ................ 9-27
9.3.5 Nonlinear Model Development ................................. 9-33
9.3.6 Nonlinear Structural Analysis .................................... 9-34
9.3.7 Performance Evaluation for SDC B ........................... 9-38
9.3.8 Performance Evaluation for SDC C ........................... 9-40
9.3.9 Evaluation of Oo Using Set of Archetype Designs .... 9-41
9.3.10 Summary Observations .............................................. 9-42
9.4 Example Application - Wood Light-Frame System ............... 9-43
9.4.1 Introduction ............................................................... 9-43
9.4.2 Overview and Approach ............................................ 9-43
9.4.3 Structural System Information ................................... 9-43
9.4.4 Identification of Wood Light-Frame Archetype
Configurations ........................................................... 9-44
9.4.5 Nonlinear Model Development ................................. 9-48
9.4.6 Nonlinear Structural Analyses ................................... 9-51
9.4.7 Performance Evaluation ............................................. 9-54
9.4.8 Calculation of O0 using Set of Archetype
Designs ...................................................................... 9-57
FEMA P695 Table of Contents xv
9.4.9 Summary Observations ............................................. 9-57
9.5 Example Applications - Summary Observations and
Conclusions ............................................................................ 9-58
9.5.1 Short Period Structures .............................................. 9-58
9.5.2 Tall Moment Frame Structures .................................. 9-58
9.5.3 Collapse Performance for Different Seismic
Design Categories...................................................... 9-59
10. Supporting Studies ......................................................................... 10-1
10.1 General ................................................................................... 10-1
10.2 Assessment of Non-Simulated Failure Modes in a Steel
Special Moment Frame System .............................................. 10-1
10.2.1 Overview and Approach ............................................ 10-1
10.2.2 Structural System Information .................................. 10-3
10.2.3 Nonlinear Analysis Model......................................... 10-4
10.2.4 Procedure for Collapse Performance Assessment,
Incorporating Non-Simulated Failure Modes ............ 10-6
10.3 Collapse Evaluation of Seismically Isolated
Structures .............................................................................. 10-12
10.3.1 Introduction ............................................................. 10-12
10.3.2 Isolator and Structural System Information ............ 10-14
10.3.3 Modeling Isolated Structure Archetypes ................. 10-16
10.3.4 Design Properties of Isolated Structure
Archetypes ............................................................... 10-21
10.3.5 Nonlinear Static Analysis for Period-Based
Ductility, SSFs, Record-to-Record Variability
and Overstrength ..................................................... 10-27
10.3.6 Collapse Evaluation Results .................................... 10-30
10.3.7 Summary and Conclusion ....................................... 10-39
11 Conclusions and Recommendations ............................................. 11-1
11.1 Assumptions and Limitations ................................................. 11-1
11.1.1 Far-Field Record Set Ground Motions ...................... 11-1
11.1.2 Influence of Secondary Systems on Collapse
Performance............................................................... 11-3
11.1.3 Buildings with Significant Irregularities ................... 11-4
11.1.4 Redundancy of the Seismic-Force-Resisting
System ....................................................................... 11-5
11.2 Observations and Conclusions ............................................... 11-5
11.2.1 Generic Findings ....................................................... 11-5
11.2.2 Specific Findings ....................................................... 11-8
11.3 Collapse Evaluation of Individual Buildings ......................... 11-9
11.3.1 Feasibility ................................................................ 11-10
11.3.2 Approach ................................................................. 11-10
11.4 Recommendations for Further Study .................................... 11-10
11.4.1 Studies Related to Improving and Refining
the Methodology ..................................................... 11-11
11.4.2 Studies Related to Advancing Seismic Design
Practice and Building Code Requirements
(ASCE/SEI 7-05) ..................................................... 11-12
Appendix A: Ground Motion Record Sets ............................................. A-1
A.1 Introduction ............................................................................. A-1
xvi Table of Contents FEMA P695
A.2 Objectives ................................................................................ A-2
A.3 Approach ................................................................................. A-3
A.4 Spectral Shape Consideration .................................................. A-4
A.5 Maximum Considered Earthquake and Design Earthquake
Demand (ASCE/SEI 7-05) ....................................................... A-4
A.6 PEER NGA Database .............................................................. A-7
A.7 Record Selection Criteria ......................................................... A-8
A.8 Scaling Method ........................................................................ A-9
A.9 Far-Field Record Set .............................................................. A-13
A.10 Near-Field Record Set ........................................................... A-20
A.11 Comparison of Far-Field and Near-Field Record Sets .......... A-27
A.12 Robustness of Far-Field Record Set ...................................... A-33
A.12.1 Approach to Evaluating Robustness ......................... A-33
A.12.2 Effects of PGA Selection Criteria Alone .................. A-34
A.12.3 Effects of PGV Selection Criteria Alone .................. A-36
A.12.4 Effects of both PGA and PGV Selection Criteria
Simultaneously, as well as Selection of Two
Records from Each Event ......................................... A-37
A.12.5 Summary of the Robustness of the Far-Field Set ..... A-38
A.13 Assessment of Record-to-Record Variability in Collapse
Fragility ................................................................................. A-39
A.14 Summary and Conclusion ...................................................... A-43
Appendix B: Adjustment of Collapse Capacity Considering Effects
of Spectral Shape ............................................................................. B-1
B.1 Introduction ............................................................................. B-1
B.2 Previous Research on Simplified Methods to Account
for Spectral Shape (Epsilon) .................................................... B-4
B.3 Development of a Simplified Method to Adjust Collapse
Capacity for Effects of Spectral Shape (Epsilon) .................... B-6
B.3.1 Epsilon Values for the Ground Motions in the
Far-Field Set ............................................................... B-7
B.3.2 Target Epsilon Values ................................................. B-7
B.3.3 Impact of Spectral Shape (.) on Median Collapse
Capacity .................................................................... B-11
B.4 Final Simplified Factors to Adjust Median Collapse
Capacity for the Effects of Spectral Shape ............................ B-21
B.5 Application to Site Specific Performance Assessment .......... B-24
Appendix C: Development of Index Archetype Configurations ........... C-1
C.1. Development of Index Archetype Configurations for
a Reinforced Concrete Moment Frame System ....................... C-1
C.1.1 Establishing the Archetype Design Space .................. C-1
C.1.2 Identifying Index Archetype Configurations
and Populating Performance Groups .......................... C-4
C.1.3 Preparing Index Archetype Designs and Index
Archetype Models ....................................................... C-7
C.2 Development of Index Archetype Configurations for a
Wood Light-Frame Shear Wall System ................................... C-9
C.2.1 Establishing the Archetype Design Space .................. C-9
C.2.2 Identifying Index Archetype Configurations
and Populating Performance Groups ........................ C-10
FEMA P695 Table of Contents xvii
C.2.3 Preparing Index Archetype Designs and Index
Archetype Models ..................................................... C-10
C.2.4 Other Considerations for Wood Light-Frame
Shear Wall Systems ................................................... C-13
Appendix D: Consideration of Behavioral Effects ................................. D-1
D.1 Identification of Structural failure Modes ............................... D-1
D.2 System Definition .................................................................... D-2
D.3 Element Deterioration Modes .................................................. D-3
D.3.1 Flexural Hinging of Beams and Columns .................. D-5
D.3.2 Compressive Failure of Columns ............................... D-5
D.3.3 Shear Failure of Beam and Columns .......................... D-5
D.3.4 Joint Panel Shear Behavior ......................................... D-6
D.3.5 Bond-Slip of Reinforcing Bars ................................... D-7
D.3.6 Punching Shear in Slab-Column Connections ........... D-7
D.4 Local and Global Collapse Scenarios ...................................... D-7
D.5 Likelihood of Collapse Scenarios ............................................ D-8
D.6 Collapse Simulation ................................................................ D-9
Appendix E: Nonlinear Modeling of Reinforced Concrete Moment
Frame Systems .................................................................................. E-1
E.1 Purpose ..................................................................................... E-1
E.2 Structural Modeling Overview ................................................. E-1
E.3 Beam-Column Element Model ................................................. E-2
E.3.1 Element and Hysteretic Model .................................... E-3
E.3.2 Calibration of Parameters for the Reinforced
Concrete Beam-Column Element Model .................... E-5
E.4 Joint Modeling ........................................................................ E-15
E.4.1 Shear Panel Spring .................................................... E-16
E.4.2 Bond-Slip Spring Model ........................................... E-16
Appendix F: Collapse Evaluation of Individual Buildings ..................... F-1
F.1 Introduction .............................................................................. F-1
F.2 Feasibility ................................................................................. F-1
F.3 Approach .................................................................................. F-1
F.4 Collapse Evaluation of Individual Building Systems ............... F-2
F.4.1 Step One: Develop Nonlinear Model(s) ...................... F-2
F.4.2 Step Two: Define Limit States and Acceptance
Criteria ......................................................................... F-3
F.4.3 Step Three: Determine Total System Uncertainty
and Acceptable Collapse Margin Ratio ....................... F-3
F.4.4 Step Four: Perform Nonlinear Static Analysis
(NSA) .......................................................................... F-4
F.4.5 Step Five: Select Record Set and Scale Records ......... F-4
F.4.6 Step Six: Perform Nonlinear Dynamic Analysis
(NDA) and Evaluate Performance ............................... F-5
xviii Table of Contents FEMA P695
Symbols ...................................................................................................... G-1
Glossary ...................................................................................................... H-1
References ................................................................................................... I-1
Project Participants .................................................................................... J-1
FEMA P695 List of Tables xxvii
List of Tables
Table 3-1 Quality Rating of Design Requirements ............................ 3-8
Table 3-2 Quality Rating of Test Data from an Experimental
Investigation Program ...................................................... 3-20
Table 4-1 Configuration Design Variables and Related Physical
Properties ........................................................................... 4-5
Table 4-2 Seismic Behavioral Effects and Related Design
Considerations ................................................................... 4-7
Table 4-3 Generic Performance Group Matrix ................................ 4-12
Table 5-1A Summary of Mapped Values of Short-Period Spectral
Acceleration, Site Coefficients and Design Parameters
for Seismic Design Categories B, C, and D ....................... 5-5
Table 5-1B Summary of Mapped Values of 1-Second Spectral
Acceleration, Site Coefficients and Design Parameters
for Seismic Design Categories B, C, and D ....................... 5-5
Table 5-2 General Considerations for Developing Index
Archetype Models ............................................................ 5-12
Table 5-3 Quality Rating of Index Archetype Models ..................... 5-23
Table 6-1 Summary of Maximum Considered Earthquake Spectral
Accelerations And Transition Periods Used for Collapse
Evaluation of Seismic Design Category D, C and B
Structure Archetypes, Respectively ................................... 6-3
Table 7-1a Spectral Shape Factor (SSF) for Archetypes Designed
for SDC B, SDC C, or SDC Dmin ....................................... 7-6
Table 7-1b Spectral shape factor (SSF) for archetypes designed
using SDC Dmax .................................................................. 7-6
Table 7-2a Total System Collapse Uncertainty (.TOT) for Model
Quality (A) Superior and Period-Based Ductility,
.T . 3 ............................................................................... 7-12
Table 7-2b Total System Collapse Uncertainty (.TOT) for Model
Quality (B) Good and Period-Based Ductility, .T . 3 ..... 7-12
Table 7-2c Total System Collapse Uncertainty (.TOT) for Model
Quality (C) Fair and Period-Based Ductility, .T . 3 ....... 7-12
xxviii List of Tables FEMA P695
Table 7-2d Total System Collapse Uncertainty (.TOT) for Model
Quality (D) Poor and Period-Based Ductility, .T . 3 ...... 7-13
Table 7-3 Acceptable Values of Adjusted Collapse Margin Ratio
(ACMR10% and ACMR20%) ................................................ 7-14
Table 9-1 Performance Groups for Evaluation for Reinforced
Concrete Special Moment Frame Archetypes .................... 9-6
Table 9-2 Reinforced Concrete Special Moment Frame Archetype
Structural Design Properties ............................................... 9-8
Table 9-3 Summary of Collapse Results for Reinforced Concrete
Special Moment Frame Archetype Designs ..................... 9-16
Table 9-4 Spectral Shape Factors (SSF) for Archetypes Designed
for Seismic Design Categories B, C, or Dmin Seismic
Criteria (from Table 7-1a) ................................................ 9-17
Table 9-5 Spectral Shape Factors (SSF) for Archetypes Designed
for Seismic Design Categories Dmax Seismic Criteria
(from Table 7-1b) ............................................................. 9-18
Table 9-6 Total System Collapse Uncertainty (ßTOT) for Model
Quality (B) Good and Period-Based Ductility, .T . 3
(from Table 7-2b) ............................................................. 9-18
Table 9-7 Acceptable Values of Adjusted Collapse Margin Ratio
(ACMR10% and ACMR20%) (from Table 7-3)..................... 9-19
Table 9-8 Summary of Final Collapse Margins and Comparison
to Acceptance Criteria for Reinforced Concrete Special
Moment Frame Archetypes .............................................. 9-20
Table 9-9 Structural Design Properties for Reinforced Concrete
Special Moment Frame Archetypes Redesigned
Considering a Minimum Base Shear Requirement .......... 9-23
Table 9-10 Summary of Final Collapse Margins and Comparison to
Acceptance Criteria for Archetypes Redesigned with an
Updated Minimum Base Shear Requirement ................... 9-24
Table 9-11 Performance Groups for Evaluation of Reinforced
Concrete Ordinary Moment Frames ................................. 9-28
Table 9-12 Reinforced Concrete Ordinary Moment Frame
Archetype Design Properties, SDC B .............................. 9-31
Table 9-13 Summary of Pushover Analysis and IDA Sidesway
Collapse Results for Reinforced Concrete Ordinary
Moment Frame Archetype Designs, SDC B .................... 9-36
FEMA P695 List of Tables xxix
Table 9-14 Effect of Non-Simulated Collapse Modes on Computed
Collapse Margin Ratios for Reinforced Concrete
Ordinary Moment Frame Archetypes, SDC B ................. 9-37
Table 9-15 Summary of Collapse Margins and Comparison to
Acceptance Criteria for Reinforced Concrete Ordinary
Moment Frame Archetypes, SDC B ................................ 9-39
Table 9-16 Reinforced Concrete Ordinary Moment Frame Archetype
Design Properties for SDC C Seismic Criteria ................ 9-40
Table 9-17 Summary of Pushover Results, Collapse Margins, and
Comparison to Acceptance Criteria for Reinforced
Concrete Ordinary Moment Frame Archetypes, SDC C . 9-41
Table 9-18 Range of Variables Considered for the Definition of
Wood Light-Frame Archetype Buildings ........................ 9-46
Table 9-19 Performance Groups Used in the Evaluation of Wood
Light-Frame Buildings ..................................................... 9-47
Table 9-20 Wood Light-Frame Archetype Structural Design
Properties ......................................................................... 9-48
Table 9-21 Sheathing-to-Framing Connector Hysteretic Parameters
Used to Construct Shear Elements for Wood Light-
Frame Archetype Models ................................................. 9-50
Table 9-22 Summary of Collapse Results for Wood Light-Frame
Archetype Designs ........................................................... 9-54
Table 9-23 Adjusted Collapse Margin Ratios and Acceptable
Collapse Margin Ratios for Wood Light-Frame
Archetype Designs ........................................................... 9-56
Table 10-1 Model Parameters for Column and Beam Plastic Hinges
in 4-Story Steel Special Moment Frame .......................... 10-5
Table 10-2 Isolation System Design Properties ............................... 10-23
Table 10-3 Summary of Moat Wall Clearance (Gap) Distances ...... 10-24
Table 10-4 Isolated Structure Design Properties for Code-
Conforming Archetypes ................................................. 10-25
Table 10-5 Isolated Structure Design Properties for Non-Code-
Conforming Archetypes ................................................. 10-26
Table 10-6a Collapse Results for Code-Conforming Archetypes:
Various Gap Sizes .......................................................... 10-32
xxx List of Tables FEMA P695
Table 10-6b Collapse Results for Code-Conforming Archetypes:
Nominal (GEN), Upper-Bound (GEN-UB) and Lower-
Bound (GEN-LB) Isolator Properties ............................ 10-32
Table 10-6c Collapse Results for Code-Conforming Archetypes:
Minimum Seismic Criteria (SDC Dmin) .......................... 10-33
Table 10-7a Collapse Results for Isolated Archetypes with Ductile
Superstructures and Normalized Design Shear Values
(Vs/W) Not Equal to the Code-Required Value
(Vs/W = 0.092) ................................................................ 10-35
Table 10-7b Collapse Results for Isolated Archetypes with Non-
Conforming (Non-Ductile) Superstructures of Various
Normalized Design Shear Values (Vs/W) ....................... 10-35
Table A-1A Summary of Mapped Values of Short-Period Spectral
Accelerations, Site Coefficients, and Design Parameters
Used for Collapse Evaluation of Seismic Design
Categories D, C and B Structure Archetypes .................... A-5
Table A-1B Summary of Mapped Values of 1-Second Spectral
Accelerations, Site Coefficients, and Design Parameters
Used for Collapse Evaluation of Seismic Design
Categories D, C and B Structure Archetypes .................... A-5
Table A-2 Example values of the Fundamental Period, T, and
Corresponding MCE Spectral Acceleration, SMT, for
Reinforced Concrete Moment Frame Structures of
Various Heights ................................................................. A-7
Table A-3 Median 5%-Damped Spectral Accelerations of
Normalized Far-Field and Near-Field Record Sets
and Scaling Factors for Anchoring the Normalized
Far-Field Record Set to MCE Spectral Demand ............. A-12
Table A-4A Summary of Earthquake Event and Recording Station
Data for the Far-Field Record Set ................................... A-14
Table A-4B Summary of Site and Source Data for the Far-Field
Record Set ....................................................................... A-15
Table A-4C Summary of PEER NGA Database Information and
Parameters of Recorded Ground Motions for the
Far-Field Record Set ....................................................... A-16
Table A-4D Summary of Factors Used to Normalize Recorded
Ground Motions, and Parameters of Normalized
Ground Motions for the Far-Field Record Set ................ A-17
FEMA P695 List of Tables xxxi
Table A-5 Far-Field Record Set (as-Recorded and After
Normalization): Comparison of Maximum, Minimum
and Average Values of Peak Ground Acceleration
(PGAmax) and Peak Ground Velocity (PGVmax),
Respectively .................................................................... A-18
Table A-6A Summary of Earthquake Event and Recording Station
Data for the Near-Field Record Set ................................ A-21
Table A-6B Summary of Site and Source Data for the Near-Field
Record Set ....................................................................... A-22
Table A-6C Summary of PEER NGA Database Information and
Parameters of Recorded Ground Motions for the Near-
Field Record Set .............................................................. A-23
Table A-6D Summary of Factors Used to Normalize Recorded
Ground Motions, and Parameters of Normalized
Ground Motions for the Near-Field Record Set ............. A-24
Table A-7 Near-Field Record Set (As-Recorded and After
Normalization): Comparison of Maximum, Minimum
and Average Values of Peak Ground Acceleration
(PGAmax) and Peak Ground Velocity (PGVmax),
Respectively .................................................................... A-25
Table A-8 Near-Source Coefficients of the 1997 UBC (from
Tables 16-S and 16-T, ICBO, 1997) ............................... A-29
Table A-9 Summary of Key Reinforced-Concrete Special
Moment Frame Archetype Properties and Seismic
Coefficients Used to Evaluate the Collapse Margin
Ratio (CMR) .................................................................... A-30
Table A-10A Summary of Selected Collapse Margin Ratios (CMRs)
for Reinforced-Concrete Special Moment Frame
Archetypes – Comparison of CMRs for Far-Field and
Near-Field Record Sets ................................................... A-31
Table A-10B Summary of Selected Collapse Margin Ratios (CMRs)
for Reinforced-Concrete Special Moment Frame
Archetypes - Comparison of CMRs for the Near-Field
Record Set and the Near-Field Pulse FN Record
Subset .............................................................................. A-32
Table A-11 Summary of Design Properties and Collapse Margins
of the Three Reinforced Concrete Special Moment
Frame Building Archetypes Used to Evaluate
Far-Field Record Set Robustness .................................... A-34
xxxii List of Tables FEMA P695
Table A-12 Effects of the PGA Selection Criterion on the
Computed CMR Values for Three Reinforced Concrete
Special Moment Frame Buildings ................................... A-34
Table A-13 Effects of the PGA Selection Criterion on the
Approximate ACMR Values for Three Reinforced
Concrete Special Moment Frame Buildings ................... A-36
Table A-14 Effects of the PGV Selection Criterion on the
Computed CMR Values for Three RC SMF Buildings ... A-36
Table A-15 Effects of the PGA and PGV Selection Criteria on the
Computed CMR and ACMR Values, for Three
Reinforced Concrete Special Moment Frame Buildings,
as well as The Effects of Selecting Two Records from
Each Event ...................................................................... A-38
Table B-1 Tabulated 0 . Values for Various Seismic Design
Categories .......................................................................... B-9
Table B-2 Tabulated Spectral Demands for Various Seismic
Design Categories ........................................................... B-10
Table B-3 Documentation of Building Information and .1
Regression Results for the Set of Reinforced Concrete
Special Moment Frame Buildings. .................................. B-16
Table B-4 Documentation of Building Information and . 1
Regression Results for the Set of Reinforced Concrete
Ordinary Moment Frame Buildings ................................ B-17
Table B-5 Documentation of Building Information and . 1
Regression Results for the Set of 1967-era Reinforced
Concrete Buildings .......................................................... B-18
Table B-6 Documentation of Building Information and . 1
Regression Results for the Set of Wood Light-frame
Buildings ......................................................................... B-19
Table B-7 Spectral Shape Factors for Seismic Design
Categories B, C , and Dmin ............................................... B-23
Table B-8 Spectral Shape Factors for Seismic Design
Category Dmax .................................................................. B-23
Table B-9 Spectral Shape Factors for Seismic Design
Category E ....................................................................... B-24
Table C-1 Important Parameters, Related Physical Properties,
and Design Variables for Reinforced Concrete
Moment Frame Systems .................................................... C-2
FEMA P695 List of Tables xxxiii
Table C-2 Key Design Variables and Ranges Considered in the
Design Space for Reinforced Concrete Moment Frame
Systems .............................................................................. C-3
Table C-3 Matrix of Index Archetype Configurations for a
Reinforced Concrete Moment Frame System .................... C-7
Table C-4 Index Archetype Design Assumptions for a Reinforced
Concrete Moment Frame System ....................................... C-8
Table C-5 Index Archetype Configurations for Wood Light-
Frame Shear Wall Systems .............................................. C-11
Table C-6 Index Archetype Designs for Wood Light-Frame
Shear Wall Systems (R=6) ............................................... C-11
Table D-1 Possible Deterioration Modes for Reinforced
Concrete Moment Frame Components ............................. D-3
Table D-2a Collapse scenarios for Reinforced Concrete Moment
Frames – Sidesway Collapse ............................................ D-8
Table D-2b Collapse scenarios for Reinforced Concrete Moment
Frames – Vertical Collapse ............................................... D-8
Table D-3 Likelihood of Column Collapse Scenarios by Frame
Type (H: High, M: Medium, L: Low) ............................... D-9
Table E-1 Prediction Uncertainties and Bias in Proposed
Equations ......................................................................... E-12
Table E-2 Predicted Model Parameters for an 8-Story Reinforced
Concrete Special Moment Frame Perimeter System
(Interior Column, 1st-Story Location) ............................. E-13
FEMA P695 Introduction 1-1
Chapter 1
Introduction
This report describes a recommended methodology for reliably quantifying
building system performance and response parameters for use in seismic
design. The recommended methodology (referred to herein as the
Methodology) provides a rational basis for establishing global seismic
performance factors (SPFs), including the response modification coefficient
(R factor), the system overstrength factor (..), and deflection amplification
factor (Cd), of new seismic-force-resisting systems proposed for inclusion in
model building codes.
1.1 Background and Purpose
The Applied Technology Council (ATC) was commissioned by the Federal
Emergency Management Agency (FEMA) under the ATC-63 Project to
develop a methodology for quantitatively determining global seismic
performance factors for use in seismic design.
Seismic performance factors are used in current building codes and standards
to estimate strength and deformation demands on seismic-force-resisting
systems that are designed using linear methods of analysis, but are
responding in the nonlinear range. R factors were initially introduced in the
ATC-3-06 report, Tentative Provisions for the Development of Seismic
Regulations for Buildings (ATC, 1978), and their values have become
fundamentally critical in the specification of design seismic loading.
Since then, the number of structural systems addressed in seismic codes has
increased dramatically. The 2003 edition of the National Earthquake
Hazards Reduction Program (NEHRP) Recommended Provisions for Seismic
Regulations for New Buildings and Other Structures (NEHRP Recommended
Provisions), (FEMA, 2004a), includes more than 75 individual systems, each
having a somewhat arbitrarily assigned R factor.
Original R factors were based largely on judgment and qualitative
comparisons with the known response capabilities of relatively few seismicforce-
resisting systems in widespread use at the time. Many recently defined
seismic-force-resisting systems have never been subjected to any significant
level of earthquake ground shaking. As a result, the seismic response
characteristics of many systems, and their ability to meet seismic design
performance objectives, are both untested and unknown.
1-2 Introduction FEMA P695
As new systems continue to be introduced during each code update cycle,
uncertainty in the seismic performance capability of the new building stock
continues to grow, and the need to quantify the seismic performance
delivered by current seismic design regulations becomes more urgent.
Advances in performance-based seismic design tools and technologies has
resulted in the ability to use nonlinear collapse simulation techniques to link
seismic performance factors to system performance capabilities on a
probabilistic basis.
The purpose of this Methodology is to provide a rational basis for
determining building system performance and response parameters that,
when properly implemented in the seismic design process, will result in
equivalent safety against collapse in an earthquake, comparable to the
inherent safety against collapse intended by current seismic codes, for
buildings with different seismic-force-resisting systems.
The Methodology is recommended for use with model building codes and
resource documents to set minimum acceptable design criteria for standard
code-approved seismic-force-resisting systems, and to provide guidance in
the selection of appropriate design criteria for other systems when linear
design methods are applied. It also provides a basis for evaluation of current
code-approved systems for their ability to achieve intended seismic
performance objectives. It is possible that results of future work based on
this Methodology could be used to modify or eliminate those systems or
requirements that cannot reliably meet these objectives.
1.2 Scope and Basis of the Methodology
The following key principles outline the scope and basis of the Methodology.
1.2.1 Applicable to New Building Structural Systems
The Methodology applies to the determination of seismic performance
factors appropriate for the design of seismic-force-resisting systems in new
building structures. While the Methodology is conceptually applicable (with
some limitations) to design of non-building structures, and to retrofit of
seismic-force-resisting systems in existing buildings, such systems were not
explicitly considered. The Methodology is not intended to apply to the
design of nonstructural systems.
1.2.2 Compatible with the NEHRP Recommended Provisions
and ASCE/SEI 7
The Methodology is based on, and intended for use with, applicable design
criteria and requirements of the most current editions of the NEHRP
FEMA P695 Introduction 1-3
Recommended Provisions for Seismic Regulations for New Buildings and
Other Structures (NEHRP Recommended Provisions), (FEMA, 2004a), and
the seismic provisions of ASCE/SEI 7-05, Minimum Design Loads for
Buildings and Other Structures, (ASCE, 2006a). The Building Seismic
Safety Council has adopted ASCE/SEI 7-05 as the “starting point” for the
development of its 2009 and future editions of the NEHRP Recommended
Provisions. At this time, ASCE/SEI 7-05 is the most current, published
source of seismic regulations for model building codes in the United States.1
ASCE/SEI 7-05 provides the basis for ground motion criteria and “generic”
structural design requirements applicable to currently accepted and future
(proposed) seismic-force-resisting systems. ASCE/SEI 7-05 provisions
include detailing requirements for currently approved systems that may also
apply to new systems. By reference, other standards, such as ACI 318,
Building Code Requirements for Structural Concrete (ACI, 2005),
AISC/ANSI 341, Seismic Provisions for Structural Steel Buildings (AISC,
2005), ACI 530/ASCE 5/TMS 402, Building Code Requirements for
Masonry Structures (ACI, 2002b), and ANSI/AF&PA, National Design
Specification for Wood Construction (ANSI/AF&PA, 2005) apply to
currently approved systems, and may also apply to new systems.
The Methodology requires the seismic-force-resisting system of interest to
comply with all applicable design requirements in ASCE/SEI 7-05, including
limits on system irregularity, drift, and height, except when such
requirements are specifically excluded and explicitly evaluated in the
application of the Methodology. For new (proposed) systems, the
Methodology requires identification and use of applicable structural design
and detailing requirements in ASCE/SEI 7-05, and development and use of
new requirements as necessary to adequately describe system limitations and
ensure predictable seismic behavior of components. The latest edition of the
NEHRP Recommended Provisions, containing modifications and
commentary to ASCE/SEI 7-05, may be a possible source for additional
design requirements.
1.2.3 Consistent with the Life Safety Performance Objective
The Methodology is consistent with the primary “life safety” performance
objective of seismic regulations in model building codes. As stated in the
Part 2: Commentary to the NEHRP Recommended Provisions for Seismic
1 This chapter and other sections of this document refer to ASCE/SEI 7-05 for
design criteria and requirements to illustrate the Methodology, and to define the
values of certain parameters used for performance evaluation. The Methodology
is intended to be generally applicable, and such references should not be construed
as limiting the Methodology to this edition of ASCE/SEI 7.
1-4 Introduction FEMA P695
Regulations for New Buildings and Other Structures (Commentary to the
NEHRP Recommended Provisions), (FEMA, 2004b), “the Provisions
provides the minimum criteria considered prudent for protection of life safety
in structures subject to earthquakes.”
Design for performance other than life safety was not explicitly considered in
the development of the Methodology. Accordingly, the Methodology does
not address special performance or functionality objectives of ASCE/SEI 7-
05 for Occupancy III and IV structures.
1.2.4 Based on Acceptably Low Probability of Structural
Collapse
The Methodology achieves the primary life safety performance objective by
requiring an acceptably low probability of collapse of the seismic-forceresisting
system when subjected to Maximum Considered Earthquake (MCE)
ground motions.
In general, life safety risk (i.e., probability of death or life-threatening injury)
is difficult to calculate accurately due to uncertainty in casualty rates given
collapse, and even greater uncertainty in assessing the effects of falling
hazards in the absence of collapse. Collapse of a structure can lead to very
different numbers of fatalities depending on variations in construction or
occupancy, such as structural system type and the number of building
occupants. Rather than attempting to quantify uniform protection of “life
safety”, the Methodology provides approximate uniform protection against
collapse of the structural system.
Collapse includes both partial and global instability of the seismic-forceresisting
system, but does not include local failure of components not
governed by global seismic performance factors, such as localized out-ofplane
failure of wall anchorage and potential life-threatening failure of nonstructural
systems.
Similarly, the Methodology does not explicitly address components that are
not included in the seismic-force-resisting system (e.g., gravity system
components and nonstructural components). It assumes that deformation
compatibility and related requirements of ASCE/SEI 7-05 adequately protect
such components against premature failure. Components that are not
designated as part of the seismic-force-resisting system are not controlled by
seismic-force-resisting system design requirements. Accordingly, they are
not considered in evaluating the overall resistance to collapse.
FEMA P695 Introduction 1-5
1.2.5 Earthquake Hazard based on MCE Ground Motions
The Methodology evaluates collapse under Maximum Considered
Earthquake (MCE) ground motions for various geographic regions of
seismicity, as defined by the coefficients and mapped acceleration parameters
of the general procedure of ASCE/SEI 7-05, which is based on the maps and
procedures contained in the NEHRP Recommended Provisions.
While seismic performance factors apply to the design response spectrum,
taken as two-thirds of the MCE spectrum, code-defined MCE ground
motions are considered the appropriate basis for evaluating structural
collapse. As noted in the Commentary to the NEHRP Recommended
Provisions, “if a structure experiences a level of ground motion 1.5 times the
design level, the structure should have a low likelihood of collapse.”
1.2.6 Concepts Consistent with Current Seismic Performance
Factor Definitions
The Methodology remains true to the definitions of seismic performance
factors given in ASCE/SEI 7-05, and the underlying nonlinear static analysis
(pushover) concepts described in the Commentary to the NEHRP
Recommended Provisions. Values of the response modification coefficient,
R factor, the system overstrength factor, .O , and the deflection amplification
factor, Cd, for currently approved seismic-force-resisting systems are
specified in Table 12.2-1 of ASCE/SEI 7-05. Section 4.2 of the Commentary
to the NEHRP Recommended Provisions provides background information
on seismic performance factors.
Figures 1-1 and 1-2 are used to explain and illustrate seismic performance
factors, and how they are used in the Methodology. Parameters are defined
in terms of equations, which in all cases are dimensionless ratios of force,
acceleration or displacement. However, in attempting to utilize the figures to
clarify and to illustrate the meanings of these ratios, graphical license is taken
in two ways. First, seismic performance factors are depicted in the figures as
incremental differences between two related parameters, rather than as ratios
of the parameters. Second, as a consequence of being depicted as
incremental differences, seismic performance factors are shown on plots with
units, when, in fact, they are dimensionless.
Figure 1-1, an adaptation of Figures C4.2-1 and C4.2-3 from the
Commentary to the NEHRP Recommended Provisions, defines seismic
performance factors in terms of the global inelastic response (idealized
pushover curve) of the seismic-force-resisting system. In this figure, the
1-6 Introduction FEMA P695
horizontal axis is lateral displacement (i.e., roof drift) and the vertical axis is
lateral force at the base of the system (i.e., base shear).
In Figure 1-1, the term VE represents the force level that would be developed
in the seismic-force-resisting system, if the system remained entirely linearly
elastic for design earthquake ground motions. The term Vmax represents the
actual, maximum strength of the fully-yielded system, and the term V is the
seismic base shear required for design. As defined in Equation 1-1, the R
factor is the ratio of the force level that would be developed in the system for
design earthquake ground motions (if the system remained entirely linearly
elastic) to the base shear prescribed for design:
V
R . VE (1-1)
and, as defined in Equation 1-2, the .O factor is the ratio of the maximum
strength of the fully-yielded system to the design base shear:
V
Vmax
O . . (1-2)
.
Lateral Displacement (Roof Drift)
Lateral Seismic Force (Base Shear)
Pushover
Curve
R
R = Response Modification
Coefficient = VE/V
Cd
Cd = Deflection Amplification
Factor = (./.E)R
.0
.O = Overstrength Factor = Vmax/V
Vmax
VE
V
.E/R .E
Design Earthquake
Ground Motions
.
Lateral Displacement (Roof Drift)
Lateral Seismic Force (Base Shear)
Pushover
Curve
R
R = Response Modification
Coefficient = VE/V
Cd
Cd = Deflection Amplification
Factor = (./.E)R
.0
.O = Overstrength Factor = Vmax/V
Vmax
VE
V
.E/R .E
Design Earthquake
Ground Motions
Figure 1-1 Illustration of seismic performance factors (R, .O, and Cd) as
defined in the Commentary to the NEHRP Recommended
Provisions (FEMA, 2004b).
In Figure 1-1, the term .E/R represents roof drift of the seismic-forceresisting
system corresponding to design base shear, V, assuming that the
system remains essentially elastic for this level of force, and the term .
represents the assumed roof drift of the yielded system corresponding to
design earthquake ground motions. As illustrated in the figure and defined
FEMA P695 Introduction 1-7
by Equation 1-3, the Cd factor is some fraction of the R factor (typically less
than 1.0):
C R
E
d .
.
. (1-3)
The Methodology develops seismic performance factors consistent with the
concepts and definitions described above. Figure 1-2 illustrates the seismic
performance factors defined by the Methodology and their relationship to
MCE ground motions.
Collapse Level
Ground Motions
Spectral Displacement
Spectral Acceleration (g)
T
Smax
SDMT/1.5R
Cs
SDMT
MCE Ground Motions
(ASCE 7-05)
SMT
1.5R
1.5Cd
.
CMR
CMR
CT Sˆ
CT SD
Collapse Level
Ground Motions
Spectral Displacement
Spectral Acceleration (g)
T
Smax
SDMT/1.5R
Cs
SDMT
MCE Ground Motions
(ASCE 7-05)
SMT
1.5R
1.5Cd
.
CMR
CMR
CT Sˆ
CT SD
Figure 1-2 Illustration of seismic performance factors (R, ., and Cd) as
defined by the Methodology.
Figure 1-2 parallels the “pushover” concept shown in Figure 1-1 using
spectral coordinates rather than lateral force (base shear) and lateral
displacement (roof drift) coordinates. Conversion to spectral coordinates is
based on the assumption that 100% of the effective seismic weight of the
structure, W, participates in fundamental mode at period, T, consistent with
Equation 12.8-1 of ASCE/SEI 7-05:
V C W s . (1-4)
In Figure 1-2, the term SMT is the Maximum Considered Earthquake (MCE)
spectral acceleration at the period of the system, T, the term Smax represents
the maximum strength of the fully-yielded system (normalized by the
effective seismic weight, W, of the structure), and the term Cs is the seismic
response coefficient. As defined in Equation 1-5, the ratio of the MCE
1-8 Introduction FEMA P695
spectral acceleration to the seismic response coefficient, which is the designlevel
acceleration, is equal to 1.5 times the R factor:
s
MT
C
1.5R . S (1-5)
The 1.5 factor in Equation 1-5 accounts for the definition of design
earthquake ground motions in ASCE/SEI 7-05, which is two-thirds of MCE
ground motions.
In Figure 1-2, the overstrength parameter, ., is defined as the ratio of the
maximum strength of the fully-yielded system, Smax (normalized by W), to the
seismic response coefficient, Cs:
s
max
C
S . . (1-6)
The Methodology calculates the overstrength parameter, ., based on
nonlinear static (pushover) analysis. Calculated values of overstrength, .,
are different from the overstrength factor, .O, of ASCE/SEI 7-05, which is
specified for design of non-ductile elements. In general, different designs of
the same system will have different calculated values of overstrength, and the
parameter, ., will vary. The single value of . that is considered to be most
appropriate for use in design of the system of interest, is the value ultimately
selected for .O.
In Figure 1-2, inelastic system displacement at the MCE level is defined as
1.5Cd times the displacement corresponding to the design seismic response
coefficient, Cs, and set equal to the MCE elastic system displacement, SDMT
(based on the “Newmark rule”), effectively redefining the Cd factor to be
equal to the R factor:
C R d . (1-7)
The equal displacement assumption is reasonable for most conventional
systems with effective damping approximately equal to the nominal 5% level
used to define response spectral acceleration and displacement. Systems
with substantially higher (or lower) levels of damping would have
significantly smaller (or larger) displacements than those with 5%-damped
elastic response. As one example, systems with viscous dampers have
significantly higher damping than 5%. For such systems, the response
modification methods of Chapter 18 of ASCE/SEI 7-05 are used to determine
an appropriate value of the Cd factor, as a fraction of the R factor.
FEMA P695 Introduction 1-9
1.2.7 Safety Expressed in Terms of Collapse Margin Ratio
The Methodology defines collapse level ground motions as the intensity that
would result in median collapse of the seismic-force-resisting system.
Median collapse occurs when one-half of the structures exposed to this
intensity of ground motion would have some form of life-threatening
collapse. As shown in Figure 1-2, collapse level ground motions are higher
than MCE ground motions. As such, MCE ground motions would result in a
comparatively smaller probability of collapse. As defined in Equation 1-8,
the collapse margin ratio, CMR, is the ratio of the median 5%-damped
spectral acceleration of the collapse level ground motions, ˆ
CT S (or
corresponding displacement, CT SD ), to the 5%-damped spectral acceleration
of the MCE ground motions, MT S (or corresponding displacement, MT SD ), at
the fundamental period of the seismic-force-resisting system:
MT
CT
MT
CT
SD
SD
S
Sˆ
CMR . . (1-8)
In one sense, the collapse margin ratio, CMR, could be thought of as the
amount MT S must be increased to achieve building collapse by 50% of the
ground motions. Collapse of the seismic-force-resisting system, and hence
CMR, is influenced by many factors, including ground motion variability and
uncertainty in design, analysis, and construction of the structure. These
factors are considered collectively in a collapse fragility curve that describes
the probability of collapse of the seismic-force-resisting system as a function
of the intensity of ground motion.
1.2.8 Performance Quantified Through Nonlinear Collapse
Simulation on a Set of Archetype Models
The Methodology determines the response modification coefficient, R factor,
and evaluates the system over-strength factor, ., using nonlinear models of
seismic-force-resisting system “archetypes.” Archetypes capture the essence
and variability of the performance characteristics of the system of interest.
The Methodology requires nonlinear analysis of a sufficient number of
archetype models, with parametric variations in design parameters, to
broadly represent the system of interest.
The Methodology requires archetype models to meet the applicable design
requirements of ASCE/SEI 7-05 and related standards, and additional criteria
developed for the system of interest. Archetype design assumes a trial value
of the R factor to determine the seismic response coefficient, Cs. The
Methodology requires detailed modeling of nonlinear behavior of archetypes,
based on representative test data sufficient to capture collapse failure modes.
1-10 Introduction FEMA P695
Collapse failure modes that cannot be explicitly modeled are evaluated using
appropriate limits on the controlling response parameter.
1.2.9 Uncertainty Considered in Performance Evaluation
The Methodology defines acceptable values of the collapse margin ratio in
terms of an acceptably low probability of collapse for MCE ground motions,
given uncertainty in the collapse fragility. Systems that have more robust
design requirements, more comprehensive test data, and more detailed
nonlinear analysis models, have less collapse uncertainty, and can achieve
the same level of life safety with smaller collapse margin ratios.
Calculated values of collapse margin ratio are compared with acceptable
values that reflect collapse uncertainty. If the calculated collapse margin is
large enough to meet the performance objective (i.e., an acceptably small
probability of collapse at the MCE), then the trial value of the R factor used
in the archetype design is acceptable. If not, a new (lower) trial value of the
R factor must be re-evaluated using the Methodology, or other limitations on
the system of interest (e.g., height restrictions in the design requirements)
must be considered.
1.3 Content and Organization
This report is written and organized to facilitate potential use and adoption
by the NEHRP Recommended Provisions. Chapter 2 provides an overview
of the Methodology, introducing the basic theory and concepts that are
described in more detail in the chapters that follow.
Chapters 3 through 7 step through the elements of the Methodology,
including required system information, structure archetype development,
nonlinear modeling, criteria for collapse assessment, nonlinear analysis, and
evaluation of seismic performance factors.
Chapter 8 defines documentation and peer review requirements, and
describes recommended qualifications, expertise, and responsibilities for
personnel involved with implementing the Methodology in the development
and review of a proposed system.
Chapter 9 provides example applications intended to assist users in
implementing the Methodology, and to validate the technical approach.
Example systems include special and ordinary reinforced concrete moment
frame systems, and wood light-frame systems.
Chapter 10 includes supporting studies on non-simulated collapse failure
modes for steel moment frame systems, and on dynamic response
FEMA P695 Introduction 1-11
characteristics, performance properties, and collapse failure modes unique to
seismically-isolated structures.
Chapter 11 provides summary conclusions, recommendations, and
limitations on the use of the Methodology.
Appendices A through F provide background information supporting the
development of the Methodology, and expanded guidance on key aspects of
the Methodology.
A glossary of definitions and list of symbols used throughout this report,
along with a list of references, are provided at the end of this report.
FEMA P695 2: Overview of Methodology 2-1
Chapter 2
Overview of Methodology
This chapter outlines the general framework of the Methodology and
describes the overall process. It introduces the key elements of the
Methodology, including required system information, development of
structural system archetypes, archetype models, nonlinear analysis of
archetypes, performance evaluation, and documentation and peer review
requirements. These elements are specified in more detail in the chapters that
follow.
2.1 General Framework
The Methodology consists of a framework for establishing seismic
performance factors (SPFs) that involves the development of detailed system
design information and probabilistic assessment of collapse risk. It utilizes
nonlinear analysis techniques, and explicitly considers uncertainties in
ground motion, modeling, design, and test data. The technical approach is a
combination of traditional code concepts, advanced nonlinear dynamic
analyses, and risk-based assessment techniques.
Reliable analysis requires valid ground motions and representative nonlinear
models of the seismic-force-resisting system. Development of representative
models requires both detailed design information and comprehensive
nonlinear test data on structural components and assemblies that make up the
system of interest. Figure 2-1 illustrates the key elements of the
Methodology.
The Methodology includes fully defined characterizations of ground motion
and methods of analysis that are generically applicable to all seismic-forceresisting
systems. Design information and test data will be different for each
system, and may not yet exist for new systems. The Methodology includes
requirements for defining the type of design information and test data that are
needed for developing representative analytical models of the seismic-forceresisting
system of interest.
Rather than establishing minimum requirements for design information and
test data, the use of better quality information is encouraged by rewarding
systems that have “done their homework.” Systems that are based on welldefined
design requirements and comprehensive test data will have
2-2 2: Overview of Methodology FEMA P695
inherently less uncertainty in their seismic performance. Such systems will
need a lower margin against collapse to achieve an equivalent level of safety,
as compared to systems with less robust data.
Due to the complexity of nonlinear dynamic analysis, the difficulty in
modeling inelastic behavior, and the need to verify the adequacy and quality
of design information and test data, the Methodology requires independent
peer review of the entire process.
Methodology
Peer Review
Requirements
Ground
Motions
Analysis
Methods
Test Data
Requirements
Design Information
Requirements
Methodology
Peer Review
Requirements
Ground
Motions
Analysis
Methods
Test Data
Requirements
Design Information
Requirements
Figure 2-1 Key elements of the Methodology.
2.2 Description of Process
The steps comprising the Methodology are shown in Figure 2-2. These steps
outline a process for developing system design information with enough
detail and specificity to identify the permissible range of application for the
proposed system, adequately simulate nonlinear response, and reliably assess
the collapse risk over the proposed range of applications. Each step is linked
to a corresponding chapter in this report, and described in the sections that
follow.
2.3 Develop System Concept
The process begins with the development of a well-defined concept for the
seismic-force-resisting system, including type of construction materials,
system configuration, inelastic dissipation mechanisms, and intended range
of application.
FEMA P695 2: Overview of Methodology 2-3
Figure 2-2 Process for quantitatively establishing and documenting seismic
performance factors (SPFs).
The amount of documentation necessary to describe the system and
characterize system components will vary, depending on the novelty and
uniqueness of the proposed system relative to other well-established
structural systems.
2.4 Obtain Required Information
Required system information is specified in Chapter 3. Required information
includes detailed design requirements and results from material, component,
and system testing. Design requirements include the rules that engineers will
use to proportion and detail structural components of the system, and limits
in the application of the system. Test results include information on
2-4 2: Overview of Methodology FEMA P695
component material properties, force-deformation behavior, and nonlinear
response.
Comprehensive design provisions are developed within the context of the
seismic provisions of ASCE/SEI 7-05 and other applicable standards. The
provisions should address all significant aspects of the design and detailing
of the seismic-force-resisting system and its components. Important
exceptions and deviations from established building code requirements
should be clearly stipulated. Design provisions should address criteria for
determining minimum strength and ensuring inelastic deformation capacity
through a combination of system design requirements, component design and
detailing requirements, and project-specific testing requirements. Design
provisions should also specify the seismic performance factors (R , .., Cd)
and other criteria (e.g., drift limits, height limits, and seismic usage
restrictions) that are proposed as part of the design basis for the new system.
Test data are necessary for characterizing the strength, stiffness and ductility
of the materials, members, and connections of the proposed system. Test
data are also necessary for establishing properties of the nonlinear analysis
models used to assess collapse risk. Test data and other substantiating
evidence should be acquired as the basis of the design provisions and for
calibrating analysis models. Design requirements should be documented
with supporting evidence to ensure sufficient strength, stiffness and ductility
of the proposed system, across the intended range of application of the
system.
2.5 Characterize Behavior
System behavior is characterized through the use of structural system
archetypes. The concept of an archetype is described in Chapter 4.
Establishment of archetypes begins with identifying the range of features and
behavioral characteristic that describe the bounds of the proposed seismicforce-
resisting system.
Archetypes provide a systematic means for characterizing permissible
configurations and other significant features of the proposed system. Like
building code provisions, archetypical systems are intended to represent
typical applications of a seismic-force-resisting system, recognizing that it is
practically impossible to envision or attempt to quantify performance of all
possible applications. They should, however, reflect the degree of
irregularity permitted within standard building code provisions.
The challenge in defining and assessing structural system archetypes is in
narrowing the range of parameters and attributes to the fewest and simplest
FEMA P695 2: Overview of Methodology 2-5
possible, while still being reasonably representative of the variations that
would be permitted in actual structures. In addition to ground motion
intensity (Seismic Design Category), the following characteristics are
considered in defining structural system archetypes: (1) building height; (2)
fundamental period; (3) structural framing configurations; (4) framing bay
sizes or wall lengths; (5) magnitude of gravity loads; and (6) member and
connection design and detailing requirements. Structural system archetypes
are assembled into bins called performance groups, which reflect major
divisions, or changes in behavior, within the archetype design space. The
collapse safety of the proposed system is then evaluated for each
performance group.
In the collapse assessment process, only framing components that are
specifically designated as part of the seismic-force-resisting system are
included in the archetypes. While it is recognized that other portions of the
building (e.g., components of the gravity system or certain nonstructural
components) can significantly affect collapse behavior, such components,
which are not controlled by seismic-force-resisting system design
requirements, cannot be relied upon for reducing collapse risk.
2.6 Develop Models
Development of structural models for collapse assessment is discussed in
Chapter 5. Structural system archetypes provide the basis for preparing a
finite number of trial designs and developing a corresponding number of
idealized nonlinear models that sufficiently represent the range of intended
applications for a proposed system. Index archetype models are developed to
provide the most basic (generic) idealization of an archetypical
configuration, while still capturing significant behavioral modes and key
design features of the proposed seismic-force-resisting system.
Designs consider the range of seismic criteria for each applicable Seismic
Design Category, variations in gravity loads, and other distinguishing
features including alternative geometric configurations, varying heights, and
different tributary areas that impact seismic design or system performance.
To the extent possible, nonlinear models include explicit simulation of all
significant deterioration mechanisms that could lead to structural collapse.
Recognizing that it is not always possible (or practical) to simulate all
possible collapse modes, the Methodology includes provisions for assessing
the effects of behaviors that are not explicitly simulated in the model, but
could trigger collapse.
2-6 2: Overview of Methodology FEMA P695
Nonlinear models must account for the seismic mass that is stabilized by the
seismic-force-resisting system, including the destabilizing P-delta effects
associated with the seismic mass. In most cases, elements are idealized with
phenomenological models to simulate complicated component behavior. In
some cases, however, two-dimensional or three-dimensional continuum finite
element models may be required to properly characterize behavior. Models
are calibrated using material, component, or assembly test data and other
substantiating evidence to verify their ability to simulate expected nonlinear
behavior.
2.7 Analyze Models
Collapse assessment is performed using both nonlinear static (pushover) and
nonlinear dynamic (response history) analysis procedures described in
Chapter 6. Nonlinear static analyses are used to help validate the behavior of
nonlinear models and to provide statistical data on system overstrength and
ductility capacity. Nonlinear dynamic analyses are used to assess median
collapse capacities and collapse margin ratios.
Nonlinear response is evaluated for a set of pre-defined ground motions that
are used for collapse assessment of all systems. Two sets of ground motion
records are provided for nonlinear dynamic analysis. One set includes 22
ground motion record pairs from sites located greater than or equal to 10 km
from fault rupture, referred to as the “Far-Field” record set. The other set
includes 28 pairs of ground motions recorded at sites less than 10 km from
fault rupture, referred to as the “Near-Field” record set. While both Far-Field
and Near-Field record sets are provided, only the Far-Field record set is
required for collapse assessment. This is done for reasons of practicality, and
in recognition of the fact that there are many unresolved issues concerning
the characterization of near-fault hazard and ground motion effects. The
Near-Field record set is provided as supplemental information to examine
issues that could arise due to near-fault directivity effects, if needed.
The record sets include records from all large-magnitude events in the Pacific
Earthquake Engineering Research Center (PEER) Next-Generation
Attenuation (NGA) database (PEER, 2006a). Records were selected to meet
a number of sometimes conflicting objectives. To avoid event bias, no more
than two of the strongest records have been taken from any one earthquake,
yet the record sets have a sufficient number of motions to permit statistical
evaluation of record-to-record (RTR) variability and collapse fragility.
Strong ground motions were not distinguished based on either site condition
or source mechanism. The Far-Field and Near-Field record sets are provided
in Appendix A, along with background information on their selection.
FEMA P695 2: Overview of Methodology 2-7
For collapse evaluation, ground motions are systematically scaled to
increasing earthquake intensities until median collapse is established.
Median collapse is the ground motion intensity in which half of the records
in the set cause collapse of an index archetype model. This process is similar
to, but distinct from the concept of incremental dynamic analysis (IDA), as
proposed by Vamvatsikos and Cornell (2002).
Figure 2-3 shows an example of IDA results for a single structure subjected
to a suite of ground motions of varying intensities. In this illustration,
sidesway collapse is the governing mechanism, and collapse prediction is
based on lateral dynamic instability, or excessive lateral displacements.
Using collapse data obtained from IDA results, a collapse fragility can be
defined through a cumulative distribution function (CDF), which relates the
ground motion intensity to the probability of collapse (Ibarra et al., 2002).
Figure 2-4 shows an example of a cumulative distribution plot obtained by
fitting a lognormal distribution to the collapse data from Figure 2-3.
While the IDA concept is useful for illustrating the collapse assessment
procedure, the Methodology only requires calculation of the median collapse
point, which can be calculated with fewer nonlinear analyses than would
otherwise be required to calculate the full IDA curve. An abbreviated
process for calculating the median collapse point is described in Chapter 6.
Figure 2-3 Incremental dynamic analysis response plot of spectral
acceleration versus maximum story drift ratio.
2-8 2: Overview of Methodology FEMA P695
Figure 2-4 Collapse fragility curve, or cumulative distribution function.
2.8 Evaluate Performance
The performance evaluation process is described in Chapter 7. It utilizes
results from nonlinear static analyses to determine an appropriate value of the
system overstrength factor, .O, and results from nonlinear dynamic analyses
to evaluate the acceptability of a trial value of the response modification
coefficient, R. The deflection amplification factor, Cd, is derived from an
acceptable value of R, with consideration of the effective damping of the
system of interest.
The trial value of the response modification coefficient, R, is evaluated in
terms of the acceptability of a calculated collapse margin ratio, which is the
ratio of the ground motion intensity that causes median collapse, to the
Maximum Considered Earthquake (MCE) ground motion intensity defined
by the building code. Acceptability is measured by comparing the collapse
margin ratio, after some adjustment, to acceptable values that depend on the
quality of information used to define the system, total system uncertainty,
and established limits on acceptable probabilities of collapse.
To account for unique characteristics of extreme ground motions that lead to
building collapse, the collapse margin ratio is converted to an adjusted
collapse margin ratio. The adjustment is based on the shape of the spectrum
of rare ground motions, and is a function of the structure ductility and period
of vibration. Systems with larger ductility and longer periods benefit by
larger adjustments. The background and development of this adjustment to
account for the effects of spectral shape are provided in Appendix B.
Acceptable values of the collapse margin ratio are defined in terms of an
acceptably low probability of collapse for MCE ground motions, considering
uncertainty in collapse fragility. Systems that have more robust design
FEMA P695 2: Overview of Methodology 2-9
requirements, more comprehensive test data, and more detailed nonlinear
analysis models, have less collapse uncertainty, and can achieve the same
level of life safety with smaller collapse margin ratios. The following
sources of uncertainty are explicitly considered: (1) record-to-record
uncertainty; (2) design requirements-related uncertainty; (3) test data-related
uncertainty; and (4) modeling uncertainty.
The probability of collapse due to MCE ground motions applied to a
population of archetypes is limited to 10%, on average. Each performance
group is required to meet this average limit, recognizing that some individual
archetypes could have collapse probabilities that exceed this value. The
probability of collapse for individual archetypes is limited to 20%, or twice
the average value, to evaluate acceptability of potential “outliers” within a
performance group. It should be noted that these limits were selected based
on judgment. Within the performance evaluation process, these values can
be adjusted to reflect different values of acceptable probabilities of collapse
that are deemed appropriate by governing jurisdictions or other authorities
employing this Methodology to establish seismic design requirements for a
proposed system.
If the adjusted collapse margin ratio is large enough to result in an acceptably
small probability of collapse at the MCE, then the trial value of R is
acceptable. If not, the system must be redefined by adjusting the design
requirements (Chapter 3), re-characterizing behavior (Chapter 4), or
redesigning with new trial values (Chapter 5), and then re-evaluated using
the Methodology. In some cases, inadequate performance could require
extensive revisions to the overall system concept.
2.9 Document Results
Documentation requirements are described in Chapter 8. The results of
system development efforts must be thoroughly documented for review and
approval by an independent peer review panel, review and approval by an
authority having jurisdiction, and eventual use in design and construction.
Documentation is required at each step of the process. It should describe
seismic design rules, range of applicability of the system, testing protocols
and results, rationale for the selection of structural system archetypes, results
of analytical investigations, evaluation of quality of information,
quantification of uncertainties, results of performance evaluations, and
proposed seismic performance factors.
Documentation should be of sufficient detail and clarity to allow an
unfamiliar structural engineer to properly implement the design, and an
2-10 2: Overview of Methodology FEMA P695
unfamiliar reviewer to evaluate compliance with the design requirements.
Documentation should also provide sufficient information to allow peer
reviewers, code authorities, or material standard organizations to assess the
viability of the proposed system and the reasonableness of the proposed
seismic performance factors.
2.10 Peer Review
Peer review by an independent team of experts is a requirement of the
Methodology, and should be an integral part of the process at each step.
Implementation of the Methodology involves much uncertainty, judgment
and potential for variation. Deciding on an appropriate level of detail to
adequately characterize performance of a proposed system should be
performed in collaboration with a peer review panel, on an ongoing basis,
during developmental efforts.
The peer review panel is responsible for reviewing and commenting on the
approach taken by the system development team, including the extent of the
experimental program, testing procedures, design requirements, development
of structural system archetypes, analytical approaches, extent of the nonlinear
analysis investigation, and the final selection of the proposed seismic
performance factors. Members of the peer review panel must be qualified to
critically evaluate the development of the proposed system including testing,
design, and analysis, and sufficiently independent from the system
development team to provide an unbiased assessment of the developmental
process.
The peer review panel, and their involvement, should be established early to
clarify expectations for the collapse assessment. The peer review team is
expected to exercise considerable judgment in evaluating all aspects of the
process, from definition of the proposed system, to establishment of design
criteria, scope of testing, and extent of analysis deemed necessary to
adequately evaluate collapse safety.
Details on the required peer review process, and guidance on the selection of
peer review panel members, are provided in Chapter 8.
FEMA P695 3: Required System Information 3-1
Chapter 3
Required System Information
This chapter identifies information that is necessary for establishing seismic
performance factors as part of the development, documentation, and review
of a proposed seismic-force-resisting system. It describes the type of
information that is required, and provides guidance on how it should be
developed.
This information is used in the development of structural system archetypes
in Chapter 4, and nonlinear analysis models in Chapter 5. It is subject to peer
review as it is developed, and is an integral part of the reporting requirements
in Chapter 8.
3.1 General
Seismic performance factors for a proposed system are established through
nonlinear simulation of response to earthquake ground motion, and
probabilistic assessment of collapse risk. Detailed system information is
necessary for reliable prediction of structural response, and for development
and validation of standardized engineering criteria that will lead to structures
that perform as expected. Required system information includes:
. a comprehensive description of the proposed system, including its
intended applications, physical and behavioral characteristics, and
construction methods;
. a clear and complete set of design requirements and specifications for the
system that provide information to quantify strength limit states,
proportion and detail components, analyze predicted response, and
confirm satisfactory behavior; and
. test data and other supporting evidence from an experimental
investigation program to validate material properties and component
behavior, calibrate nonlinear analysis models, and establish performance
acceptance criteria.
The process for obtaining required system information is shown in Figure
3-1. It involves development of detailed system design requirements,
acquisition of test data, and assessment of the quality of this information.
3-2 3: Required System Information FEMA P695
Figure 3-1 Process for obtaining required system information
Design requirements and test data are used as inputs for the development of
structural system archetypes in Chapter 4. Quality ratings for design
requirements and test data are used to assess total uncertainty in Chapter 7.
3.2 Intended Applications and Expected Performance
A description of the intended applications and expected performance of a
proposed seismic-force-resisting system is required. This description should
include: (1) the anticipated function and occupancy; (2) physical and
behavioral characteristics of the system; (3) typical geometric configurations;
and (4) any similarities or differences between the proposed system and
current code-conforming systems. The description should also indicate how
the structural system and its key components are expected to perform in an
earthquake.
The following information should be used as a guide for describing the
intended applications of a proposed seismic-force-resisting system:
. intended occupancies and use of facilities to be constructed using the
proposed system,
FEMA P695 3: Required System Information 3-3
. horizontal and vertical configurations (e.g., framing layout, spans, story
heights, overall heights) of typical facilities to be constructed using the
proposed system,
. structural gravity framing systems to be used in combination with the
proposed system, including typical dead and live loads,
. geometric configurations of the proposed seismic-force-resisting system,
. expected inelastic behavior under seismic loading of varying intensity,
and
. methods of construction.
3.3 Design Requirements
Design requirements establish the fundamental information that will be used
to proportion and detail components, analyze the predicted response, and
confirm the behavior of a proposed system. They also set boundaries in the
application of the system. Design requirements are an essential input to the
development of structural system archetypes in Chapter 4. Information
needed to define and document system design requirements is specified in
the sections that follow.
3.3.1 Basis for Design Requirements
Design requirements should be based on criteria specified in applicable
sections of the latest edition of ASCE/SEI 7, Minimum Design Loads for
Buildings and Other Structures (ASCE, 2006a), and other applicable material
reference standards, such as ACI 318, Building Code Requirements for
Structural Concrete (ACI, 2005), AISC/ANSI 341, Seismic Provisions for
Structural Steel Buildings (AISC, 2005), ACI 530/ASCE 5/TMS 402,
Building Code Requirements for Masonry Structures (ACI, 2002b), and
ANSI/AF&PA, National Design Specification for Wood Construction
(ANSI/AF&PA, 2005). The following statements, taken mostly from the
NEHRP Recommended Provisions for Seismic Regulations for New
Buildings and Other Structures (FEMA, 2004a), should be used as a basis for
developing design requirements:
. The structure shall include complete lateral- and vertical-force-resisting
systems capable of providing adequate strength, stiffness, and energy
dissipation capacity to withstand the design ground motions within the
prescribed limits of deformation and strength demand.
. Design ground motions shall be assumed to occur along any direction of
the structure.
3-4 3: Required System Information FEMA P695
. Adequacy of the systems shall be demonstrated through construction of a
mathematical model, and evaluation of this model for the effects of
design ground motions. This evaluation shall be based on analysis in
which design seismic forces are distributed and applied throughout the
height of the structure in accordance with the ASCE/SEI 7-05.
. Deformations and internal forces in all members of the structure shall be
determined and evaluated against acceptance criteria contained or
referred to in ASCE/SEI 7-05, and as developed for the system under
consideration.
. A continuous load path, or paths, shall be provided with adequate
strength and stiffness to transfer all forces from the point of application
to the final point of resistance.
. The foundation shall be designed to accommodate forces developed or
movements imparted to the structure by design ground motions. In
determining foundation design criteria, special recognition shall be given
to the dynamic nature of the forces, the expected ground motions, and the
design basis for strength and energy dissipation capacity of the structure.
. Design of a structure shall consider potentially adverse effects on the
stability of the structure due to failure of a member, connection, or
component of the seismic-force-resisting system.
3.3.2 Application Limits and Strength Limit States
The boundaries of the intended application of the proposed seismic-forceresisting
system must be clearly stated, including, for example, any proposed
height limitations or restrictions to certain Seismic Design Categories.
Design requirements must address material properties, components,
connections, assemblies, and seismic-force-resisting system overall behavior.
With generally accepted modeling criteria and good engineering judgment,
design requirements should be of sufficient detail that analytical models of
component behavior can be developed. They must address the details of
stiffness models for members, connections, assemblies, and the overall
system, recognizing that seismic performance factors will be used in the
context of linear analyses and response to equivalent static forces. Where
size effects are important, they must be included.
Design requirements must provide information necessary to quantify all
pertinent strength limit states, including:
. tension, compression, bending, shear;
. yield, rupture, brittle fracture;
FEMA P695 3: Required System Information 3-5
. local, member, and global instability.
Proposed systems that rely on standard structural materials, or minor
modifications to existing, proven systems, can reference much of the
requirements to existing standards. However, such references must be
clearly justifiable and verifiable. New systems that behave outside the
bounds of existing system behavior must include consideration of behavioral
effects on other elements of building construction, including the gravity load
system and nonstructural components.
3.3.3 Overstrength Design Criteria
It is expected that most seismic-force-resisting systems will rely on inelastic
behavior somewhere within the system. Overstrength criteria should be
applied to the design of components that are judged to have small inelastic
deformation capacity followed by rapid deterioration in strength. This is
especially important if the component is also an essential part of the gravity
load system. Design requirements should be written in a manner that clearly
identifies all such components so that it is not left up to the judgment of the
designer to make this identification.
Design procedures utilizing linear static equivalent lateral force analyses
should follow the current standard method of requiring that such components
be designed for gravity loads plus .0
times the seismic loads, or for the
maximum forces that can be delivered to the component by other elements in
the system. It is understood that the overstrength factor used for this purpose
is based on judgment, and can vary by a large amount depending on system
configuration. To provide adequate protection, this factor should be a high
estimate of the expected ratio of maximum force to design force, particularly
for systems or materials that are non-ductile or have significant variability or
uncertainty in response.
3.3.4 Configuration Issues
Design requirements should comprehensively address all expected system
configurations. Emphasis should be placed on criteria that protect against the
occurrence of non-ductile failure modes and unintended concentration of
inelastic action in limited portions of the system.
When determining the design strength of components that are affected by
combined actions, such as axial-shear force interaction, consideration should
be given to system configurations that might have an effect on the magnitude
of combined loads. Beneficial effects of gravity loading must not be
permitted in configurations that result in little or no gravity load on seismic-
3-6 3: Required System Information FEMA P695
force-resisting components. Similarly, possible detrimental effects of
induced vertical loads should be considered in configurations that generate
high axial loads on seismic-force-resisting components.
Design requirements should address issues of multi-directional loading, and
simultaneous in-plane and out-of-plane loading, unless the combined load
effects are demonstrated to be unimportant.
3.3.5 Material Properties
Design requirements should document all material properties that will serve
as reference values for design of components, as well as criteria for
determining and measuring these properties. Documentation is not needed
for material properties that are prescribed in existing codes and material
reference standards. To the extent possible, the experimental determination
of material properties should be based on testing procedures specified in
ASTM standards. Material properties of interest include:
. tensile, compressive, and shear stress and strain properties,
. friction properties between parts that might possibly slide,
. bond properties at the interface of two materials, and
. other properties on which component behavior depends strongly.
In the determination of material properties, consideration should be given to
the simulation of common field conditions during testing, including
confinement conditions (e.g., bi-axial or tri-axial states of stress or strain),
environmental effects (e.g., temperature, moisture, solar radiation), and
cyclic loading. Effects of aging should be quantified, if deemed important to
seismic behavior.
Design requirements should include criteria for field testing of material
properties for systems in which the reference properties depend strongly on
case-specific mix proportions, placement, curing or other similar aspects of
construction.
3.3.6 Strength and Stiffness Requirements
Design requirements should contain comprehensive guidance for the
determination of design strength and effective elastic stiffness of structural
components, and assemblies of components.
. Stiffness requirements. Guidance on determination of the effective
elastic stiffness of structural components should be provided. The
effective elastic stiffness is defined as the stiffness that, if utilized in an
FEMA P695 3: Required System Information 3-7
analytical model, will provide a good estimate of the story drift demand
at the design level.
. Component strength requirements. The nominal strength of a
component should be expressed in terms of material properties, and
quantified for the range of loads, and combinations of loads, that might
be experienced as the system is subjected to collapse-level ground
motions.
Uncertainty inherent in a strength design equation, as well as the severity
of the consequence of failure, should be reflected in the resistance factor
(.-factor) associated with the strength design equation. Resistance
factors calibrated for use with common gravity load combinations are
recommended for use. Although these factors may not be anchored in
reliability analyses for seismic load combinations, design requirements
will be utilized in conjunction with linear analyses and equivalent static
forces, and the use of resistance factors will have an important effect on
the overall capacity of the system.
If the strength or deformation capacity of one component is affected
significantly by interaction with other components, then this interaction
should be accounted for in design equations.
If a component is subjected to a load effect, or combination of load
effects, that will cause rapid deterioration in strength in the inelastic
range, then this load effect, or this combination of load effects, must be
clearly identified as a non-ductile mode, and should trigger overstrength
design criteria.
Design requirements should be specific about component detailing
needed to ensure adequate strength during inelastic deformation.
. Connection strength requirements. Design requirements should be
specific about design of connections. In general, connections are
considered to be non-ductile. If a proposed system is based on a ductile
connection, then design requirements must clearly result in connections
that will have sufficient deformation capacity to avoid significant
deterioration before any of the connected components reach their
expected limits.
. Sensitivity to gravity loads. Where the strength of a member or
connection is sensitive to compression or tension from gravity loads,
design requirements must account for the effect of vertical ground
motions. Standard factors in existing standards for additive effects (1.2 +
0.2SDS) and counteracting effects (0.9 – 0.2SDS) may be used only if
justified through studies on structural system archetypes.
3-8 3: Required System Information FEMA P695
3.3.7 Approximate Fundamental Period
The design requirements should include formulas for calculation of the
approximate fundamental period, Ta, when the formulas of Section 12.8.2.1
of ASCE/SEI 7-05 do not apply to the system of interest (e.g., when the
approximate period parameters given in Table 12.8-2 of ASCE/SEI 7-05 for
"all other structural systems" are not appropriate). The formula(s) for Ta
should provide an estimate of the mean minus one sigma value of the actual
first mode period of buildings in which the system is utilized.
3.4 Quality Rating for Design Requirements
Quality of information is related to uncertainty, which factors into the
performance evaluation for a proposed seismic-force-resisting system. The
quality of the proposed design requirements is rated in accordance with the
requirements of this section, and approved by the peer review panel.
Design requirements are rated between (A) Superior and (D) Poor, as shown
in Table 3-1. The selection of a quality rating considers the completeness
and robustness of the design requirements, and confidence in the basis for the
design equations. Quantitative values of design requirements-related
collapse uncertainty are: (A) Superior, .DR = 0.10; (B) Good, .DR = 0.20; (C)
Fair, .DR = 0.35; and (D) Poor, .DR = 0.50. Use of these values is described
in Section 7.3.
Table 3-1 Quality Rating of Design Requirements
Completeness and Robustness
Confidence in Basis of Design Requirements
High Medium Low
High. Extensive safeguards
against unanticipated failure
modes. All important design
and quality assurance issues
are addressed.
(A) Superior
.DR = 0.10
(B) Good
.DR = 0.20
(C) Fair
.DR = 0.35
Medium. Reasonable
safeguards against
unanticipated failure modes.
Most of the important
design and quality assurance
issues are addressed.
(B) Good
.DR = 0.20
(C) Fair
.DR = 0.35
(D) Poor
.DR = 0.50
Low. Questionable
safeguards against
unanticipated failure modes.
Many important design and
quality assurance issues are
not addressed.
(C) Fair
.DR = 0.35
(D) Poor
.DR = 0.50
--
The highest rating of (A) Superior applies to systems that include a
comprehensive set of design requirements that provide safeguards against
unanticipated failure modes. Therefore, for a superior rating, there must be a
FEMA P695 3: Required System Information 3-9
high level of confidence that the design requirements produce the anticipated
structural behavior. Existing code requirements for special concrete moment
frames and special steel moment frames, for example, have been vetted with
detailed experimental results and documented performance in earthquakes.
Design and detailing provisions include capacity design requirements to
safeguard against unanticipated behaviors. A set of design requirements such
as these would be rated (A) Superior.
The lowest rating of (D) Poor applies to design requirements that have
minimal safeguards against unanticipated failure modes, do not ensure a
hierarchy of yielding and failure, and would generally be associated with
systems that exhibit behavior that is difficult to predict.
3.4.1 Completeness and Robustness Characteristics
Completeness and robustness characteristics are related to how well the
design requirements address issues that could potentially lead to
unanticipated failure modes, as well as proper implementation of designs
through fabrication, erection and final construction. Completeness and
robustness characteristics are rated from high to low, as follows:
. High. Design requirements are extensive, well-vetted and provide
extensive safeguards against unanticipated failure modes. They establish
a definite hierarchy of component yielding and failure. All important
issues regarding system behavior have been addressed, resulting in a
high reliability in the behavior of the system. Through mature
construction practices, and tightly specified quality assurance
requirements, there is a high likelihood that the design provisions will be
well executed through fabrication, erection and final construction.
. Medium. Design requirements are reasonably extensive and provide
reasonable safeguards against unanticipated failure modes, leaving some
limited potential for the occurrence of such modes. Design requirements
establish a suggested hierarchy of component yielding and failure.
While most important behavioral issues have been addressed, some have
not, which somewhat reduces the reliability of the system. Quality
assurance requirements are specified but do not fully address all the
important aspects of fabrication, erection and final construction.
. Low. Design requirements provide questionable safeguards against
unanticipated failure modes. Hierarchy of component yielding and
failure has been only marginally addressed (if at all), and there is a
likelihood of the occurrence of unanticipated failure modes. Design
requirements do not address all important behaviors, resulting in
3-10 3: Required System Information FEMA P695
marginally reliable behavior of the system. Quality assurance is lacking,
written guidance is not provided, and construction practices are not welldeveloped
for the type of system and materials.
Simplified, but conservative, design requirements by themselves are not a
reason for a low Completeness and Robustness rating as long as
conservatism is quantifiable in the context of unanticipated failure modes.
3.4.2 Confidence in Design Requirements
Confidence in the basis of the design requirements refers to the degree to
which the prescribed material properties, strength criteria, stiffness
parameters, and design equations are representative of actual behavior and
will achieve the intended result. Confidence is rated from high to low, as
follows:
. High. There is substantiating evidence (experimental data, history of
use, similarity with other systems) that results in a high level of
confidence that the properties, criteria, and equations provided in the
design requirements will result in component designs that perform as
intended.
. Medium. There is some substantiating evidence that results in a
moderate level of confidence that the properties, criteria, and equations
provided in the design requirements will result in component designs that
perform as intended.
. Low. There is little substantiating evidence (little experimental data, no
history of use, no similarity with other systems) that results in a low level
of confidence that the properties, criteria, and equations provided in the
design requirements will result in component designs that perform as
intended.
3.5 Data from Experimental Investigation
Analytical modeling alone is not adequate for predicting nonlinear seismic
response with confidence, particularly for structural systems that have not
been subjected to past earthquakes. A comprehensive experimental
investigation program is necessary to establish material properties, develop
design criteria, calibrate and validate component models, confirm behavior,
and calibrate analyses for a proposed seismic-force-resisting system.
Experimental results from other testing programs can be used to supplement
an experimental investigation program, but these results must come from
reliable sources, and their applicability to the system under consideration
must be demonstrated.
FEMA P695 3: Required System Information 3-11
There are practical limitations on how comprehensive an experimental
testing program can be. It must be understood, however, that limitations on
available experimental data will affect the uncertainty and reliability of the
collapse assessment of a proposed system, and will factor directly into the
performance evaluation process. The scope of an experimental investigation
program should be developed in consultation with the peer review panel.
3.5.1 Objectives of Testing Program
Testing is used to develop basic information so that the combination of
experimental and analytical data is sufficient to achieve the following two
objectives:
. Predict the seismic response of structures in the regime of interest to the
establishment of seismic performance factors, which occurs when the
structure, or any portion of the structure, is subjected to large seismic
demands and approaches a state of lateral dynamic instability (collapse).
This implies the need to model strength and stiffness properties of
important components, and reliably capture structural response, from
elastic behavior through the state of lateral dynamic instability, over the
entire range of possible structural configurations permitted by the design
requirements.
. Develop and validate standardized engineering design criteria that can be
used to design structures that perform as expected, given the specified
seismic performance factors.
Achievement of these objectives requires a coordinated material, component,
connection, assembly, and system testing program that will provide the
following information:
. Material test data. Data that serve as reliable reference values for the
prediction of strength, stiffness, and deformation properties of structural
components and connections under earthquake loading.
. Component and connection test data. Information needed to develop
and calibrate design criteria and analytical models of cyclic loaddeformation
characteristics for components and connections that form an
essential part of the seismic-force-resisting system.
. Assembly and system test data. Information needed to quantify
interactions between structural components and connections that cannot
be predicted by analysis with confidence.
3-12 3: Required System Information FEMA P695
3.5.2 General Testing Issues
In developing a comprehensive testing program, the following issues should
be considered:
. Cumulative damage effects. Structural materials and components
experience history-dependent cumulative damage during repeated cyclic
loading. The loading history used in testing should be representative of
the cyclic response that a material, component, connection, or assembly
would experience as part of a typical structural system subjected to a
severe earthquake.
. Size effects. Tests should be performed on full-size specimens unless it
can be shown by theory and experiment that testing of reduced-scale
specimens will not significantly affect behavior.
. Strain rate effects. If the load-deformation characteristics of the
specimen are sensitive to strain rate effects, then testing should be done
at strain rates commensurate with those that would be experienced in a
severe earthquake.
. Boundary conditions. The boundary conditions of component and
assembly tests should be: (1) representative of constraints that a
component or assembly would experience in a typical structural system;
and (2) sufficiently general so that the results can be applied to boundary
conditions that might be experienced in other system configurations.
Boundary conditions should not impose beneficial effects on seismic
behavior that would not exist in common system configurations.
. Load application. Loads should be applied to test specimens in a
manner that replicates the transfer of forces to the component or
assembly commonly occurring in in-situ conditions.
. Configuration and number of component/assembly test specimens.
The configuration and number of component and assembly test
specimens should be such that all common failure modes that could
occur in typical system configurations are represented and evaluated.
Emphasis should be on the detection and evaluation of failure modes that
lead to a rapid deterioration in strength (e.g., brittle failure modes).
. Interaction between structural components. Test configurations
should consider important interactions between structural components,
unless these interactions can be predicted with confidence by analysis.
. Direction(s) of loading. Structural components that resist seismic forces
in more than one direction (e.g., concrete core walls) should be tested
FEMA P695 3: Required System Information 3-13
such that the combined load effects are adequately considered, unless
these effects can be predicted with confidence by analysis.
. In-plane and out-of-plane load effects. Planar structural components
(e.g., walls, diaphragms) should be subjected to simultaneous in-plane
and out-of-plane loading, unless these effects can be superimposed with
confidence by analysis.
. Gravity load effects. Effects of gravity loads should be considered in
the experimental program, unless these effects can be superimposed on
lateral load effects with confidence by analysis.
. Statistical variability. A sufficiently large number of tests should be
performed so that statistical variations can be evaluated from the data
directly, or can be deduced in combination with data from other sources.
. Environmental conditions. If environmental conditions during
construction or service (e.g., temperature, humidity) will significantly
affect behavior, then the range of conditions that could exist in practice
should be simulated during testing.
. Test specimen construction. Specimens should be constructed in a
setting that simulates commonly encountered field conditions. For
example, if field conditions necessitate overhead welding then this type
of welding should be applied in test specimen construction.
. Quality of test specimen construction. Construction of component,
connection, assembly, and system specimens should match the level of
quality that will be commonly implemented in the field. Special
construction techniques or quality control measures should not be
employed, unless they are part of the design requirements.
. Past experience. Laboratory testing cannot fully replace experience
gained from observation of system behavior in actual use. A benefit
should be given to structural systems whose performance has been
documented in past earthquakes or other use.
. Documentation of tests and test results. Documentation of
experiments should be comprehensive, and should include: (1) geometric
data, test setup, and boundary conditions; (2) important details of the test
specimen, including construction process and fabrication details; (3) type
and location of instruments used for measurement of important response
parameters; (4) material test data needed for performance evaluation; (5)
a written record of all important events prior and during the test; (6) a
comprehensive log of all important visual observations; and (7) a
3-14 3: Required System Information FEMA P695
comprehensive set of digital experimental data needed for performance
evaluation.
The above list should be used as a guide. There may be other issues that are
equally important to a given system, but cannot be placed in a general
context. Assistance in the identification of important testing issues can be
obtained from references available in the literature (e.g., ACI, 2001; AISC,
2005; ASTM, 2003; ATC, 1992; Clark et al., 1997; FEMA, 2007; and ICC,
2009). Specimen fabrication, testing procedures, loading protocol, and test
documentation should follow guidelines established in these references as
appropriate for the system being evaluated, and as approved by the peer
review panel.
Testing laboratories used to conduct an experimental investigation program
should comply with national or international accreditation criteria, such as
the ISO/IEC 17025 General Requirements for the Competence of Testing and
Calibration Laboratories (ISO/IEC, 2005). Testing Laboratories that are not
accredited may be used for the experimental investigation program, subject
to approval by the peer review panel.
3.5.3 Material Testing Program
A material testing program is required to provide reliable stress-strain
relationships for the prediction of strength, stiffness, and deformation
properties of structural components and connections under the type of
loading experienced during an earthquake. In addition to general testing
issues, materials testing should consider the following:
. low-cycle fatigue and fracture properties,
. bi-axial and tri-axial stress conditions,
. utilization of applicable ASTM Standards,
. evaluation of variability in material properties,
. effects of aging, and
. effects of environmental conditions.
Material testing programs should be performed in accordance with all
applicable ASTM Standards and other testing criteria specified in nationally
accepted industry standards and specifications. Material test data available
from past tests that conform to all applicable standards may be used.
FEMA P695 3: Required System Information 3-15
3.5.4 Component, Connection, and Assembly Testing
Program
A component, connection, and assembly testing program is required to
provide information for the development, calibration, and validation of
analytical models of cyclic load-deformation characteristics of components
and connections that form an essential part of the seismic-force-resisting
system, as well as development of the design criteria for such components
and connections.
Components, connections, and assemblies that have low inelastic
deformation capacities (commonly referred to as non-ductile elements) must
be identified, and should undergo sufficient testing to validate both the
design strength of such elements as well as the strength properties used in the
analytical models employed in the collapse assessment.
Testing of Structural Components
Component testing serves to identify and quantify component parameters that
significantly affect seismic response. Cyclic behavior is characterized by a
basic hysteresis loop, which deteriorates with the number and amplitude of
cycles. It is critical that a test is continued until severe strength deterioration
is evident, and all important characteristics that enter design equations and
analytical models have been verified experimentally. Two hysteretic
responses of a structural component (in this case a steel beam and a plywood
panel) are shown in Figure 3-2. In the left figure it appears that the loops
stabilize at very large amplitudes, and that more (and larger) deformation
cycles can be sustained. However, the possibility of fracture at large
deformations is high, and once this fracture occurs, the resistance will
deteriorate, rapidly approaching zero. For this reason, no credit should be
given to residual strength beyond the deformation at which the test is
terminated.
Displacement
Force
Displacement
Force
00
Force
Displacement
00
Force
Displacement
Figure 3-2 Characteristics of force-deformation response: (a) steel beam
(Uang et al. 2000); (b) plywood shear wall panel (Gatto and
Uang, 2002).
3-16 3: Required System Information FEMA P695
A sufficiently large number of tests should be performed so that important
statistical variations can be evaluated from the data directly, or can be
deduced in combination with data from other sources. A minimum of two
tests is required for each set of primary variables in a test configuration. If
rapid deterioration occurs, such as that caused by brittle fracture, a minimum
of three tests should be performed. If rapid deterioration is not of concern, it
is recommended that one of the tests be a monotonic loading test, which
facilitates analytical modeling of the type discussed in Sections 5.4 and 5.5.
Component tests are conducted for the purpose of evaluating all force and
deformation characteristics that have a significant effect on the seismic
response up to the state of incipient collapse. Gravity loads should be
represented, unless it can be shown that their effect is not detrimental to the
seismic behavior or can be predicted with confidence from analytical models.
The load application and loading history should be representative of what
components will experience as part of a typical structural system subjected to
a severe earthquake. Instrumentation should permit the measurement of all
relevant stiffness and strength properties.
In contrast to “qualification testing”, which is intended to gain approval of
the use of certain components for specific applications, the main objectives
of these component tests are to develop design criteria and to calibrate and
validate nonlinear models that are used in collapse assessment. Types and
configurations of component tests, together with the loading protocol, should
be planned in conjunction with development of the nonlinear analysis models
in Chapter 5.
Testing of Connections
A connection is the medium that transfers forces and deformations between
adjacent components. Connections should be tested in configurations that
simulate gravity load effects as well as seismic load effects, unless gravity
loading results in more favorable connection behavior. Connection tests
should provide all information necessary to develop connection design
criteria in conformance with the latest edition of ASCE/SEI 7-05, and to
permit simulation of connections in analytical models.
Testing of Assemblies
An assembly is an arrangement of structural components whose seismic
behavior can be described in terms of a single response quantity, such as
story drift. An assembly testing program is required if important interactions
between adjacent components (or between components and connections)
cannot be deduced with confidence from a combination of material,
FEMA P695 3: Required System Information 3-17
component, and connection tests in combination with analytical modeling.
Unless strain-rate effects are important, assembly tests can be performed by
imposing load(s) to control point(s) in a quasi-static manner, following a
predetermined loading history.
3.5.5 Loading History
Structural elements have limited strength and deformation capacities.
Collapse safety depends on the ability to assess these capacities with some
confidence. Strength and deformation capacities depend on cumulative
damage, which implies that every component has a “memory” of past
damaging events, and that all past excursions (or cycles) that have
contributed to its current state of health will affect future behavior. Thus,
performance depends on the history of previously applied damaging cycles,
and assessing the consequences of loading history requires replication of the
load and deformation cycles that a component will undergo in an earthquake
(or several earthquakes, if appropriate). The objective of a loading history is
to achieve this in an approximate, but consistent manner.
There is no unique or best loading history, because no two earthquakes are
alike, and a specimen may be part of many different structural
configurations. The overriding issue is to account for cumulative damage
effects through cyclic loading. The number and amplitude of cycles applied
to the specimen may be derived from analytical studies in which models of
representative structural systems are subjected to representative earthquake
ground motions, and the response is evaluated statistically. In analytical
modeling, it should be assumed that specimen resistance deteriorates to zero
following the maximum amplitude executed in the test. No credit should be
given to deformation capability beyond the largest deformation that a
specimen experiences in a test.
Many loading protocols have been proposed in the literature, and several
have been used in multi-institutional testing programs (e.g., ATC, 1992;
Clark et al., 1997; Krawinkler et al., 2000), or are contained in standards or
are proposed for standards (e.g., AISC, 2005; ASTM, 2003; FEMA, 2007;
ICC, 2009). These protocols recommend somewhat different loading
histories, but they differ more in detail than in concept. Comprehensive
discussions of loading histories and their origin and objectives are presented
in Filiatrault et al. (2008), Krawinkler et al. (2000), and Krawinkler (1996).
The loading protocols referenced above have been developed with a design
level or maximum considered ground motion in mind, and not for the
purpose of collapse evaluation. As a result the loading histories are
3-18 3: Required System Information FEMA P695
symmetric with step-wise increasing deformation cycles (with the exception
of the SAC near-fault loading protocol presented in Clark et al., 1997).
While such histories are not representative of cyclic response approaching
collapse, they do serve the purpose of quantifying deterioration properties for
analytical modeling, particularly if a cyclic test is complemented with a
monotonic loading test that provides information on the force-displacement
capacity boundary (FEMA, 2009).
Loading history should be deformation-controlled, with the following two
exceptions:
. Force-control may be applied for small excursions in which a component
will remain essentially elastic. Force-control is encouraged for stiff
specimens tested in a relatively flexible test set-up, in order to facilitate
measurements and test control in the early stages of testing.
. Force-control may be necessary to test components that are an essential
part of the load path, but are fully force-controlled in the in-situ
condition, and have no reliable inelastic deformation capacity. One such
example would be an anchor controlled by the maximum force exerted
by a connected component. In such a case, the loading history should be
determined based on the strength and deformation capacities of the
connected component. Criteria for a force-controlled loading history are
presented in FEMA 461 Interim Testing Protocols for Determining the
Seismic Performance Characteristics of Structural and Nonstructural
Components (FEMA, 2007).
3.5.6 System Testing Program
Testing of an essentially complete structural system should be performed if
important response characteristics or important interactions between
components and connections cannot be evaluated with good confidence by
analytical models that have been calibrated through material, component,
connection, or assembly tests.
System tests should be used as a validation tool for a proposed analytical
model rather than as an exploratory test from which analytical models will be
developed. System tests should not be used to replace any testing at the
material, component, connection, or assembly level.
Dynamic System Tests
A dynamic system test should be performed if the response near collapse
depends strongly on dynamic characteristics that cannot be predicted with
good confidence by analytical models or from component or assembly tests.
FEMA P695 3: Required System Information 3-19
Such a test should be performed on an earthquake simulator (shake table)
utilizing realistic MCE-level (or collapse level) ground motions, unless it can
be demonstrated that equivalent response evaluation can be achieved through
alternative means.
Quasi-Static System Tests
Quasi-static cyclic testing consists of loads that are applied to one or more
control points by means of hydraulic actuators whose displacement or load
values are varied in a cyclic manner in accordance with a predetermined
loading history. The loading history used in testing should be representative
of the cyclic response that a material, component, connection, or assembly
would experience as part of a typical structural system subjected to a severe
earthquake.
A quasi-static system test should be performed if important behaviors or
interactions between components and connections cannot be evaluated with
good confidence by means of calibrated analytical models. Examples
include interactions between horizontal and vertical components (floor
diaphragms and vertical seismic-force-resisting units) and between vertical
components that resist seismic forces in orthogonal directions.
3.6 Quality Rating of Test Data
Quality of test data is related to uncertainty, which factors into the
performance evaluation for a proposed seismic-force-resisting system. The
quality of test data obtained from an experimental investigation program is
rated in accordance with the requirements of this section, and approved by
the peer review panel.
Test data are rated between (A) Superior and (D) Poor, as shown in Table
3-2. This rating depends not only on the quality of the testing program, but
on how well the tests address key parameters and behavioral issues. The
selection of a quality rating for test data considers the completeness and
robustness of the overall testing program, and confidence in the test results.
Quantitative values of test data-related collapse uncertainty are: (A) Superior,
.TD = 0.10; (B) Good, .TD = 0.20; (C) Fair, .TD = 0.35; and (D) Poor, .TD =
0.50. Use of these values is described in Section 7.3.
3-20 3: Required System Information FEMA P695
Table 3-2 Quality Rating of Test Data from an Experimental Investigation
Program
Completeness and Robustness
Confidence in Test Results
High Medium Low
High. Material, component,
connection, assembly, and system
behavior well understood and
accounted for. All, or nearly all,
important testing issues addressed.
(A) Superior
.TD = 0.10
(B) Good
.TD = 0.20
(C) Fair
.TD = 0.35
Medium. Material,
component, connection,
assembly, and system behavior
generally understood and
accounted for. Most important
testing issues addressed.
(B) Good
.TD = 0.20
(C) Fair
.TD = 0.35
(D) Poor
.TD = 0.50
Low. Material, component,
connection, assembly, and
system behavior fairly
understood and accounted for.
Several important testing issues
not addressed.
(C) Fair
.TD = 0.35
(D) Poor
.TD = 0.50
--
3.6.1 Completeness and Robustness Characteristics
Completeness and robustness characteristics are related to: (1) the degree to
which relevant testing issues have been considered in the development of the
testing program; and (2) the extent to which the testing program and other
documented experimental evidence quantify the necessary material,
component, connection, assembly, and system properties and important
behavior and failure modes. Completeness and robustness characteristics are
rated from high to low, as follows:
. High. All, or nearly all, important general testing issues of Section 3.5.2
are addressed comprehensively in the testing program and other
supporting evidence. Experimental evidence is sufficient so that all, or
nearly all, important behavior aspects at all levels (from material to
system) are well understood, and the results can be used to quantify all
important parameters that affect design requirements and analytical
modeling.
. Medium. Most of the important general testing issues of Section 3.5.2
are addressed adequately in the testing program and other supporting
evidence. Experimental evidence is sufficient so that all, or nearly all,
important behavior aspects at all levels (from material to system) are
generally understood, and the results can be used to quantify or deduce
most of the important parameters that significantly affect design
requirements and analytical modeling.
. Low. Several important general testing issues of Section 3.5.2 are not
addressed adequately in the testing program and other supporting
FEMA P695 3: Required System Information 3-21
evidence. Experimental evidence is sufficient so that the most important
behavior aspects at all levels (from material to system) are fairly well
understood, but the results are not adequate to quantify or deduce, with
high confidence, many of the important parameters that significantly
affect design requirements and analytical modeling.
3.6.2 Confidence in Test Results
Confidence in test results is related to the reliability and repeatability of the
results obtained from the testing program, and corroboration with available
results from other relevant testing programs. It includes consideration as to
whether or not experimental results consistently record performance to
failure for all modes of behavior (limited ductility to large ductility), and if
sufficient information is provided to assess uncertainties in the design
requirements (e.g., . factors) and analytical models. Confidence in test
results is rated from high to low, as follows:
. High. Reliable experimental information is produced on all important
parameters that affect design requirements and analytical modeling.
Comparable tests from other testing programs have produced results that
are fully compatible with those from the system-specific testing program.
A sufficient number of tests are performed so that statistical variations in
important parameters can be assessed. Test results are fully supported by
basic principles of mechanics.
. Medium. Moderately reliable experimental information is produced on
all important parameters that affect design requirements and analytical
modeling. Comparable tests from other testing programs do not
contradict, but do not fully corroborate, results from the system-specific
testing program. A measure of uncertainty in important parameters can
be estimated from the test results. Test results are supported by basic
principles of mechanics.
. Low. Experimental information produced on many of the important
parameters that affect design requirements and analytical modeling is of
limited reliability. Comparable tests from other testing programs do not
support the results from the system-specific testing program. Insufficient
data exists to assess uncertainty in many important parameters. Basic
principles of mechanics do not support some of the results of the testing
program.
FEMA P695 4: Archetype Development 4-1
Chapter 4
Archetype Development
This chapter describes the development of structural system archetypes,
which provide a systematic means for characterizing key features and
behaviors related to collapse performance of a proposed seismic-forceresisting
system. It defines how archetype descriptions and performance
characteristics are used to develop a set of building configurations (index
archetype configurations) that together describe the overall range of
permissible configurations (archetype design space) of a system, which is
then separated into groups sharing common features or behavioral
characteristics (performance groups) for assessing collapse performance.
Specific structural designs (index archetype designs) are then developed for
each configuration based on the specified design criteria, and these designs
then form the basis of nonlinear analysis models (index archetype models)
that are analyzed to assess collapse performance. Guidelines related to index
archetype designs and index archetype models are presented in
Chapter 5.
4.1 Development of Structural System Archetypes
Behavior of a proposed seismic-force-resisting system is investigated
through the use of archetypes. An archetype is a prototypical representation
of a seismic-force-resisting system. Archetypes are intended to reflect the
range of design parameters and system attributes that are judged to be
reasonable representations of the feasible design space and have a
measurable impact on system response. They are used to bridge the gap
between collapse performance of a single specific building and the
generalized predictions of behavior needed to quantify performance for an
entire class of buildings.
An index archetype configuration is a prototypical representation of a
seismic-force-resisting system configuration that embodies key features and
behaviors related to collapse performance when the system is subjected to
earthquake ground motions. Given that building codes permit significant
latitude with respect to system configurations within a building class, index
archetype configurations are not intended to represent every conceivable
configuration of the system of interest. Rather, the intent is to investigate a
reasonably broad range of parameters that are permitted by the specified
design requirements and represent conditions that are feasible in design and
construction practice.
4-2 4: Archetype Development FEMA P695
Collectively, the set of index archetype configurations describe the archetype
design space, which defines the overall range of permissible configurations,
structural design parameters, and other properties that define the application
limits for a seismic-force-resisting system. For performance evaluation, the
archetype design space is divided into performance groups, which are groups
of index archetype configurations that share a common set of features or
behavioral characteristics.
Development of structural system archetypes follows the process outlined in
Figure 4-1. Using the design requirements and test data developed under
Chapter 3 as inputs, the development of structural system archetypes
considers both structural configuration issues and seismic behavioral effects.
Figure 4-1 Process for development of structural system archetypes.
4.2 Index Archetype Configurations
Index archetype configurations must be sufficiently broad in scope to capture
the range of situations that are feasible under the design requirements, but
sufficiently limited to be practical to evaluate. The intent is to both: (1)
FEMA P695 4: Archetype Development 4-3
assess situations that will be generally representative of practice that meets
minimum specified requirements for seismic design and construction of a
proposed seismic-force-resisting system; and (2) assess designs that are at the
limits of the range of design configurations that are allowable based on the
design requirements.
While index archetype configurations are not intended to represent every
conceivable combination of design parameters, the archetype configurations
must encompass the full design space permitted by the design requirements.
An exception to this occurs when the collapse safety trends for the assessed
archetype designs show that certain configurations will not control the
system performance assessment.
Index archetype configurations should not incorporate “standard” practices
that may routinely exceed minimum code requirements. For example, use of
one member size at multiple locations in a building is a design and
construction practice that can result in member overstrengths, which exceed
minimum design requirements, and should not be built into index archetype
configurations.
It is expected that the set of index archetype configurations will generally
include about twenty to thirty specific structural configurations, though the
specific number will depend on the characteristics of the seismic-forceresisting
system and the limits of the archetype design space. Where the
seismic design requirements are relatively loose and cover a broad range of
possible design situations, the number of required index archetype
configurations may be significantly larger than for systems with more limited
applications. The final selection of index archetype configurations, and their
corresponding design parameters, should be reviewed and approved by the
peer review panel.
The development of index archetype configurations involves the following
steps:
. Identify key design variables and related physical properties based on
structural configuration issues summarized in Section 4.2.1. Investigate
physical properties that affect collapse performance to identify critical
design variables that should be reflected in index archetype
configurations.
. Establish bounds for key design variables that define the archetype
design space. The design space is limited primarily by the seismic
design requirements and practical constraints on design and construction.
4-4 4: Archetype Development FEMA P695
. Identify behavioral issues and related design considerations based on
behavioral effects summarized in Section 4.2.2. Investigate possible
deterioration modes that could result in local and global collapse
scenarios, and assess the likelihood of those scenarios.
. Develop a set of index archetype configurations based on key design
variables and behavioral effects that are likely to result in local or global
collapse scenarios.
4.2.1 Structural Configuration Issues
Structural configuration issues include occupancy and program influences,
framing type and geometric variations, and gravity and lateral load
intensities. Typical configuration design variables that can affect the
behavior of a seismic-force-resisting system are summarized in Table 4-1.
These structural issues should be used as a guide in establishing index
archetype configurations, as follows:
. Occupancy and Use. Building occupancy and use can influence the
structural layout, framing system, configuration, and loading intensity.
Framing spans, story heights, and live loads for seismic-force-resisting
systems intended for residential occupancies are usually quite different
from office occupancies. Similarly, steel moment frames and associated
gravity framing used for industrial occupancies can be different from
those used for office or institutional buildings. Occupancies with large
live load demands may have larger inherent overstrength in comparison
with systems for other occupancies. Major changes in structural
configuration resulting from different occupancies and use should be
reflected in the index archetype configurations.
. Elevation and Plan Configuration. The range of elevation and plan
configurations permitted by the system design requirements should be
reflected in the index archetype configurations. This could include, for
example: (1) the range of framing span lengths of the seismic-forceresisting
system; (2) alternative configurations of steel bracing (e.g.,
chevron versus x-bracing); (3) variations in shear wall aspect ratios that
result in flexure- versus shear-dominated behavior; (4) floor diaphragm
characteristics that are addressed in the seismic system designation; and
(5) the extent of gravity loading tributary to the system. The extent to
which such factors are significant will vary depending on the type of
seismic-force-resisting system.
. Building Height. The range of story heights and number of stories
permitted by the system design requirements should be reflected in the
index archetype configurations to the extent that these parameters affect
FEMA P695 4: Archetype Development 4-5
the structural period and the localization of inelastic deformations. Since
the inelastic response of structures with short periods (in the constant
acceleration region of the hazard spectrum) tends to be different from
structures with longer periods (in the constant velocity region of the
hazard spectrum), the response of short-period and long-period structures
are characterized separately. Due to significant differences in the
response characteristics of very tall buildings, and limited low frequency
content in the ground motions specified for collapse assessment, use of
this Methodology should be limited to buildings with a fundamental
period of T = 0.4 seconds (i.e., building heights of about 20 to 30 stories
for moment frame systems, and 30 to 40 stories for braced frame and
shear wall systems).
Table 4-1 Configuration Design Variables and Related Physical Properties
Design Variable Related Physical Properties
Occupancy and Use
. Typical framing layout
. Distribution of seismic-force-resisting system
components
. Gravity load intensity
. Component overstrength
Elevation and Plan
Configuration
. Distribution of seismic-force-resisting
components
. Typical framing layout
. Permitted vertical (strength and stiffness)
irregularities
. Beam spans, number of framing bays, system
regularity
. Wall length, aspect ratio, plan geometry, wall
coupling
. Braced bay size, number of braced bays,
bracing configuration
. Diaphragm proportions, strength, and stiffness
(or flexibility)
. Ratio of seismic mass to seismic-force-resisting
components
. Ratio of tributary gravity load to seismic load
Building Height
. Story heights
. Number of stories
Structural Component
Type
. Moment frame connection types
. Bracing component types
. Shear wall sheathing and fastener types
. Isolator properties and types
Seismic Design
Category
. Design ground motion intensity
. Special design/detailing requirements
. Application limits
Gravity Load
. Gravity load intensity
. Typical framing layout
. Ratio of tributary gravity load to seismic load
. Component overstrength
4-6 4: Archetype Development FEMA P695
. Structural Component Type. The extent that structural component
types can vary within a given seismic-force-resisting system should be
reflected in index archetype configurations. Examples include different
types of moment connection details (e.g., welded, bolted, or reducedbeam
section), steel bracing members (e.g., HSS, pipe, or W-shape), and
light-frame wood shear wall sheathing, framing, and fasteners.
. Seismic Design Category. Systems should be evaluated for the highest
(most severe) Seismic Design Category for which they are proposed, and
then verified in lower Seismic Design Categories. Index archetype
configurations within a Seismic Design Category should consider
spectral intensities corresponding to the maximum and minimum values
for that category, associated design and detailing requirements, and any
restrictions on use that are keyed to Seismic Design Category.
. Gravity Load. Large gravity load demands can result in overstrength
with respect to seismic demands. For some components (e.g., columns
in moment frames) axial load ratio can significantly impact inelastic
deformation capacity. The nature, magnitude and variation of gravity
loads, including structure self weight, occupancy-related superimposed
dead loads, and occupancy-related live loads should be considered, and
design parameters that affect tributary gravity load, such as bay sizes and
building height, should be reflected in the index archetype
configurations. It is anticipated that, for most systems, two levels of
gravity load (high and low) would be sufficient.
An example of how configuration issues are considered in the development
of index archetype configurations for a special reinforced concrete moment
frame system conforming to design requirements contained in ASCE/SEI
7-05 is provided in Appendix C.
4.2.2 Seismic Behavioral Effects
Consideration of seismic behavioral effects includes identifying dominant
deterioration and collapse mechanisms that are possible, and assessing the
likelihood that they will occur. How a component or system behaves under
seismic loading is often influenced by configuration decisions, so behavioral
effects and configuration issues should be considered concurrently in the
development of index archetype configurations.
Seismic collapse resistance depends on the strength, stiffness, and
deformation capacity of individual structural components and the overall
seismic-force-resisting system. Each of these properties can be addressed
directly through system design requirements, but each are also influenced by
FEMA P695 4: Archetype Development 4-7
aspects of the configuration that could change the way a system behaves
across the range of an archetype design space. For example, requirements
for ductile confinement of reinforced concrete columns will directly affect
inelastic deformation capacity, but the magnitude of column axial load,
which is influenced by elevation and plan configurations, also has a large
impact.
Consideration of behavioral effects is used to help identify major changes in
system behavior as the configuration varies. Once potential deterioration and
collapse mechanisms are identified, they are addressed through one of the
following methods: (1) by ruling out failure modes that are unlikely to occur
based on system design and detailing requirements; (2) explicit simulation of
failure modes through nonlinear analyses; or (3) evaluation of “nonsimulated”
failure modes using alternative limit state checks on demand
quantities from nonlinear analyses.
Typical behavioral issues and related design considerations that can have an
effect on the behavior of a seismic force-resisting system are summarized in
Table 4-2.
Table 4-2 Seismic Behavioral Effects and Related Design Considerations
Behavioral Issue Related Design Considerations
Strength
. Minimum design member forces
. Calculated member forces
. Capacity design requirements
. Component overstrength
Stiffness
. Design member forces
. Drift limits
. Plan and elevation configuration
. Calculated inter-story drifts
. Diaphragm stiffness (or flexibility)
. Foundation stiffness (or flexibility)
Inelastic deformation
capacity
. Component detailing requirements
. Member geometric proportions
. Capacity design requirements
. Calculated member forces
. Redundancy of the seismic forceresisting
system
Seismic Design Category
. Design ground motion intensity
. Special design/detailing requirements
Inelastic system mobilization
. Building height and period
. Diaphragm strength and stiffness
. Permitted strength and stiffness
irregularities
. Capacity design requirements
4-8 4: Archetype Development FEMA P695
These behavioral issues should be used as a guide in establishing index
archetype configurations, as follows:
. Strength. Differences between design strength and calculated seismic
demands should be reflected in the index archetype configurations.
Design strength is a function of the design earthquake intensity,
component detailing requirements, capacity design requirements, and
overstrength resulting from gravity loads and other minimum load
requirements. Calculated demands are a function of the structural
configuration and gravity load (dead and live load) intensity. In cases
where gravity load, minimum seismic or wind loads, other minimum
loads, or stiffness considerations control, there can be significant
overstrength relative to seismic design forces. Capacity design
provisions control yielding by requiring strengths of certain components
to be greater than would otherwise be required by minimum seismic
design forces to protect them from inelastic demands.
. Stiffness. The elastic lateral stiffness of the seismic-force-resisting
system affects the dynamic behavior, sensitivity to sidesway stability
(P-delta) effects, and induced deformation demands on critical
components. Stiffness is a function of the design earthquake intensity
and imposed drift limits, system configuration, and relative stiffness (or
flexibility) of certain key components, such as diaphragms or
foundations. Index archetype configurations should take into account
system types that are more sensitive to drift limits than others, and
should identify configurations that probe limits on minimum stiffness
within the system design requirements. Where behavior of certain
elements (e.g., diaphragm or foundation flexibility), is likely to influence
the performance of the system, these effects should be considered in the
development of index archetype configurations. For example, in lowrise
industrial structures with large floor plans and stiff seismic-forceresisting
elements, flexibility of the floor and roof diaphragms could
significantly alter the period of vibration of the system, and should,
therefore, be considered.
. Inelastic Deformation Capacity. Inelastic deformation capacity of
components is a function of design requirements, including detailing
rules and capacity design provisions, member geometric proportions, and
calculated member forces that can vary with structural configuration.
Where the inelastic deformation capacity of components is influenced by
the force distribution (such as differences in the level of axial forces in
walls or columns), factors that influence force distribution (such as plan
configuration) should be considered in the index archetype
FEMA P695 4: Archetype Development 4-9
configurations. The impact of structural redundancy on the distribution
of inelastic deformations should also be reflected in the index archetype
configurations.
. Seismic Design Category. Applicable Seismic Design Categories
establish the design ground motion intensities, which influence seismicforce-
resisting system strength and stiffness. Seismic Design Category
designations can also trigger special design and detailing requirements
that will influence component inelastic deformation capacity. Index
archetype configurations should reflect behavioral effects that are
influenced by the Seismic Design Categories for which a system is being
proposed.
. Inelastic System Mobilization. Inelastic system mobilization is the
extent to which inelastic action is distributed throughout the seismicforce-
resisting system. Inelastic system mobilization is influenced by
limits on stiffness and strength irregularities and other system design
requirements such as strong-column-weak-beam criteria. Design and
configuration decisions can affect whether yielding is distributed
vertically across many stories or tends to concentrate in just a few
stories. To the extent that the diaphragm design is specified as part of
the lateral system, diaphragm strength and stiffness will influence the
dynamic response and distribution of forces among the seismic-forceresisting
elements. Index archetype configurations should identify and
test configurations that are permitted by system design requirements, and
will result in a lower bound of inelastic system mobilization.
An example of how behavioral effects are considered in the development of
index archetype configurations for a special reinforced concrete moment
frame system conforming to design requirements contained in ASCE/SEI
7-05 is provided in Appendix D.
4.2.3 Load Path and Components Not Designated as Part of
the Seismic-Force-Resisting System
The complete load path of the seismic-force-resisting system should be
considered when defining index archetype configurations. Only those
components that are either specified in the design requirements of the
seismic-force-resisting system, or otherwise have a significant effect on the
system, should be incorporated in the archetype assessment. Portions of the
structure that comprise the gravity load system, but are not specifically
designed to resist seismic forces, should not be included as part of the index
archetype configuration. However, the seismic mass and destabilizing
4-10 4: Archetype Development FEMA P695
P-delta effects of the gravity system should be included in the index
archetype models.
Conversely, potential failure modes of the gravity system, floor diaphragms,
and collector elements need not be reflected in the index archetype
configurations, unless those elements are specifically defined in the design
requirements for the seismic-force-resisting system. While it is recognized
that failure of gravity elements may trigger collapse, their design is usually
considered through separate code requirements that are common to many
seismic-force-resisting systems and specific to none. If the deformation
demands at collapse of a proposed seismic-force-resisting system, however,
are excessive relative to those normally experienced by other systems, typical
controls for deformation compatibility of gravity framing may not be
adequate. In such cases, special deformation criteria for gravity system
components should be included in the design requirements for the proposed
seismic-force-resisting system.
4.2.4 Overstrength Due to Non-Seismic Loading
Index archetype designs should reflect the inherent overstrength that results
from earthquake and gravity load design requirements in addition to any
other minimum force requirements that are specified by building code
provisions that will always be applicable to a structure. For example, the
minimum wind load requirements specified in ASCE/SEI 7-05 (10 psf on the
projected surface area exposed to wind) are required in all structures, and
should be considered when evaluating potential overstrength. Similarly,
minimum seismic design force requirements, which are specified generally
for all seismic force resisting systems, should be incorporated in index
archetype designs.
Overstrength from design loading that could exist, but is not assured in all
cases, should not be considered in the development of index archetype
configurations. Such an example could include hurricane wind forces, or
other similar location- or use-specific non-seismic loads, that would not
apply to all buildings everywhere. Only non-seismic requirements that apply
universally to all structures should be considered when evaluating potential
overstrength relative to earthquake loading.
4.3 Performance Groups
Index archetype configurations are assembled into performance groups (or
bins) that reflect major differences in configuration, design gravity and
seismic load intensity, structural period, and other factors that may
significantly affect seismic behavior within the archetype design space.
FEMA P695 4: Archetype Development 4-11
Performance groups should contain multiple index archetype configurations
that reflect the expected range of permissible variation in size and other key
parameters defined by the archetype design space. For example, each
performance group should contain index archetype configurations that cover
the range of building heights (up to the limits that can be assessed using the
specified ground motions) permitted by the system design requirements.
Binning of index archetype configurations into performance groups provides
the basis for statistical assessment of minimum and average collapse margin
ratios for performance evaluation in Chapter 7. Performance group
populations should not be made larger than necessary, biased towards certain
configurations, or otherwise manipulated to bias the average collapse
statistics for the bin. Binning of index archetype configurations into
performance groups should be reviewed and approved by the peer review
panel.
4.3.1 Identification of Performance Groups
As illustrated by the generic performance group table shown in Table 4-3,
performance groups should be organized to consider: (1) basic structural
configuration; (2) gravity load level; (3) seismic design category; and (4)
period domain. The number of basic structural configurations will vary by
system (i.e., 1 through N), and variation in gravity load levels may (or may
not) affect the performance of certain systems. As a minimum, assuming one
basic structural configuration and no dependence on gravity loads, all
systems will have at least four performance groups based on combinations of
two seismic design levels and two period domains. These parameters should
be used as a guide in establishing performance groups, as follows:
. Basic Structural Configuration. Changes in the basic structural
configuration are intended to capture major variations in the seismicforce-
resisting system that are permissible within the design
requirements and are likely to affect the structural response. Examples
of alternative configurations that could be separated into performance
groups include: (1) variations in bracing configurations (e.g., X-bracing
versus chevron bracing) in steel concentrically braced frames; (2)
variations in framing spans and story heights in moment frames; and (3)
variations in shear wall aspect ratios that may influence whether a wall
responds in shear or flexural behavior.
. Gravity Load Level. To the extent that the gravity load intensity and
distribution of gravity loads affect the response of a seismic-forceresisting
system, gravity loads should be varied in the index archetype
configurations and separated into different performance groups.
4-12 4: Archetype Development FEMA P695
Variation in gravity load level is related to the intensity of the gravity
load (as affected by the type of gravity framing and use of the structure)
and the amount of seismic mass that is tributary to the seismic-forceresisting
system in the form of directly applied gravity loads. In moment
frames, tributary gravity loads can be distinguished through the use of
perimeter frames versus space frame configurations. In wall systems,
differences can be distinguished through the use of bearing walls (where
most of the gravity loads are directly supported by the walls) versus nonbearing
walls (where gravity loads are supported by other means, such as
frames).
Table 4-3 Generic Performance Group Matrix
Performance Group Summary
Group
No.
Grouping Criteria
Number of
Basic Archetypes
Configuration
Design Load Level Period
Gravity Seismic Domain
PG-1
Type 1
High
Max SDC
Short ..3.
PG-2 Long ..3.
PG-3
Min SDC
Short ..3.
PG-4 Long ..3.
PG-5
Low
Max SDC
Short ..3.
PG-6 Long ..3.
PG-7
Min SDC
Short ..3.
PG-8 Long ..3.
PG-9
Type 2
High
Max SDC
Short ..3.
PG-10 Long ..3.
PG-11
Min SDC
Short ..3.
PG-12 Long ..3.
PG-13
Low
Max SDC
Short ..3.
PG-14 Long ..3.
PG-15
Min SDC
Short ..3.
PG-16 Long ..3.
PG-17
Type N
High
Max SDC
Short ..3.
PG-18 Long ..3.
PG-19
Min SDC
Short ..3.
PG-20 Long ..3.
PG-21
Low
Max SDC
Short ..3.
PG-22 Long ..3.
PG-23
Min SDC
Short ..3.
PG-24 Long ..3.
. Seismic Design Category. In concept, the full range of Seismic Design
Categories for which the seismic-force-resisting system will be permitted
FEMA P695 4: Archetype Development 4-13
should be reflected in the index archetype configurations and separated
into different performance groups. Generally, however, it should suffice
to check a system for the maximum and minimum spectral intensities of
the highest Seismic Design Category (SDC) in which the system will be
permitted. For example, systems intended for SDC D should be
designed and assessed for the maximum and minimum spectral
acceleration values (SDC Dmax and SDC Dmin) as given in Table 5-1A
and Table 5-1B of Chapter 5. Usually, designs for the maximum spectral
acceleration of the highest Seismic Design Category will control the
collapse performance of the system, which is an indication that
assessment of performance in lower Seismic Design Categories will not
be required. If, however, the minimum spectral acceleration of the
highest Seismic Design Category controls performance, then the system
should also be designed and assessed for the minimum spectral
acceleration of the next lowest Seismic Design Category.
. Period Domain. Differences in fundamental period, T, between shortperiod
and long-period systems should be reflected in the index
archetype configurations and separated into different performance
groups. Period domain is described in Chapter 5, and defined by the
boundary between the constant acceleration and constant velocity regions
of the design spectrum. Since, within a given structural system, building
period varies with the building height, bins of short-period and longperiod
archetypes will typically be distinguished by building height (or
number of stories). The range of index archetypes should generally
extend from short-period configurations for one-story buildings up to
long-period configurations for the tallest practical buildings within the
archetype design space. For the potentially limited number of shortperiod
building archetypes, development of index archetypes should
consider variations in both the number of stories and the story height.
Each performance group should include at least three index archetypes.
There is no maximum number of archetypes in each performance group, but
it is expected that each group will typically have three to six index archetype
configurations. This minimum requirement may be waived if it is infeasible
to have three alternative designs within a specific performance group. For
example, in the case of flexible moment frame systems, it may not be
possible to have three distinct index archetype configurations within the
short-period domain. Further guidance on the minimum number of
performance groups, and index archetype configurations in each group, is
provided in Chapter 7.
FEMA P695 5: Nonlinear Model Development 5-1
Chapter 5
Nonlinear Model Development
This chapter describes the development of analytical models for collapse
assessment of a proposed seismic-force-resisting system. It first defines how
index archetype designs are prepared from index archetype configurations
using the proposed design requirements for the system of interest. It then
outlines how index archetype models are developed using nonlinear
component properties and limit state criteria developed and calibrated with
test data. Index archetype models are based on structural system archetypes
defined in Chapter 4, and are used to perform nonlinear analyses described in
Chapter 6.
5.1 Development of Nonlinear Models for Collapse
Simulation
Since the index archetype configurations are developed with explicit
consideration of features to be investigated through nonlinear collapse
simulation, the development of nonlinear models and structural system
archetypes is interdependent. Nonlinear model development includes
preparation of: (1) index archetype designs, which are index archetype
configurations that have been proportioned and detailed using the design
requirements for a proposed seismic-force-resisting system; and (2) index
archetype models, which are idealized mathematical representations of index
archetype designs used to simulate collapse in nonlinear static and dynamic
analyses.
Development of nonlinear models for collapse simulation follows the process
outlined in Figure 5-1. Using system design requirements (Chapter 3) and
structural system archetypes (Chapter 4), design criteria and material test
data are applied to each index archetype configuration to develop index
archetype designs based on trial values of R, Cd..and .o
. Each design is then
idealized into an index archetype model for nonlinear analysis (Chapter 6).
Model parameters and collapse assessment criteria are calibrated to
component, connection, assembly, and system tests. Quality ratings for
index archetype models are used to assess modeling uncertainty in Chapter 7.
5.2 Index Archetype Designs
Index archetype designs for a seismic-force-resisting system are prepared by
applying the proposed design requirements, substantiated by test data, to a set
5-2 5: Nonlinear Model Development FEMA P695
of index archetype configurations. Index archetype designs should include
all significant design features that are likely to affect structural response and
collapse behavior. All seismic-force-resisting components and connections
should be designed in strict accordance with the minimum requirements of
the seismic design provisions for the proposed system. Designs should also
meet applicable provisions of ASCE/SEI 7-05 (or other model building code)
and any other referenced standards.
Figure 5-1 Process for development of index archetype models.
While index archetype designs are intended to interrogate the feasible range
of the archetype design space, they are not intended to capture all feasible
“outliers” of superior or poor seismic performance. Nor are they intended to
interrogate seismic design criteria that are common across all seismic-forceresisting
systems.
FEMA P695 5: Nonlinear Model Development 5-3
Explicit assessment of redundancy, structural irregularities, soil-structure
interaction, importance factors, and other general seismic design criteria are
not addressed in the Methodology. It is assumed that such generic design
requirements of ASCE/SEI 7-05 (or other applicable building codes) are
equally effective for all seismic-force-resisting systems. It is, therefore,
important that index archetype designs comply with all relevant limitations in
ASCE/SEI 7-05 provisions that are generic to all systems.
The following sections describe design methods, seismic criteria, design
loads, load combinations and related requirements for preparing index
archetype designs. They are included here to illustrate how seismic design
base shears are calculated relative to the Maximum Considered Earthquake
(MCE) spectrum used to assess collapse margin, and to highlight specific
aspects of seismic design criteria that are relevant to the Methodology.
These criteria are primarily based on the design requirements contained
within ASCE/SEI 7-05, modified as appropriate for design of index
archetypes. While the specific requirements discussed here are based on
ASCE/SEI 7-05, this is not meant to imply a limitation on the application of
the Methodology, which is intended to be generally applicable to any set of
comprehensive seismic design procedures.
5.2.1 Seismic Design Methods
In general, the Equivalent Lateral Force (ELF) method of Section 12.8,
ASCE/SEI 7-05, should be used to develop index archetype designs, except
as noted below:
. The Response Spectrum Analysis (RSA) method of Section 12.9,
ASCE/SEI 7-05, should be used to develop archetype designs when the
ELF method is not permitted by ASCE/SEI 7-05, or by the specific
design requirements of the system of interest. For example, the ELF
method is not permitted for the design of taller structures in Seismic
Design Category D that have a fundamental period, T, greater than 3.5Ts
(Table 12.6-1, ASCE/SEI 7-05).
. Response history analysis (RHA) methods should be used to develop
archetype designs when ELF and RSA methods are not permitted by
ASCE/SEI 7-05, or by the specific design requirements of the system of
interest. For example, response history methods are required for design
of seismically isolated structures with certain performance characteristics
(Section 17.4.2.2, ASCE/SEI 7-05).
5-4 5: Nonlinear Model Development FEMA P695
. The RSA method (or RHA methods) may be used to develop index
archetype designs when such methods are (or are expected to be)
commonly used in practice in lieu of the ELF method.
5.2.2 Criteria for Seismic Design Loading
The provisions of ASCE/SEI 7-05 specify seismic loads and design criteria
in terms of Seismic Design Category (SDC), which is a function of the level
of design earthquake (DE) ground motions and the Occupancy Category of
the structure. The Methodology is based on life safety performance and
assumes all structures to be either Occupancy Category I or II (i.e., structures
that do not have special functionality requirements) with a corresponding
importance factor equal to unity. Seismic Design Categories for Occupancy
I and II structures vary from SDC A to SDC E in regions of the lowest and
highest seismicity, respectively.
The Methodology defines MCE and DE ground motions for structures in
Seismic Design Categories B, C, and D. The Methodology ignores both
SDC A structures, which are not subject to seismic design (other than the
minimum, 1% lateral load specified by Section 11.7.2 of ASCE/SEI 7-05),
and SDC E structures which are located in deterministic MCE ground motion
regions near active faults.
The provisions of ASCE/SEI 7-05 define MCE demand in terms of mapped
values of short-period spectral acceleration, SS, and 1-second spectral
acceleration, S1, site coefficients, Fa and Fv, and a standard response
spectrum shape. For seismic design of the structural system, ASCE/SEI 7-05
defines the DE demand as two-thirds of the MCE demand. The
Methodology requires archetypical systems to be designed for DE seismic
criteria, and then evaluated for collapse with respect to MCE demand.
For Seismic Design Categories B, C, and D, maximum and minimum ground
motions are based on the respective upper-bound and lower-bound values of
MCE and DE spectral acceleration, as given in Table 11.6-1 of ASCE/SEI
7-05, for short-period response, and in Table 11.6-2 of ASCE/SEI 7-05, for
1-second response. MCE spectral accelerations are derived from DE spectral
accelerations for site coefficients corresponding to Site Class D (stiff soil)
following the requirements of Section 11.4 of ASCE/SEI 7-05. For the
purpose of assessing performance across all possible site classifications, the
Methodology uses values based on the default Site Class D uniformly for
design of all archetypes.
Tables 5-1A and 5-1B list values of maximum and minimum spectral
acceleration, site coefficients, and design parameters for Seismic Design
FEMA P695 5: Nonlinear Model Development 5-5
Categories B, C, and D. Figure 5-2 shows DE response spectra for ground
motions associated with these parameters, based on the standard shape of the
design response spectrum shown in Figure 11.4-1 of ASCE/SEI 7-05.
Table 5-1A Summary of Mapped Values of Short-Period Spectral
Acceleration, Site Coefficients and Design Parameters for
Seismic Design Categories B, C, and D
Seismic Design Category Maximum Considered Earthquake Design
Maximum Minimum SS (g) Fa SMS (g) SDS (g)
D 1.5 1.0 1.5 1.0
C D 0.55 1.36 0.75 0.50
B C 0.33 1.53 0.50 0.33
B 0.156 1.6 0.25 0.167
Table 5-1B Summary of Mapped Values of 1-Second Spectral Acceleration,
Site Coefficients and Design Parameters for Seismic Design
Categories B, C, and D
Seismic Design Category Maximum Considered Earthquake Design
Maximum Minimum S1 (g) Fv SM1 (g) SD1 (g)
D 0.601 1.50 0.90 0.60
C D 0.132 2.28 0.30 0.20
B C 0.083 2.4 0.20 0.133
B 0.042 2.4 0.10 0.067
1. Value of 1-second MCE spectral acceleration rounded to 0.60 g.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
0 0.5 1 1.5 2 2.5 3 3.5 4
Period (seconds)
Spectral Acceleration (g)
DE SDC D (maximum)
DE SDC D (min) or SDC C (max)
DE SDC C (min) or SDC B (max)
DE SDC B (minimum)
Figure 5-2 Plots of design earthquake (DE) response spectral accelerations
used for design of Seismic Design Category D, C and B structure
archetypes, respectively.
5-6 5: Nonlinear Model Development FEMA P695
Maximum values of spectral acceleration for SDC D (Ss = 1.5 g and S1 = 0.60
g) are based on the effective boundary between deterministic (near-source)
and probabilistic regions of MCE ground motions, as defined in Section 21.2
of ASCE/SEI 7-05. The Methodology purposely excludes SDC D structures
at deterministic (near-source) sites defined by 1-second spectral acceleration
equal to or greater than 0.60 g, although Section 11.6 of ASCE/SEI 7-05
defines SDC D structures as having 1-second spectral acceleration values as
high as 0.75 g. The 1-second value of MCE spectral acceleration shown in
Table 5-1B, S1 = 0.60, is rounded upward slightly for convenience, and
should be taken as S1 < 0.60g, for the purpose of evaluating minimum base
shear design requirements. This is done to avoid triggering Equation 12.8-6
of ASCE-SEI 7-05 for design of index archetypes.
An internal study, documented in Appendix A, found that the collapse
margin ratio (CMR) was somewhat smaller in systems designed using SDC E
seismic criteria and evaluated using near-field ground motions, than in
systems designed using SDC D seismic criteria and evaluated using far-field
ground motions. By ignoring SDC E structures, the Methodology implicitly
accepts a somewhat greater collapse risk for buildings located close to active
faults. This is consistent with the approach in ASCE/SEI 7-05, which
implicitly accepts greater risk for buildings near active faults by limiting
MCE ground motions to deterministic values of seismic hazard.
5.2.3 Transition Period, Ts
The Methodology requires statistical evaluation of short-period archetypes
separately from long-period archetypes, and distinguishes between them on
the basis of the transition period, Ts. The transition period defines the
boundary between the region of constant acceleration and the region of
constant-velocity of the design (or MCE) response spectrum, as illustrated in
Figure 5-2. The transition period, Ts, is defined as:
D1 M1
s
DS MS
T S S
S S
. . (5-1)
where the values of SD1 and SDS (and SM1 and SMS ) are given in Table 5-1A
and Table 5-1B, respectively, for each Seismic Design Category. The value
of Ts is 0.6 seconds for the upper bound of Seismic Design Category D
(SDC Dmax), 0.4 seconds for the lower bound of Seismic Design Category D
(SDC Dmin), and 0.4 seconds for the upper and lower bounds of the other
Seismic Design Categories.
FEMA P695 5: Nonlinear Model Development 5-7
5.2.4 Seismic Base Shear, V
Index archetypes are designed using the seismic base shear, V, as defined by
Equation 12.8-1 of ASCE/SEI 7-05:
V . CsW (5-2)
where Cs is the seismic response coefficient and W is the effective seismic
weight. This base shear equation is used as the basis of the applied forces
when the Equivalent Lateral Force procedure is used for design, and to scale
design values in accordance with Section 12.9.4 of ASCE/SEI 7-05 when the
Response Spectrum Analysis procedure is used for design. The seismic
coefficient, Cs, is defined for short-period archetypes (T = Ts) as:
R
C SDS
s . (5-3)
and for long-period archetypes (T > Ts) as:
D1 0.44
s DS
C S S
T R
. . (5-4)
where SD1 and SDS are given in Tables 5-1A and 5-1B, respectively, R is the
trial value of the response modification factor, and T is the fundamental
period of the index archetype.
Equations 5-3 and 5-4 have an implied occupancy importance factor of I=1.0,
since the index archetype designs are defined for Occupancy Categories I and
II. These equations are also constrained by the minimum seismic coefficient,
Cs = 0.01, required by Equation 12.8-5 of ASCE/SEI 7-05 and by the
minimum static lateral force, Fx = 0.01wx, required by Equation 11.7-1 of
ASCE/SEI 7-05 for design of all structures. Note that Equation 12.8-6 of
ASCE-SEI 7-05 does not apply since the Methodology defines 1-second
MCE spectral acceleration as S1 < 0.60 g, in all cases.
ASCE/SEI 7-05 reduces the value of Cs for very long period structures that
have a fundamental period, T > TL, where TL is the transition period between
the constant velocity and constant displacement response domains. Values of
TL range from 4 seconds to 16 seconds, as defined in Section 11.4.5 of
ASCE/SEI 7-05. Reduced values of Cs for very long period structures do not
apply, since the Methodology limits index archetype designs to
configurations with fundamental periods less than 4 seconds (due to possible
limitations on the low frequency content of ground motion records used for
the nonlinear dynamic analysis).
5-8 5: Nonlinear Model Development FEMA P695
5.2.5 Fundamental Period, T
The fundamental period, T, is used within the Methodology in two ways.
First, it is used in establishing the design base shear through Equation 5-4.
Second, it is used in defining the ground motion spectral intensity to establish
the collapse margin ratio (CMR) in nonlinear dynamic analysis procedures
(Chapter 6). In both cases, the Methodology defines the fundamental period,
T, as:
x 0.25 seconds
u a u t n T . C T . C C h . (5-5)
where hn is the building height, the values of the coefficient, Cu, are given in
Table 12.8-1 of ASCE/SEI 7-05, and values of period parameters Ct and x are
given in Table 12.8-2 of ASCE/SEI 7-05. The approximate fundamental
period, Ta, is based on regression analysis of actual building data, and
represents lower-bound (mean minus one standard deviation) values of
building period (Chopra et al., 1998). The value of the coefficient, Cu, ranges
from 1.4 in high seismic regions to 1.7 in low seismic regions, and the
product, CuTa, approximates the average value of building period.
Alternative formulas for estimating the approximate fundamental period of
masonry or concrete shear wall structures are given in Section 12.8.2.1 of
ASCE/SEI 7-05, and may be used in lieu of Ta in Equation 5-5. Alternative
formulas for the approximate fundamental period of proposed seismic-forceresisting
systems with different dynamic characteristics should be specified
as part of the design requirements for such systems.
Use of Equation 5-5 provides a consistent basis for determining the building
period for the purpose of calculating the design base shear and evaluating the
MCE spectral intensity at collapse. Based on scatter in ground motion
spectra at small periods, the fundamental period, T, as calculated by Equation
5-5 (or other alterative formulas) includes a lower limit of 0.25 seconds for
the purpose of evaluating the CMR.
5.2.6 Loads and Load Combinations
Index archetype designs should be prepared considering gravity and seismic
loading, in accordance with the seismic load effects and load combinations of
Section 12.4 of ASCE/SEI 7-05 and guidance provided in this section.
Basic seismic load combinations for strength design (ignoring snow load, S,
and foundation load, H) are:
.1.2 0.2 . DS E . S D .Q . L (5-6a)
.0.9 0.2 . DS E . S D.Q (5-6b)
FEMA P695 5: Nonlinear Model Development 5-9
where D includes the structural self weight and superimposed dead loads, L
is the live load (including appropriate live load reduction factors), and QE is
the effect of horizontal seismic forces resulting from the base shear, V. The
basic load combinations, defined above, purposely do not include the
redundancy factor, ., which is conservatively assumed to be 1.0 in all cases.
Seismic loads, and hence capacity of index archetype designs, should not be
increased for a possible lack of redundancy that may not exist in all
applications of the proposed system.
Where the seismic load effect with overstrength is required, basic load
combinations for strength design (ignoring snow load, S, and foundation
load, H) are:
.1.2 0.2 . DS O E . S D.. Q . L (5-7a)
.0.9 0.2 . DS OE . S D.. Q (5-7b)
where .o
is the overstrength factor. Snow load, S, and foundations loads, H,
are not required for design of index archetypes. Snow load is an
environmental load that varies independently of seismic intensity, and is not
considered to be a primary factor affecting seismic performance. Foundation
loads, H, are not typical and when present, generally apply to design of
structural components that do not affect performance of the seismic-forceresisting
system.
Wind load is also not required for design of index archetypes, since wind
load does not occur (in load combinations) with earthquake load. However,
index archetypes may be designed for minimum values of wind load in lieu
of earthquake load when minimum values of wind load exceed earthquake
load. If wind load is used for archetype design, minimum values of wind
load should be based on the lowest basic wind speed in Chapter 6 of
ASCE/SEI 7-05 for all regions of the United States, and consider building
(plan) configurations which minimize lateral forces due to wind load.
5.2.7 Trial Values of Seismic Performance Factors
Preparation of index archetype designs requires selection of trial values of
the response modification coefficient, R, displacement amplification
coefficient, Cd, and overstrength factor, .o
. Initial values of R, Cd, and .o
may need to be revised, and the index archetypes redesigned, based on the
outcome of the performance evaluation process in Chapter 7.
A trial value of the response modification factor, R, is required for all index
archetype designs to determine seismic base shear, V, and the related effect
5-10 5: Nonlinear Model Development FEMA P695
of horizontal seismic force, QE. For certain archetypes, Equations 5-7a and
5-7b require a trial value of the overstrength factor, .O, for design of
structural components that are subject to rapid deterioration and are sensitive
to overload conditions. For index archetype designs governed by drift,
Section 12.8.6 of ASCE/SEI 7-05 requires a trial value of the deflection
amplification factor, Cd, to determine story drift. Guidance on initial
selection of trial values may be found in the acceptance criteria of Chapter 7.
5.2.8 Performance Group Design Variations
As described in Chapter 4, performance groups for each archetype
configuration include design load variations based on Seismic Design
Category and gravity load intensity, and design height variations that
influence the fundamental period of the structure.
Maximum and Minimum Seismic Loads. Strictly speaking, index
archetype designs should reflect all Seismic Design Categories for which the
seismic-force-resisting system will be permitted. Typically, however, the
highest SDC will have the smallest collapse margin ratios, and govern
system performance. To reasonably cover the design space, the performance
groupings (Chapter 4) require assessment of index archetypes that are
designed for both the maximum and minimum spectral intensities of the
highest SDC in which the system is allowed. For example, index archetypes
for systems permitted in all SDCs must be designed for SDC Dmax and SDC
Dmin. Similarly, index archetypes for systems permitted in SDC A or SDC B
must be designed for SDC Bmax and SDC Bmin.
In certain cases, performance might be controlled by lower SDCs. For
example, if collapse margin ratios calculated for index archetypes designed
using SDC Dmax are not consistently larger than those for index archetypes
designed using SDC Dmin, then lower SDCs might control collapse
performance. In such cases, index archetypes should also be designed and
assessed for minimum spectral accelerations corresponding to the next lowest
SDC. If index archetypes designed for the next lower SDC are found to have
even lower collapse margin ratios, then additional index archetypes designs
should be prepared to confirm whether or not lower SDCs control system
performance.
High and Low Gravity Loads. Where gravity loads tributary to the
seismic-force-resisting system significantly influence collapse behavior,
index archetype designs must be prepared for high gravity and low gravity
load intensities. High and low gravity load intensities reflect differences in
the self weight of alternative gravity framing components (e.g., metal deck
FEMA P695 5: Nonlinear Model Development 5-11
versus concrete slabs) as well as framing configurations (e.g., space frame
versus perimeter frame configurations or bearing wall versus non-bearing
wall components). Where gravity load effects are significant, design dead
load, D, and live load, L, intensities should be established considering
different building occupancies and the range of likely gravity load intensities
in the archetype design space.
Building Height. Index archetypes must be designed to populate
performance groups in the short-period (T = Ts) and long-period (T > Ts)
ranges. Equation 5-5 can be used to relate building height to number of
stories and story heights. Ideally, each performance group should have at
least 3 index archetype designs. However, flexible systems (e.g., steel
moment frames) can have relatively long fundamental periods, T, greater
then Ts, potentially limiting the number of index archetype designs in shortperiod
performance groups. Conversely, systems with building height limits
may have few archetypes with fundamental periods, T, greater than Ts,
limiting the number of archetypes in long-period performance groups. In
such cases, performance groups are permitted to have less than three
archetypes, provided that the group has at least one index archetype design
for each feasible building height (combination of number of stories and story
heights).
Unless restricted by height, long-period performance groups should contain
index archetypes of different heights (number of stories) that have
fundamental periods ranging from about Ts to about 4 seconds. Index
archetypes should be designed such that the fundamental periods of the
performance group are well distributed over the full range of periods. It
should be noted that more than three archetypes per long-period performance
group may be required to properly evaluate the performance of taller
systems.
5.3 Index Archetype Models
General considerations for developing index archetype models are
summarized in Table 5-2. These considerations should be used as a guide in
establishing index archetype models, as follows:
. Model Idealization. Definition of index archetype models includes
selection of the type of idealization used to represent structural behavior.
At the one extreme are nonlinear continuum finite element models,
which, in theory, are capable of representing the underlying structural
mechanics most directly. At the other extreme are phenomenological
models, which represent the overall force-deformation response through
concentrated nonlinear springs. A nonlinear beam-column hinge model
5-12 5: Nonlinear Model Development FEMA P695
is an example of such a phenomenological model, in which momentrotation
behavior is related to beam-column design parameters through
semi-empirical models that are calibrated to beam-column subassembly
tests.
In between these two extremes are models that utilize both continuum
and phenomenological representations. A “fiber-type” model of a
reinforced concrete shear wall is an example of such a combined model,
where flexural effects are modeled with uniaxial stress-strain behavior
for reinforcing steel and concrete, and where shear behavior (or
combined shear-flexural behavior) is represented through a stressresultant
(force-based) phenomenological model. Regardless of type,
models must be validated against test data and other substantiating
evidence to assess how accurately they capture nonlinear response and
critical limit state behavior.
Table 5-2 General Considerations for Developing Index Archetype Models
Model Attributes Considerations
Mathematical Idealization
. Continuum (physics-based)
versus phenomenological
elements
Plan and Elevation Configurations
. Number of moment frame bays,
regularity.
. Planar versus 3-D wall
representations, openings,
coupling beams, regularity.
. Number of bracing bays,
bracing configuration, regularity.
. Variations to reflect diaphragm
effects on stiffness and 3-D
force distributions
2-D versus 3-D
Component Behavior
. Prevalence of 2-D versus 3-D
systems in design practice
. Impact on structural response,
including provisions for 3-D
(out-of-plane failures) in 2-D
models
2-D versus 3-D
System Behavior
. Characteristics of index
archetype configurations, such
as diaphragm flexibility
. Impact on structural response
that is specific to certain
structural systems
. Elevation and Plan Configurations. Representation of elevation and
plan configurations in index archetype models will depend on both the
index archetype configurations and the structural system behavior.
While vertical and horizontal irregularities will certainly influence
collapse, for the purpose of evaluating general design provisions,
currently permissible elevation and plan irregularities in ASCE/SEI 7-05
FEMA P695 5: Nonlinear Model Development 5-13
are not addressed in index archetype models. As illustrated by examples
in Chapter 9, two-dimensional, three-bay frames of regular proportions
are judged sufficient to represent typical behavior of reinforced concrete
moment frame systems. The extent to which this type of model will
suffice for studies of other moment frame types should be established
based on the specific behavioral effects of the specific moment frame
system. For walls, the issue of planar versus three-dimensional response
is a key consideration, as is the presence of wall openings, boundary
elements, and coupling beams. For example, reinforced concrete walls
with large boundary members (e.g., flanged walls) are likely to exhibit
more shear-critical behavior than planar walls without boundary
members. For braced frame systems, one or two bays of framing are
likely to be sufficient unless the system relies on the specific interaction
between two adjacent bays. Representation of alternative brace
configurations is likely to be a dominant variable in collapse assessment
of braced-frame systems. Where diaphragm flexibility has a significant
effect on the lateral system response and performance, this flexibility
should be incorporated in the index archetype model.
. Two-Dimensional versus Three-Dimensional Component Behavior.
The need for models that simulate two-dimensional versus threedimensional
behavior will generally depend on: (1) the type of structural
configurations common in the design space; and (2) the expected
influence of three-dimensional effects on structural response. For most
structural framing types, two-dimensional models are likely to be
sufficient. However, there may be cases where three-dimensional
behavior (e.g., out-of-plane torsional-flexural instability of laterally
unbraced beam-columns or braces) or three-dimensional geometry (e.g.,
reinforced-concrete C-shaped core walls) are important to simulate. For
wall systems, two-dimensional wall models may be sufficiently accurate
for some system configurations (e.g., wooden shear walls, planar
reinforced concrete walls) but less accurate and perhaps inappropriate for
others (e.g., C-shaped and I-shaped reinforced concrete core walls).
. Two-Dimensional versus Three-Dimensional System Behavior.
System behavior involves the interaction of multiple seismic-forceresisting
components distributed spatially within a structure.
Introduction of different spatial combinations, however, could lead to an
intractable number of index archetype configurations and corresponding
index archetype models. Building code provisions regarding plan
configuration and three-dimensional effects (e.g., redundancy, accidental
torsion) are usually not system specific, so in most cases, a twodimensional
system representation should be adequate. Diaphragm
5-14 5: Nonlinear Model Development FEMA P695
flexibility may require three-dimensional index archetype model
configurations if important diaphragm effects cannot be suitably
incorporated in two-dimensional models.
5.3.1 Index Archetype Model Idealization
Index archetype models should provide the most basic (generic)
representation of an index archetype configuration that is still capable of
distinguishing between significant behavioral modes and key design features
of the proposed seismic-force-resisting system. Index archetype models
should be developed in cooperation with the peer review panel.
The mathematical idealization of index archetype models should capture all
significant nonlinear effects related to the collapse behavior of the system.
This can be done through: (1) explicit simulation of failure modes through
nonlinear analyses; or (2) evaluation of non-simulated1 failure modes using
alternative limit state checks on demand quantities from nonlinear analyses.
Analytical models are generally distinguished by overall topology and
element type. Topology refers to two-dimensional or three-dimensional
modeling configurations. The choice of topology (2-D or 3-D) is largely a
function of the index archetype configurations. The choice of element type
depends on structural component behavior and the nature of component
degradation. Two-dimensional topologies (e.g., planar frames or walls) do
not preclude the modeling of three-dimensional effects (e.g., out-of-plane
instabilities). Conversely, three-dimensional topologies (e.g., space frames
or C-shaped walls) do not necessarily employ element types that capture all
three-dimensional behavioral effects. Thus, the modeling decisions should
be made on a case-by-case basis, depending on the specific features of the
structure system archetypes.
For simulating collapse, component models must capture strength and
stiffness degradation under large deformations. Structural components are
usually idealized as a combination of one-dimensional line-type elements
(beam-columns or axial struts) and two-dimensional continuum elements
(plane-stress or plate/shell finite elements). Three-dimensional continuum
elements (brick finite elements) may be appropriate and necessary in some
cases. Within each element type, element formulations can be further
distinguished by the extent to which the underlying structural behavior is
modeled explicitly or through phenomenological representations. For
1 The term “non-simulated” is used to describe potential modes of collapse failure
that are not explicitly captured by the index archetype model (i.e., not explicitly
simulated), but is evaluated by alternative methods of analysis and included in the
evaluation of collapse performance.
FEMA P695 5: Nonlinear Model Development 5-15
example, nonlinear beam-column elements can range in sophistication from
fiber-type continuum elements, in which the geometry and materials in the
cross section are modeled explicitly, to concentrated spring models, in which
the inelastic response is idealized through uni-axial or multi-axial springs.
Provided that they are accurately calibrated to the appropriate range of design
and behavioral parameters, concentrated spring models will usually be
sufficient for simulating nonlinear response of columns, beams, and beamcolumn
connections in frame systems. These models have the practical
advantage of providing a straight-forward approach to characterizing strength
and inelastic deformation characteristics. However, concentrated spring
models generally cannot represent behavioral effects beyond those present in
the underlying data. Continuum models, which generally model the physical
behavior at a more fundamental level, can, if properly formulated and
validated, represent a broader range of behavioral effects that do not rely as
much on tests to represent the specific parameters of the index archetype
designs.
Wall systems will typically require two-dimensional continuum models that
can capture significant nonlinear stress and strain variations within the walls.
Continuum models may include traditional two-dimensional plane
stress/strain finite elements, or alternative formulations that utilize
combinations of formal finite element approaches and engineering
assumptions to represent the nonlinear behavior (including the effects of
strength and stiffness degradation).
In the case of moment frame systems, for example, an index archetype model
might consist of the two-dimensional, three-bay frame shown in Figure 5-3.
This model incorporates one-dimensional line-type elements with either
concentrated spring or discrete component models to simulate the nonlinear
degrading response of beams, columns, beam-column connections, and panel
zones. Significant frame behaviors are captured in a two-dimensional
representation, and the three-bay configuration captures differences between
interior and exterior columns. The additional leaning column elements
capture P-delta effects of the seismic mass that is not tributary to the frame.
For shear wall systems, an index archetype model might be as simple as a
cantilever element that accounts for inelastic flexure and shear behavior at
the base of the wall. However, where punched shear wall geometries are
included in the index archetype configurations, then the corresponding index
archetype models would need to be more complicated.
5-16 5: Nonlinear Model Development FEMA P695
Figure 5-3 Example of index archetype model for moment resisting frame
systems
5.4 Simulated Collapse Modes
To the extent possible, index archetype models should directly simulate all
significant deterioration modes that contribute to collapse behavior.
Typically, this is accomplished through structural component models that
simulate stiffness, strength, and inelastic deformation under reverse cyclic
loading. Research has demonstrated that the most significant factors
influencing collapse response are the strength at yield, Fy, maximum strength
(at capping point), Fc, plastic deformation capacity, .p
, the post-capping
tangent stiffness, Kpc, and the residual strength, Fr (Ibarra et al., 2005). These
parameters can be used to define a component backbone curve, such as the
one shown in Figure 5-4. Recently, such a curve has been designated a
force-displacement capacity boundary (FEMA, 2009).
Cyclic deterioration, which reduces stiffness values and lowers the forcedisplacement
capacity boundary established by the monotonic backbone
curve, should also be included to the extent that it influences the collapse
response in nonlinear dynamic analyses. An example of degrading hysteretic
response is shown in Figure 5-5. Characterization of component backbone
curves and hysteretic responses should represent the median response
properties of structural components. While illustrated in an aggregate sense
in Figure 5-4 and Figure 5-5, the behavior can be modeled through elements
of varying degrees of sophistication using phenomenological or physicsbased
approaches.
FEMA P695 5: Nonlinear Model Development 5-17
F
.
Fc
Fy
Fr
.y .c .r .u
Ke
.p .pc
Effective yield strength and deformation (Fy and .y
)
Effective elastic stiffness, Ke = Fy/ .y
Strength cap and associated deformation for monotonic loading (Fc and .c
)
Pre-capping plastic deformation for monotonic loading, .p
Effective post-yield tangent stiffness, Kp = (Fc-Fy)/. .p
Post-capping deformation range, .pc
Effective post-capping tangent stiffness, Kpc = Fc/..pc
Residual strength, Fr
Ultimate deformation, .u
Figure 5-4 Parameters of an idealized component backbone curve
-1.5
-1
-0.5
0
0.5
1
1.5
-8 -6 -4 -2 0 2 4 6 8
Chord Rotation (radians)
Normalized Moment (M/My)
Non-Deteriorated
Backbone
Figure 5-5 Idealized inelastic hysteretic response of structural components
with cyclic strength and stiffness degradation.
While of lesser importance than the definition of the maximum force and
deformation at the capping point, the initial stiffness can have a significant
effect on the ductility capacity. Element-level initial stiffness should reflect
all important contributors to deformation (e.g., flexure, bond-slip, and shear),
and should be validated against component and assembly test data. An
effective initial stiffness defined as the secant stiffness from the origin
through the point of 40% of the yield strength of the element should be
5-18 5: Nonlinear Model Development FEMA P695
considered in phenomenological concentrated spring models. In continuum
models, initial stiffness is usually modeled directly. Where results are
sensitive to initial stiffness, attention should be given to effects related to
initiation of cracking or yielding that may not be considered in the model,
such as shrinkage cracking due to concrete curing and residual stresses due to
fabrication.
Figure 5-4 and Figure 5-5 are intentionally portrayed in a generic sense, since
critical response parameters will vary for each specific component and
configuration. For example, in ductile reinforced concrete components (i.e.,
special moment frames), nonlinear response is typically associated with
moment-rotation in the hinge regions where degradation occurs at large
deformations through a combination of concrete crushing, confinement tie
yielding/rupture, and longitudinal bar buckling. However, in less ductile
reinforced concrete components (i.e., ordinary moment frames), nonlinear
response may include shear failures and axial failure following shear failure.
Where the seismic-force-resisting system carries significant gravity load,
characteristic force and deformation quantities may need to represent vertical
deformation effects as well as horizontal response effects.
The development of analytical models is case specific, and no single model is
universally applicable. For many steel, reinforced concrete, and wood
components, the deterioration model proposed by Ibarra et al. (2005)
satisfactorily matches experimental results and analytical predictions.
However, this model should be utilized for a proposed system only if it can
be justified based on experimental evidence.
Referring to Figure 5-5, the backbone curve defines a boundary within which
hysteresis loops are confined. The implication is that in the analytical model,
the load-deformation response is not permitted to move outside this curve.
Such boundaries can be based on monotonic behavior, but ideally they
should be based on series of tests including monotonic loading and cyclic
loading with different loading protocols (FEMA, 2009). If such boundaries
are fixed in the analytical model (i.e., cyclic deterioration is not incorporated
explicitly), then estimates of the backbone curve parameters should account
for average cyclic deterioration, to produce a modified backbone curve. If
the initial stiffness is very different from the effective elastic stiffness, then it
may affect the response close to collapse, and should become part of the
modeling effort.
Figure 5-6 illustrates the effect of cyclic loading relative to a backbone curve
obtained from monotonic loading. In almost all cases, the plastic
deformation capacity, .p
, is reduced by cyclic loading, and in many cases it is
FEMA P695 5: Nonlinear Model Development 5-19
reduced by a considerable amount from the monotonic loading case. A
backbone curve is difficult to construct from a cyclic test (unless experience
exists from other similar specimens) and often necessitates the execution of
an additional monotonic test. If monotonic tests are not available, a curve
enveloping the cyclic test (cyclic envelope) may be used as a conservative
estimate of the modified backbone curve.
Figure 5-6 Comparison of monotonic and cyclic response, along with a
cyclic envelope curve (adapted from Gatto and Uang 2002).
If the backbone curve is obtained from a monotonic test (or is deduced based
on a cyclic deterioration model), then cyclic deterioration must be built into
the analytical model representing component behavior. Most cyclic
deterioration models are energy based (e.g., Ibarra et al., 2005; Sivaselvan
and Reinhorn, 2000). Validity of the component model must be
demonstrated through satisfactory matching of component, connection, or
assembly test date from the experimental program.
Figure 5-6 also illustrates a simplified measure of performance, which is the
deformation associated with a force value of 80% of the maximum strength
measured in the test, Fc
c. In the figure, the deformation value, .c
c, which is
obtained from the intersection of a horizontal line at 0.8Fc
c with the cyclic
envelope, can be viewed as a conservative estimate of the ultimate
deformation capacity of a component. In simplified analytical models it can
be assumed that no deterioration occurs up to this value of deformation,
provided that the strength of the component is assumed to drop to zero at
deformations larger than this value. Both Fc
c and .c
c may be different in the
positive and negative directions.
The monotonic backbone curve of Figures 5-4 and 5-6 is similar but distinct
from the generalized force-displacement curves specified in ASCE/SEI 41,
5-20 5: Nonlinear Model Development FEMA P695
Seismic Rehabilitation of Existing Buildings, (ASCE, 2006b). In ASCE/SEI
41-06, generalized force-displacement curves utilize cyclic envelopes that
incorporate some degree of cyclic degradation and, in most cases, result in
conservative estimates of median response. In this Methodology, backbone
curves are intended to represent median properties of monotonic loading
response, where cyclic strength and stiffness degradation are directly
modeled in the analysis, and statistical variations of the component response
are explicitly accounted for in the assessment process.
The type of backbone curve and cyclic hysteretic model used will also
impact the amount of equivalent viscous damping used in the model. Models
that have backbone curves with a large initial elastic region (which do not
dissipate energy under cyclic loading) will generally use higher equivalent
viscous damping than models with small initial elastic regions (which do
dissipate energy under small cycles).
While component models are expected to be rigorously calibrated to test
data, available data may not be comprehensive enough to fully calibrate the
models. Data are often particularly scarce for evaluating the capping point
and post-capping behavior that occurs at large deformations in ductile
components. In such cases, test data should be augmented by engineering
analysis and judgment to establish the modeling parameters.
An example of the development and calibration of nonlinear component
models for reinforced concrete moment frame systems is provided in
Appendix E.
5.5 Non-Simulated Collapse Modes
In cases where it is not possible (or not practical) to directly simulate all
significant deterioration modes contributing to collapse behavior, nonsimulated
collapse modes can be indirectly evaluated using alternative limit
state checks on structural response quantities measured in the analyses.
Examples of possible non-simulated collapse modes might include shear
failure and subsequent axial failure in reinforced concrete columns, fracture
in the connections or hinge regions of steel moment frame components, or
failure of tie-downs in light-frame wood shear walls. Component failures
such as these may be difficult to simulate directly.
In Figure 5-7, a non-simulated collapse (NSC) limit state is shown to occur
prior to the deformation at peak strength (and subsequent deterioration) that
is directly simulated in the model. Non-simulated limit state checks are
similar to the assessment approach of ASCE/SEI 41-06, in which component
acceptance criteria are used to evaluate specific performance targets based on
FEMA P695 5: Nonlinear Model Development 5-21
demand quantities extracted from the analyses. This approach is more of an
approximation to the actual behavior of the system. It increases the
uncertainty in analytical results and tends to provide conservative estimates
of collapse limit states. While not ideal, it is a practical approach that
provides a consistent method for evaluating the effects of deterioration and
collapse mechanisms that are otherwise difficult (or impossible) to
incorporate directly in the analytical model.
Compared to collapse that is simulated directly, non-simulated limit state
checks will generally result in lower estimates of median collapse. Nonsimulated
collapse modes are usually associated with component failure
modes, and a commonly applied assumption is that the first occurrence of
this failure mode will lead to collapse of the structure. Collapse of an entire
structure predicated on the failure of a single component can, in many cases,
be overly conservative. For this reason, local failure modes should be
directly simulated, if at all possible, to permit redistribution of forces to other
components after a limit state has been reached. Alternatively, local limit
state checks can be redefined to indirectly account for possible redistribution
of forces that is known to occur prior to collapse.
Figure 5-7 Component backbone curve showing a deterioration mode that
is not directly simulated in the analysis model.
When considered in the context of incremental dynamic analyses, nonsimulated
component limit state checks are essentially stipulating a collapse
limit prior to the point where an analysis would otherwise simulate collapse.
Figure 5-8 shows a plot of the results from an incremental dynamic analysis
of an index archetype model, which is subjected to a single ground motion
that is scaled to increasing intensities. The point denoted SC corresponds to
the collapse limit that is directly simulated in the model. The point denoted
NSC represents the collapse limit as determined by applying a component
limit state check on a potential collapse mode that is not directly simulated in
the model. In this example, the limit state check is based on story drift, but
5-22 5: Nonlinear Model Development FEMA P695
non-simulated collapse checks could be based on any other structural
response parameter measured in the analysis, such as peak force demand in
an element, or peak plastic hinge rotation demand.
Limit state checks for non-simulated collapse modes should be established
based on test data and other supporting evidence, and should be calibrated to
represent the median value of the governing response parameter that is
associated with the collapse response. When establishing limit state checks,
judgment should be exercised in relating the critical condition of a
component to the collapse response of the building system, since there are
many cases where critical limit states for isolated components will not
immediately trigger overall system collapse. Non-simulated collapse limit
states should be developed in cooperation with the Methodology peer review
panel.
Figure 5-8 Incremental dynamic analysis results showing simulated (SC) and
non-simulated (NSC) collapse modes.
5.6 Characterization of Modeling Uncertainties
In this Methodology, nonlinear analysis is used to determine the median
ground motion intensity associated with collapse of a proposed seismicforce-
resisting system. Index archetype models should, therefore, represent
the median response of structural components that constitute the proposed
system. Variability in collapse response, due to ground motion variability,
modeling, and other uncertainties, is factored into the performance evaluation
process in Chapter 7. When a model calibrated to median properties is used,
nonlinear dynamic analysis under multiple ground motions is intended to
provide a median estimate of the collapse capacity of an index archetype.
FEMA P695 5: Nonlinear Model Development 5-23
5.7 Quality Rating of Index Archetype Models
Quality of index archetype models is related to uncertainty, which factors
into the performance evaluation for a proposed seismic-force-resisting
system. The quality of index archetype models is rated in accordance with
the requirements of this section, and approved by the peer review panel.
Index archetype models are rated between (A) Superior and (D) Poor, as
shown in Table 5-3. This rating is a combined assessment of: (1) how well
index archetype models represent the range of structural collapse
characteristics and associated design parameters of the archetype design
space; and (2) how well the analysis models capture structural collapse
behavior through both direct simulation and non-simulated limit state checks.
The quantitative values of modeling-related collapse uncertainty are: (A)
Superior, .MDL = 0.10; (B) Good, .MDL = 0.20; (C) Fair, .MDL = 0.35; and (D)
Poor, .MDL = 0.50. Use of these values is described in Section 7.3.
Table 5-3 Quality Rating of Index Archetype Models
Representation of Collapse
Characteristics
Accuracy and Robustness of Models
High Medium Low
High. Index models capture the
full range of the archetype
design space and structural
behavioral effects that
contribute to collapse.
(A) Superior
.MDL = 0.10
(B) Good
.MDL = 0.20
(C) Fair
.MDL = 0.35
Medium. Index models are
generally comprehensive and
representative of the design
space and behavioral effects
that contribute to collapse.
(B) Good
.MDL = 0.20
(C) Fair
.MDL = 0.35
(D) Poor
.MDL = 0.50
Low. Significant aspects of the
design space and/or collapse
behavior are not captured in
the index models.
(C) Fair
.MDL = 0.35
(D) Poor
.MDL = 0.50
--
The highest rating of (A) Superior applies to instances in which the index
archetype models represent the complete range of structural configuration
and collapse behavior, there is a high confidence in the ability of established
models to simulate behavior, and the nonlinear model is of high-fidelity. The
combination of low quality representation of collapse characteristics along
with low quality modeling in terms of accuracy and robustness is not
permitted.
An adaptation of the Methodology to assess building-specific collapse
performance of an individual building is presented in Appendix F.
5-24 5: Nonlinear Model Development FEMA P695
Differences in assigning quality ratings for an analytical model of an
individual building are discussed there.
5.7.1 Representation of Collapse Characteristics
Representation of collapse characteristics refers to how completely and
comprehensively the index archetype models capture the full range of design
parameters and associated structural collapse behavior that is envisioned
within the archetype design space. The quality of the representation is
characterized as follows:
. High. The set of index archetype configurations and associated
archetype models provides a complete and comprehensive representation
of the full range of structural configurations, design parameters and
behavioral characteristics that affect structural collapse. The index
archetype models cover a comprehensive range of building heights,
lateral system configurations, and design alternatives that are permitted
by the design requirements. To the extent that 3-D component and
system effects are significant, they are reflected in the index archetype
models, as are other significant system effects such as diaphragm
flexibility,
. Medium. The set of index archetype models provides a reasonably
broad and complete representation of the design space. Where the
complete design space is not fully represented in the set of models, there
is reasonable confidence that the range of response captured by the
models is indicative of the primary structural behavior characteristics
that affect collapse.
. Low. The set of index archetype models does not capture the full range
of structural configurations and collapse behavior for the system due to
the combined effects of a loosely defined design space and a less than
complete set of index archetype configurations. Loosely defined limits
on system configurations and design parameters present a challenge in
that the number of possible alternative configurations and structural
design parameters are so large as to preclude systematic interrogation
with a manageable number of index archetype configurations. Seismicforce-
resisting systems permitted in low Seismic Design Categories that
have limited requirements on design (e.g., steel ordinary moment frame
systems) may fall into this category. Even for well controlled design
criteria, however, representation of collapse characteristics may be low if
the number and variety of index archetype configurations are not
insufficient to capture the possible range in collapse behavior.
FEMA P695 5: Nonlinear Model Development 5-25
5.7.2 Accuracy and Robustness of Models
Accuracy and robustness is related to the degree to which nonlinear
behaviors are directly simulated in the model, or otherwise accounted for in
the assessment. Use of non-simulated collapse limit state checks will lower
the accuracy and robustness of a nonlinear model. If conservatively applied,
however, non-simulated collapse checks should not necessarily lower the
overall quality rating of the assessment procedure. Model accuracy and
robustness are characterized as follows:
. High. Nonlinear models directly simulate all predominate inelastic
effects, from the onset of yielding through strength and stiffness
degradation causing collapse. Models employ either concentrated hinges
or distributed finite elements to provide spatial resolution appropriate for
the proposed system. Computational solution algorithms are sufficiently
robust to accurately track inelastic force redistribution, including cyclic
loading and unloading, without convergence problems, up to the point of
collapse.
. Medium. Nonlinear models capture most, but not all, nonlinear
deterioration and response mechanisms leading to collapse. Models may
not be sufficiently robust to track the full extent of deterioration, so that
some component-based limit state checks are necessary to assess
collapse.
. Low. Nonlinear models capture the onset of yielding and subsequent
strain hardening, but do not simulate degrading response. Onset of
degradation is primarily evaluated using non-simulated component limit
state checks. Overall uncertainty in response quantities is increased due
to inability to capture the effects of deterioration and redistribution.
FEMA P695 6: Nonlinear Analysis 6-1
Chapter 6
Nonlinear Analysis
This chapter describes nonlinear analysis procedures for collapse assessment
of a proposed seismic-force-resisting system. It defines the set of input
ground motions, and specifies how nonlinear static analyses and nonlinear
dynamic analyses are conducted on index archetype models developed in
Chapter 5. Nonlinear analyses are used to define the median collapse
capacity and other parameters that are needed for performance evaluation in
Chapter 7.
6.1 Nonlinear Analysis Procedures
Nonlinear analysis for collapse assessment follows the process outlined in
Figure 6-1.
Figure 6-1 Process for performing nonlinear analyses for collapse
assessment.
6-2 6: Nonlinear Analysis FEMA P695
Nonlinear static (pushover) and dynamic (response history) analyses of all
index archetype models are performed to obtain statistics for system
overstrength, period elongation, and collapse capacity. Nonlinear static
analyses are performed first, to help validate the model and to provide
statistical data on system overstrength, ., and period-based ductility, .T.
Nonlinear dynamic analyses are then performed to assess median collapse
capacities, CT Sˆ , and collapse margin ratios, CMR. Median collapse capacity
is defined as the ground motion intensity where half of the ground motions in
the record set cause collapse of an index archetype model.
In all cases, modeling parameters, including the seismic mass and imposed
gravity loads, should represent the median values of the structure and its
components. The gravity loads for analysis are different from design gravity
loads, and are given by the following load combination:
1.05D + 0.25L (6-1)
where D is the nominal dead load of the structure and the superimposed dead
load, and L is the nominal live load. Load factors in Equation 6-1 are based
on expected values (equivalent to median values for normally distributed
random variables) reported in a study on the development of a predecessor
document to ASCE/SEI 7-05 (Ellingwood et al., 1980). The nominal live
load in Equation 6-1 can be reduced by reduction factors based on influence
area (subject to the limitations in ASCE/SEI 7-05), but should not be reduced
by additional reduction factors.
As described in Chapter 5, index archetype models should account for all
seismic mass and P-delta effects associated with gravity loads that are
stabilized by the seismic-force-resisting system. This includes gravity loads
that are directly tributary to the components of the seismic-force-resisting
system, as well as gravity loads that rely on the seismic-force-resisting
system for lateral stability. Models should also account, either directly or
indirectly, for stiffness and strength degradation leading to the onset of
collapse along with energy dissipated by the building.
6.1.1 Nonlinear Analysis Software
Software for nonlinear analysis can be of any type that is: (1) capable of
static pushover and dynamic response history analyses; and (2) capable of
capturing strength and stiffness degradation in structural components at large
deformations. A significant computational challenge is to accurately capture
the negative post-peak response, sometimes referred to as strain-softening
response, in component backbone curves. Strain-softening response leads to
the need for robust iterative numerical solution strategies to minimize errors
FEMA P695 6: Nonlinear Analysis 6-3
and achieve convergence at large inelastic deformations. Problems with
strain-softening response have the potential for non-unique solutions and
damage localization that is sensitive to numerical issues. While most modern
analysis software can overcome these issues, care must be taken to
investigate the sensitivity of the solution to modeling parameters and
numerical aspects of the computational solution algorithms.
6.2 Input Ground Motions
Nonlinear dynamic response of index archetypes is evaluated for a set of predefined
ground motions that are systematically scaled to increasing
intensities until median collapse is established. The ratio between median
collapse intensity, CT Sˆ , and Maximum Considered Earthquake (MCE)
ground motion intensity, SMT, is defined as the collapse margin ratio, CMR,
which is the primary parameter used to characterize the collapse safety of the
structure.
6.2.1 MCE Ground Motion Intensity
Collapse performance is evaluated relative to ground motion intensity
associated with the MCE, as defined in ASCE/SEI 7-05, and related to the
seismic criteria used for design of index archetypes in Chapter 5.
As described in Chapter 5, the Methodology defines DE and MCE ground
motion intensities for three ranges of spectral acceleration associated with
Seismic Design Categories B, C and D. Table 6-1 summarizes MCE spectral
acceleration for maximum and minimum ground motions for these Seismic
Design Categories, and Figure 6-1 shows MCE response spectra for the
corresponding ground motion intensities.
Table 6-1 Summary of Maximum Considered Earthquake Spectral
Accelerations And Transition Periods Used for Collapse
Evaluation of Seismic Design Category D, C, and B Structure
Archetypes, Respectively
Seismic Design Category Maximum Considered
Earthquake
Transition
Period
Maximum Minimum SMS (g) SM1 (g) Ts (sec.)
D 1.5 0.9 0.6
C D 0.75 0.30 0.4
B C 0.50 0.20 0.4
B 0.25 0.10 0.4
6-4 6: Nonlinear Analysis FEMA P695
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4
Period (seconds)
Spectral Acceleration (g)
MCE SDC D (maximum)
MCE SDC D (min)/SDC C (max)
MCE SDC C (min)/SDC B (max)
MCE SDC B (minimum)
Figure 6-2 MCE response spectra required for collapse evaluation of index
archetypes designed for Seismic Design Category (SDC) B, C,
and D.
MCE ground motion intensity, SMT, is defined for short-period archetypes
(T = Ts) as:
SMT . SMS (6-2)
and for long-period archetypes (T > Ts) as:
T
S SM
MT
. 1 (6-3)
where values of SM1 and SMS are given in Table 6-1, and T is the fundamental
period of an index archetype as defined in Equation 5-5.
6.2.2 Ground Motion Record Sets
The Methodology provides two sets of ground motion records for collapse
assessment using nonlinear dynamic analysis, referred to as the Far-Field
record set and the Near-Field record set. The Far-Field record set includes
twenty-two component pairs of horizontal ground motions from sites located
greater than or equal to 10 km from fault rupture. The Near-Field record set
includes twenty-eight component pairs of horizontal ground motions
recorded at sites less than 10 km from fault rupture. The record sets do not
include the vertical component of ground motion since this direction of
earthquake shaking is not considered of primary importance for collapse
evaluation, and is not required by the Methodology for nonlinear dynamic
analysis.
FEMA P695 6: Nonlinear Analysis 6-5
The ground motion record sets each include a sufficient number of records to
permit evaluation of record-to-record (RTR) variability and calculation of
median collapse intensity, CT Sˆ . Explicit calculation of record-to-record
variability, however, is not required for collapse evaluation of index
archetypes. Instead, an estimate of record-to-record variability, based on
previous research and developmental studies, is built into the process for
calculating total system collapse uncertainty in Chapter 7. The record sets,
along with selection criteria and background information on their selection,
are provided in Appendix A.
The Methodology specifies use of the Far-Field record set for collapse
evaluation of index archetypes designed for Seismic Design Category (SDC)
B, C or D criteria (i.e., structures at sites that are located away from active
faults). The Near-Field record set is provided as supplemental information,
and is used in special studies of Appendix A to evaluate potential differences
in the CMR for SDC E structures. Figure 6-3 shows the 44 individual
response spectra (i.e., 22 records, 2 components each) of the Far-Field record
set, the median response spectrum, and spectra representing one standard
deviation and two-standard deviations above the median.
0.01
0.1
1
10
0.01 0.1 1 10
Period (seconds)
Spectral Acceleration (g)
Median Spectrum - Far-Field Set
+ 1 LnStdDev Spectrum - FF Set
+ 2 LnStdDev Spectrum - FF Set
Figure 6-3 Far-Field record set response spectra.
Both ground motion record sets include strong-motion records (i.e., records
with PGA > 0.2 g and PGV > 15 cm/sec) from all large-magnitude (M > 6.5)
events in the Pacific Earthquake Engineering Research Center (PEER) Next-
Generation Attenuation (NGA) database (PEER, 2006a). Large-magnitude
events dominate collapse risk and generally have longer durations of shaking,
which is important for collapse evaluation of nonlinear degrading models.
6-6 6: Nonlinear Analysis FEMA P695
The sets include records from soft rock and stiff soil sites (predominantly
Site Class C and D conditions), and from shallow crustal sources
(predominantly strike-slip and thrust mechanisms). To avoid event bias, no
more than two of the strongest records are taken from each earthquake.
The primary function of the Far-Field record set is to provide a fully-defined
set of records for use in a consistent manner to evaluate collapse across all
applicable Seismic Design Categories, located in any seismic region, and
founded on any soil site classification. Actual earthquake records are used,
in contrast with artificial or synthetic records, recognizing that regional
variation of ground motions would not be addressed. In the United States,
strong-motion records date back to the 1933 Long Beach Earthquake, with
only a few records obtained from each event until the 1971 San Fernando
Earthquake. Large magnitude events are rare, and few existing earthquake
ground motion records are strong enough to collapse a large percentage of
modern, code-compliant buildings.
Even with many instruments, existing strong motion instrumentation
networks (e.g., Taiwan and California) provide coverage for only a small
fraction of all regions of high seismicity. Considering the size of the earth
and period of geologic time, the available sample of strong motion records
from large-magnitude earthquakes is still quite limited, and potentially biased
by records from more recent, relatively well-recorded events. Due to the
limited number of very large earthquakes, and the frequency ranges of
ground motion recording devices, the ground motion record sets are
primarily intended for buildings with natural (first-mode) periods less than or
equal to 4 seconds. Thus, the record set is not necessarily appropriate for tall
buildings with long fundamental periods of vibration greater than 4 seconds.
6.2.3 Ground Motion Record Scaling
Ground motion records are scaled to represent a specific intensity (e.g., the
collapse intensity of the index archetypes of interest). Record scaling
involves two steps. First, individual records in each set are “normalized” by
their respective peak ground velocities, as described in Appendix A. This
step is intended to remove unwarranted variability between records due to
inherent differences in event magnitude, distance to source, source type and
site conditions, without eliminating overall record-to-record variability.
Second, normalized ground motions are collectively scaled (or “anchored”)
to a specific ground motion intensity such that the median spectral
acceleration of the record set matches the spectral acceleration at the
fundamental period, T, of the index archetype that is being analyzed.
FEMA P695 6: Nonlinear Analysis 6-7
The first step was performed as part of the ground motion development
process, so the record sets contained in Appendix A already reflect this
normalization. The second step is performed as part of the nonlinear
dynamic analysis procedure. This two-step scaling process parallels the
ground motion scaling requirements of Section 16.1.3.2 of ASCE/SEI 7-05.
Figure 6-4 shows the median spectrum of the Far-Field record set anchored
to maximum and minimum MCE response spectra of Seismic Design
Categories B, C and D, at a period of 1 second.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4
Period (seconds)
Spectral Acceleration (g)
FF Record Set Scaled to MCE SDC Dmax
FF Set Scaled to MCE SDC Dmin/Cmax
FF Set Scaled to MCE SDC Cmin/Bmax
FF Record Set Scaled to MCE SDC Bmin
Figure 6-4 Median spectrum of the Far-Field record set anchored to
maximum and minimum MCE response spectra of Seismic
Design Categories B, C and D, at a period of 1 second.
6.3 Nonlinear Static (Pushover) Analyses
Nonlinear static (pushover) analyses are conducted under the factored gravity
load combination of Equation 6-1 and static lateral forces. In general,
pushover analysis should be performed following the nonlinear static
procedure (NSP) of Section 3.3.3 of ASCE/SEI 41-06.
The vertical distribution of the lateral force, Fx, at each story level, x, should
be in proportion to the fundamental mode shape of the index archetype
model:
x , x x m F 1 .
. (6-4)
where mx is the mass at level x; and .1,x is the ordinate of the fundamental
mode at level x.
6-8 6: Nonlinear Analysis FEMA P695
Figure 6-5 shows an idealized pushover curve and definitions of the
maximum base shear capacity, Vmax and the ultimate displacement, .u
. Vmax is
taken as the maximum base shear strength at any point on the pushover
curve, and .u
is taken as the roof displacement at the point of 20% strength
loss (0.8Vmax).
Vmax
.y,eff .u
0.8Vmax
V
Base
Shear
Roof Displacement
Figure 6-5 Idealized nonlinear static pushover curve
A nonlinear static pushover analysis is used to quantify Vmax and .u
, which
are then used to compute archetype overstrength, ., and period-based
ductility, .T. In order to quantify these values, the lateral loads are applied
monotonically until a loss of 20% of the base shear capacity (0.8Vmax) is
achieved.
The overstrength factor for a given index archetype model, ., is defined as
the ratio of the maximum base shear resistance, Vmax, to the design base
shear, V:
V
. . Vmax (6-5)
The period-based ductility for a given index archetype model, .T, is defined
as the ratio of ultimate roof drift displacement, .u
, (defined as shown in
Figure 6-5) to the effective yield roof drift displacement .y,eff :
y,eff
u
T .
.
. . (6-6)
FEMA P695 6: Nonlinear Analysis 6-9
The effective yield roof drift displacement is as given by the formula:
max 2
, 2 (max( , ))
4 y eff O 1
C V g T T
W
.
.
. . . .. ..
(6-7)
where C0 relates fundamental-mode (SDOF) displacement to roof
displacement, Vmax/W is the maximum base shear normalized by building
weight, g is the gravity constant, T is the fundamental period (CuTa, defined
by Equation 5-5), and T1 is the fundamental period of the archetype model
computed using eigenvalue analysis.
The coefficient C0 is based on Equation C3-4 of ASCE/SEI 41-06, as
follows:
.
.
. N
x ,x
N
x ,x
,r
m
m
C
1
2
1
1
1
0 1
.
.
. (6-8)
where mx is the mass at level x; and .1,x (.1,r) is the ordinate of the
fundamental mode at level x (roof), and N is the number of levels.
Additional background on period-based ductility is included in Appendix B.
Since pushover analyses are intended to verify the models and provide a
conservative bound on the system overstrength factor, checks for nonsimulated
collapse modes are not incorporated directly. Non-simulated
collapse modes should be considered when evaluating ultimate roof drift
displacement, .u
.
Where three-dimensional analyses are used, separate nonlinear static
analyses should be performed to evaluate overstrength and ultimate roof drift
displacement independently along the two principle axes of the index
archetype model. The resulting values for overstrength and ultimate roof
drift displacement are then calculated by averaging the values from each of
the principle loading directions.
6.4 Nonlinear Dynamic (Response History) Analyses
Nonlinear dynamic (response history) analyses are conducted under the
factored gravity load combination of Equation 6-1 and input ground motions
from the Far-Field record set in Appendix A. Nonlinear dynamic analyses
are used to establish the median collapse capacity, CT Sˆ , and collapse margin
ratio, CMR, for each of the index archetype models. Ground motion
intensity, ST, is defined based on the median spectral intensity of the Far-
6-10 6: Nonlinear Analysis FEMA P695
Field record set, measured at the fundamental period of the structure (CuTa,
defined by Equation 5-5). Determination of the collapse margin ratio for
each index archetype model is expected to require approximately 200
nonlinear response history analyses (approximately 5 analyses of varying
intensity for each component of the 22 pairs of earthquake ground motion
records).
The following sections present background on the collapse assessment
methodology followed by specific guidelines for conducting nonlinear
dynamic analyses to calculate the collapse parameters, CT Sˆ and CMR, for
index archetype models.
6.4.1 Background on Assessment of Collapse Capacity
The median collapse intensity can be visualized through the concept of
incremental dynamic analysis (IDA) (Vamvatsikos and Cornell, 2002), in
which individual ground motions are scaled to increasing intensities until the
structure reaches a collapse point. Results from a set of an incremental
dynamic analyses are illustrated in Figure 6-6, where each point in the figure
corresponds to the results of one nonlinear dynamic analysis of one index
archetype model subjected to one ground motion record that is scaled to one
intensity level.
Figure 6-6 Incremental dynamic analysis response plot of spectral
acceleration versus maximum story drift ratio.
In Figure 6-6, the results of each analysis are plotted in terms of the spectral
intensity of the ground motion (on the vertical axis) versus maximum story
drift ratio recorded in the analysis (on the horizontal axis). Each line in
Figure 6-6 connects results for a given ground motion scaled to increasing
FEMA P695 6: Nonlinear Analysis 6-11
spectral intensities. Differences between the lines reflect differences in the
response of the same index archetype model when subjected to different
ground motions with different frequency characteristics. Collapse under
each ground motion is judged to occur either directly from dynamic analysis
results as evidenced by excessive lateral displacements (lateral dynamic
instability) or assessed indirectly through non-simulated component limit
state criteria. In Figure 6-6, the median collapse capacity of CT Sˆ = 2.8g is
defined as the spectral intensity when half of the ground motions cause the
structure to collapse.
Using collapse data from IDA results, a collapse fragility curve can be
defined through a cumulative distribution function (CDF), which relates the
ground motion intensity to the probability of collapse (Ibarra et al., 2002).
Figure 6-7 shows an example of a cumulative distribution plot obtained by
fitting a lognormal distribution through the collapse data points from Figure
6-6.
Figure 6-7 Collapse fragility curve, or cumulative distribution function.
The lognormal collapse fragility is defined by two parameters, which are the
median collapse intensity, CT Sˆ , and the standard deviation of the natural
logarithm, .RTR. The median collapse capacity ( CT Sˆ = 2.8g in the figure)
corresponds to a 50% probability of collapse. The slope of the lognormal
distribution is measured by .RTR, and reflects the dispersion in results due to
record-to-record (RTR) variability (uncertainty). In this Methodology only
the median collapse intensity, CT Sˆ , is calculated, and record-to-record
variability, .RTR, is set to a fixed value (i.e., .RTR = 0.4 for systems with
period-based ductility . 3).
Values of record-to-record variability are fixed for several reasons. First,
previous studies have shown that the record-to-record variability is fairly
constant for different structural models and record sets. Second, more
6-12 6: Nonlinear Analysis FEMA P695
precise calculation of record-to-record variability would not significantly
affect calculation of the CMR when combined with other sources of collapse
uncertainty. Finally, record-to-record variability is fixed because accurate
calculation of .RTR would require collapse data from a larger number of
ground motions than is necessary to calculate an accurate median collapse
intensity, CT Sˆ .
6.4.2 Calculation of Median Collapse Capacity and CMR
While the IDA concept is useful for illustrating the collapse assessment
procedure, the Methodology only requires identification of the median
collapse intensity, CT Sˆ , which can be calculated with fewer nonlinear
analyses than would otherwise be necessary for developing the full IDA
curve. Referring to Figure 6-6, CT Sˆ can be obtained by scaling all the records
in the Far-Field record set to the MCE intensity, SMT, and then increasing the
intensity until just over one-half of the scaled ground motion records cause
collapse. The lowest intensity at which one-half of the records cause
collapse is the median collapse intensity, CT Sˆ . Judicious selection of
earthquake intensities close to and approaching the median collapse intensity
leads to a significant reduction in the number of analyses that are required.
As a result, nonlinear response history analyses for median collapse
assessment are computationally much less involved than the full IDA
approach. While the full IDA curve is not required, a sufficient number of
response points should be plotted at increasing intensities to help validate the
accuracy in calculating the median collapse intensity.
The MCE intensity is obtained from the response spectrum of MCE ground
motions at the fundamental period, T. In Figure 6-6, the MCE intensity, SMT,
is 1.1 g, taken directly from the response spectrum for SDC Dmax in Figure
6-4. The ratio between the median collapse intensity and the MCE intensity
is the collapse margin ratio, CMR, which is the primary parameter used to
characterize the collapse safety of the structure.
MT
CT
S
CMR S
ˆ
. (6-9)
6.4.3 Ground Motion Record Intensity and Scaling
In the Methodology, ground motion intensities are defined in terms of the
median spectral intensity of the Far-Field record set, rather than the spectral
intensity of each individual record. Conceptually, this envisions the Far-
Field record set as representative of a suite of records from a characteristic
earthquake in which spectral intensities of individual records will exhibit
FEMA P695 6: Nonlinear Analysis 6-13
dispersion about the median value of the set. Thus, the median collapse
capacity of the structure is equal to the median capacity of the Far-Field
record set at the point where half of the records in the set cause collapse of
the index archetype model.
The spectral scaling intensity for the ground motion records is determined
based on the median spectral acceleration of the Far-Field record set at the
fundamental period, T, of the building. For purposes of scaling the spectra
and calculating the corresponding MCE hazard spectra, the value of
fundamental period is the same as that required for archetype design (CuTa,
defined by Equation 5-5). The spectral acceleration of the record set at the
specified period is ST, and the intensity at the collapse point for each record is
SCT (i.e., ST of the average spectra corresponding to the point when the
specified record triggers collapse). The median collapse intensity for the
entire record set (the median of 44 records) is CT Sˆ . The spectral acceleration
of the MCE hazard, SMT, is given in Table 6-1.
6.4.4 Energy Dissipation and Viscous Damping
To the extent that index archetype models are accurately calibrated for the
loading histories encountered in the dynamic analysis, most of the structural
damping will be modeled directly in the analysis through hysteretic response
of the structural components. Thus, assumed viscous damping for nonlinear
collapse analyses should be less than would typically be used in linear
dynamic analyses. Depending on the type and characteristics of the
nonlinear model, additional viscous damping may be used to simulate the
portion of energy dissipation arising from both structural and nonstructural
components (e.g., cladding, partitions) that is not otherwise incorporated in
the model. When used, viscous damping should be consistent with the
inherent damping in the structure that is not already captured by the
nonlinear hysteretic response that is directly simulated in the model.
For nonlinear dynamic analyses, equivalent viscous damping is typically
assumed to be in the range of 2% to 5% of critical damping for the first few
vibration modes that tend to dominate the response. Care should be taken to
ensure that added viscous damping does not increase beyond acceptable
levels as the model yields. The appropriate amount of damping, and
strategies to incorporate it in the assessment, should be confirmed with the
peer review panel.
6-14 6: Nonlinear Analysis FEMA P695
6.4.5 Guidelines for CMR Calculation using Three-
Dimensional Nonlinear Dynamic Analyses
For two-dimensional analyses, all forty-four ground motion records (twentytwo
pairs) are applied as independent events to calculate the median collapse
intensity, CT Sˆ , for each index archetype model. For three-dimensional
analyses, the twenty-two record pairs are applied twice to each model, once
with the ground motion records oriented along one principal direction, and
then again with the records rotated 90 degrees.
Because ground motions records are applied in pairs in three-dimensional
nonlinear dynamic analyses, collapse behavior of each index archetype
model resulting from each ground motion component is coupled.
Notwithstanding other variations between the two-dimensional and threedimensional
analyses, studies have shown that the median collapse intensity
resulting from three-dimensional analyses is on average about 20% less than
the median collapse intensity resulting from two-dimensional analyses.
Thus, the application of pairs of ground motion records in three-dimensional
analyses introduces a conservative bias as compared to results from twodimensional
analyses.
To achieve parity with the two-dimensional analyses, an adjustment should
be made when calculating the collapse margin ratio using three-dimensional
analyses. The CMR calculated based on median collapse intensity, CT Sˆ ,
obtained from three-dimensional analyses should be multiplied by a factor of
1.2. This multiplier is applied in addition to the spectral shape factor, SSF,
which is used to calculate the adjusted collapse margin ratio, ACMR, as part
of the performance evaluation in Chapter 7.
6.4.6 Summary of Procedure for Nonlinear Dynamic Analysis
Nonlinear dynamic analysis of each index archetype model includes the
following steps:
1. Using the normalized Far-Field earthquake record set in Appendix A,
scale all records to an initial scale factor. A suggested initial scale factor
is ST = 1.3SMT,, which can be adjusted up or down based on the results of
initial analyses.
2. Perform nonlinear response history analyses on each index archetype
model using all twenty-two pairs of records in the Far-Field record set.
For two-dimensional analyses, models should be analyzed separately for
each ground motion component in each pair, for a total of forty-four
analyses. For three-dimensional analyses, the twenty-two record pairs
should be applied twice to each model, once with the ground motion
FEMA P695 6: Nonlinear Analysis 6-15
records oriented along one principal direction, and then again with the
records rotated 90 degrees. Process results to check for simulated
collapse (lateral dynamic instability or excessive lateral deformations
signaling a sidesway collapse mechanism) or non-simulated collapse
(demands that exceed certain component limit state criteria applied
external to the analysis).
3. Based on results from the first set of analyses, adjust the ground motion
scale factor, and perform additional analyses until collapse is detected for
twenty-two of the forty-four ground motions. The median collapse
intensity, CT Sˆ , should be taken as the collapse spectral intensity, ST,
observed for the set of analyses where twenty-two records caused
collapse. To reduced the number of analyses, the median collapse
intensity, CT Sˆ , can be conservatively estimated from the collapse
spectral intensity, ST, observed for any set of analyses where less than
twenty-two records cause collapse. This may expedite the assessment
for index archetype designs whose actual median collapse intensity
(based on 22 records) exceeds the acceptance criteria by a significant
margin.
Other strategies can be used for systematically scaling the records to
determine the median collapse capacity. The most straightforward approach
is to systematically scale up the entire record set, in specified increments,
until collapse is detected for twenty-two of the records (or record pairs).
6.5 Documentation of Analysis Results
Nonlinear analysis results serve as the basis for performance evaluation in
Chapter 7, and are subject to review by the peer review panel. As a
minimum, information on model development, and data from nonlinear static
and dynamic analyses, should be documented in accordance with this
section.
6.5.1 Documentation of Nonlinear Models
The following information should be reported on the development of the
nonlinear index archetype models:
. Description of index archetype models, including graphical
representations of idealized models showing support and loading
conditions and member types
. Summary of modeling parameters and substantiating test data
including material strengths and stress-strain properties, component
6-16 6: Nonlinear Analysis FEMA P695
and connection strengths and deformation capacities, gravity loads
and masses, and damping parameters
. Summary of criteria and substantiating test data for non-simulated
collapse modes
. General documentation of analysis software
6.5.2 Data from Nonlinear Static Analyses
The following information should be reported from nonlinear static
(pushover) and eigenvalue analyses of each index archetype model:
. Fundamental period of vibration, T, model period of vibration, T1, and
design base shear, V
. Distribution of lateral (pushover) loads
. Plot of base shear versus roof drift
. Fully yielded strength, Vmax, and static overstrength factor . = Vmax/V
. The effective yield, .y,eff , ultimate roof displacements, .u, and periodbased
ductility, .T
. Story drift ratios at the design base shear, the maximum load Vmax, and
0.8Vmax (used to gage system behavior)
6.5.3 Data from Nonlinear Dynamic Analyses
The following information should be reported from nonlinear dynamic
(response history) analyses of each index archetype model:
. MCE ground motion intensity (MCE spectral acceleration), SMT, and the
period used to calculate this value
. Median collapse intensity, CT Sˆ , and collapse margin ratio, CMR
. Data used to compute the median collapse capacity, CT Sˆ , along with the
response parameter used to identify the collapse condition (e.g.,
maximum story drift ratio for simulated collapse, and limit-state criteria
for non-simulated collapse). Accompanying notes, plots, or narratives
describing the governing mode(s) of failure
. Representative plots of hysteresis curves for selected structural
components up to the collapse point
FEMA P695 7: Performance Evaluation 7-1
Chapter 7
Performance Evaluation
This chapter describes the process for evaluating the performance of a
proposed seismic-force-resisting system, assessing the acceptability of a trial
value of the response modification coefficient, R, and determining
appropriate values of the system overstrength factor, .O, and the deflection
amplification factor, Cd.
Performance evaluation is based on the results of nonlinear static and
dynamic analyses conducted in accordance with Chapter 6. It requires
judgment in interpreting analytical results, assessing uncertainty, and
rounding of values for design. Performance evaluation, and selection of
appropriate seismic performance factors, requires the concurrence of the peer
review panel.
7.1 Overview of the Performance Evaluation Process
The performance evaluation process utilizes results from nonlinear static
(pushover) analyses to determine an appropriate value of the system
overstrength factor, .O, and results from nonlinear static and nonlinear
dynamic (response history) analyses to evaluate the acceptability of a trial
value of the response modification coefficient, R. The deflection
amplification factor, Cd, is derived from an acceptable value of R, with
consideration of the effective damping of the system of interest.
The trial value of the response modification coefficient, R, used to design
index archetypes, is evaluated in terms of the acceptability of the collapse
margin ratio. Acceptability is measured by comparing the collapse margin
ratio, after adjustment for the effects of spectral shape, to acceptable values
that depend on the quality of the information used to define the system, total
system uncertainty, and established limits on collapse probability.
Performance evaluation follows the process outlined in Figure 7-1, and
includes the following steps:
. Obtain calculated values of system overstrength, ., period-based
ductility, .T, and collapse margin ratio, CMR, for each index archetype,
from results of nonlinear analyses (Chapter 6).
7-2 7: Performance Evaluation FEMA P695
Figure 7-1 Process for performance evaluation
. Calculate the adjusted collapse margin ratio, ACMR, for each archetype
using the spectral shape factor, SSF, which depends on the fundamental
period, T, and period-based ductility, .., as provided in Section 7.2.
. Calculate total system collapse uncertainty, .TOT, based on the quality
ratings of design requirements and test data (Chapter 3), and the quality
rating of index archetype models (Chapter 5), as provided in Section 7.3.
FEMA P695 7: Performance Evaluation 7-3
. Determine acceptable values of adjusted collapse margin ratio, ACMR10%
and ACMR20%, respectively, based on total collapse system uncertainty,
.TOT, as provided in Section 7.4.
. Evaluate the adjusted collapse margin ratio, ACMR, for each archetype
and average values of ACMR for each archetype performance group
relative to acceptable values as provided in Section 7.5.
. Evaluate the system overstrength factor, .O, as provided in Section 7.6.
. Evaluate the displacement amplification factor, Cd, as provided in
Section 7.7.
If the evaluation of ACMR finds trial values of seismic performance factors
to be unacceptable, then the system should be redefined and reanalyzed, as
needed, and then re-evaluated by repeating performance evaluation process.
Systems could be redefined by adjusting the design requirements (Chapter 3),
re-characterizing behavior (Chapter 4), or redesigning with new trial values
of seismic performance factors (Chapter 5). In general, it is expected that
more than one iteration of the evaluation process will be required to
determine optimal (and acceptable) values of the seismic performance factors
for the system of interest.
7.1.1 Performance Group Evaluation Criteria
In Chapter 4, index archetype configurations are assembled into performance
groups that reflect major differences in configuration, design and seismic
load intensity and structural period. The binning of index archetype
configurations provides the basis for statistical assessment of minimum and
average properties of seismic performance factors.
In general, trial values of seismic performance factors are evaluated for each
performance group. Results within each performance group are averaged to
determine the value for the group, which is the primary basis for judging
acceptability of each trial value. The trial value of the response modification
factor, R, must be found acceptable for all performance groups. The system
overstrength factor, .O, is based on the largest average value of overstrength,
., for all performance groups (subject to certain limits). The deflection
amplification factor, Cd, is derived from the acceptable value of R and
consideration of the effective damping of the system.
The governing performance group for the response modification factor, R, is
the one with the smallest average value of ACMR. The governing
performance group for the overstrength factor, .O, is the one with the largest
average value of ....It is likely that the response modification factor, R, and
7-4 7: Performance Evaluation FEMA P695
the system overstrength factor, .O, will be governed by different
performance groups.
Results are also evaluated to identify potential outliers within each
performance group (i.e., individual index archetypes that perform
significantly worse than the average performance of the group). Outliers can
be accommodated by adopting more conservative values of seismic
performance factors, or they can be eliminated from the archetype design
space by revising the design requirements (e.g., implementation of height
limits or other restrictions on use). Revision of seismic performance factors
or design requirements will necessitate re-design and re-analysis of index
archetypes, and re-evaluation of system performance.
It is not required that all index archetype configurations be used for
evaluation, if it can be shown by selective analysis that certain design
combinations (configurations) are not critical and do not control the
performance evaluation. Caution should be exercised in removing noncritical
index archetype configurations from a governing performance group
since their removal could adversely affect the average value of the group
used in the evaluation.
7.1.2 Acceptable Probability of Collapse
The fundamental premise of the performance evaluation process is that an
acceptably low, yet reasonable, probability of collapse can be established as a
criterion for assessing the collapse performance of a proposed system.
In this Methodology, it is suggested that the probability of collapse due to
Maximum Considered Earthquake (MCE) ground motions be limited to 10%.
Each performance group is required to meet this collapse probability limit,
on average, recognizing that some individual archetypes could have collapse
probabilities that exceed this value. A limit of twice that value, or 20%, is
suggested as a criterion for evaluating the acceptability of potential “outliers”
within a performance group.
It should be noted that these limits were selected based on judgment. Within
the performance evaluation process, these values could be adjusted to reflect
different values of acceptable probabilities of collapse that are deemed
appropriate by governing jurisdictions, or other authorities employing this
Methodology to establish seismic design requirements and seismic
performance factors for a proposed system.
FEMA P695 7: Performance Evaluation 7-5
7.2 Adjusted Collapse Margin Ratio
Collapse capacity, and the calculation of collapse margin ratio, can be
significantly influenced by the frequency content (spectral shape) of the
ground motion record set. To account for the effects of spectral shape, the
collapse margin ratio, CMR, is modified to obtain an adjusted collapse
margin ratio, ACMR, for each index archetype, i:
ACMRi . SSFi .CMRi (7-1)
This adjustment is in addition to the adjustment of the CMR made to account
for three-dimensional nonlinear dynamic analysis effects specified in Chapter
6.
7.2.1 Effect of Spectral Shape on Collapse Margin
Baker and Cornell (2006) have shown that rare ground motions in the
Western United States, such as those corresponding to the MCE, have a
distinctive spectral shape that differs from the shape of the design spectrum
used for structural design in ASCE/SEI 7-05. In essence, the shape of the
spectrum of rare ground motions is peaked at the period of interest, and drops
off more rapidly (and has less energy) at periods that are longer or shorter
than the period of interest. Where ground motion intensities are defined
based on the spectral acceleration at the first-mode period of a structure, and
where structures have sufficient ductility to inelastically soften into longer
periods of vibration, this peaked spectral shape, and more rapid drop at other
periods, causes rare records to be less damaging than would otherwise be
expected based on the shape of the standard design spectrum.
The most direct approach to account for spectral shape would be to select a
unique set of ground motions that have the appropriate shape for each site,
hazard level, and structural period of interest. This, however, is not feasible
in a generalized procedure for assessing the collapse performance of a class
of structures, with a range of possible configurations, located in different
geographic regions, with different soil site classifications. To remove this
conservative bias, simplified spectral shape factors, SSF, which depend on
fundamental period and period-based ductility, are used to adjust collapse
margin ratios. Background and development of spectral shape factors are
described in Appendix B.
7.2.2 Spectral Shape Factors
Spectral shape factors, SSF, are a function of the fundamental period, T, the
period-based ductility, .T, and the applicable Seismic Design Category.
7-6 7: Performance Evaluation FEMA P695
Table 7-1a and Table 7-1b provide values of SSF for use in adjusting the
collapse margin ratio, CMR.
Table 7-1a Spectral Shape Factor (SSF) for Archetypes Designed for SDC B,
SDC C, or SDC Dmin
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.02 1.04 1.06 1.08 1.09 1.12 1.14
0.6 1.00 1.02 1.05 1.07 1.09 1.11 1.13 1.16
0.7 1.00 1.03 1.06 1.08 1.10 1.12 1.15 1.18
0.8 1.00 1.03 1.06 1.08 1.11 1.14 1.17 1.20
0.9 1.00 1.03 1.07 1.09 1.13 1.15 1.19 1.22
1.0 1.00 1.04 1.08 1.10 1.14 1.17 1.21 1.25
1.1 1.00 1.04 1.08 1.11 1.15 1.18 1.23 1.27
1.2 1.00 1.04 1.09 1.12 1.17 1.20 1.25 1.30
1.3 1.00 1.05 1.10 1.13 1.18 1.22 1.27 1.32
1.4 1.00 1.05 1.10 1.14 1.19 1.23 1.30 1.35
. 1.5 1.00 1.05 1.11 1.15 1.21 1.25 1.32 1.37
Table 7-1b Spectral Shape Factor (SSF) for Archetypes Designed using SDC
Dmax
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.05 1.1 1.13 1.18 1.22 1.28 1.33
0.6 1.00 1.05 1.11 1.14 1.2 1.24 1.3 1.36
0.7 1.00 1.06 1.11 1.15 1.21 1.25 1.32 1.38
0.8 1.00 1.06 1.12 1.16 1.22 1.27 1.35 1.41
0.9 1.00 1.06 1.13 1.17 1.24 1.29 1.37 1.44
1.0 1.00 1.07 1.13 1.18 1.25 1.31 1.39 1.46
1.1 1.00 1.07 1.14 1.19 1.27 1.32 1.41 1.49
1.2 1.00 1.07 1.15 1.2 1.28 1.34 1.44 1.52
1.3 1.00 1.08 1.16 1.21 1.29 1.36 1.46 1.55
1.4 1.00 1.08 1.16 1.22 1.31 1.38 1.49 1.58
. 1.5 1.00 1.08 1.17 1.23 1.32 1.4 1.51 1.61
Since spectral shape factors are considerably different between SDC Dmax and
other Seismic Design Categories, the governing performance group for the
adjusted collapse margin ratio, ACMR, may not be the same as the governing
performance group for the collapse margin ratio, CMR, before adjustment.
FEMA P695 7: Performance Evaluation 7-7
7.3 Total System Collapse Uncertainty
Many sources of uncertainty contribute to variability in collapse capacity.
Larger variability in the overall collapse prediction will necessitate larger
collapse margins in order to limit the collapse probability to an acceptable
level at the MCE intensity. It is important to evaluate all significant sources
of uncertainty in collapse response, and to incorporate their effects in the
collapse assessment process.
7.3.1 Sources of Uncertainty
The following sources of uncertainty are considered in the collapse
assessment process:
. Record-to-Record Uncertainty (RTR). Record-to-record uncertainty is
due to variability in the response of index archetypes to different ground
motion records. Record-to-record variability is evident in incremental
dynamic response plots (as shown in Figure 6-5). Variability in response
is due to the combined effects of: (1) variations in frequency content and
dynamic characteristics of the various records; and (2) variability in the
hazard characterization as reflected in the Far-Field ground motion
record set. Values of record-to-record variability, .RTR, ranging from
0.35 to 0.45 are fairly consistent among various building types (Haselton,
2006; Ibarra and Krawinkler, 2005a and 2005b; Zareian et al., 2006;
Zareian, 2006). Based on available research and studies of example
archetype evaluations using the Far-Field ground motion record set in
Appendix A, a fixed value of .RTR = 0.40 is assumed in the performance
evaluation of systems with significant period elongation (i.e., periodbased
ductility, .T . 3). Most systems, even those with limited ductile
capacity, have significant period elongation before collapse, and are
appropriately evaluated using this value.
Studies in Appendix A also found that record-to-record variability can be
significantly less than .RTR = 0.40 for systems that have little, or no,
period elongation (e.g., systems with very limited ductility and certain
base-isolated systems). For these systems, values of record-to-record
variability can be reduced as follows:
. RTR . 0.1. 0.1.T . 0.40 (7-2)
where ßRTR must be greater than or equal to 0.20.
. Design Requirements Uncertainty (DR). Design requirements
uncertainty is related to the completeness and robustness of the design
requirements, and the extent to which they provide safeguards against
7-8 7: Performance Evaluation FEMA P695
unanticipated failure modes. Design requirements-related uncertainty is
quantified in terms of the quality of design requirements, rated in
accordance with the requirements in Chapter 3.
. Test Data Uncertainty (TD). Test data uncertainty is related to the
completeness and robustness of the test data used to define the system.
Uncertainty in test data is closely associated with, but distinct from,
modeling-related uncertainty. Test data-related uncertainty is quantified
in terms of the quality of test data, rated in accordance with the
requirements in Chapter 3.
. Modeling Uncertainty (MDL). Modeling uncertainty is related to how
well index archetype models represent the full range of structural
response characteristics and associated design parameters of the
archetype design space, and how well the analysis models capture
structural collapse behavior through direct simulation or non-simulated
component checks. Modeling-related uncertainty is quantified in terms
of the quality of index archetype models, rated in accordance with the
requirements in Chapter 5.
7.3.2 Combining Uncertainties in Collapse Evaluation
The total uncertainty is obtained by combining RTR, DR, TD, and MDL
uncertainties. Formally, the collapse fragility of each index archetype is
defined by the random variable, SCT, assumed to be equal to the product of
the median value of the collapse ground motion intensity, ˆ
CT S , as calculated
by nonlinear dynamic analysis, and the random variable, .TOT:
CT CT TOT S . Sˆ . (7-3)
where .TOT is assumed to be lognormally distributed with a median value of
unity and a lognormal standard deviation of .TOT. The lognormal random
variable, .TOT, is defined as the product of four component random variables
as:
TOT RTR DR TD MDL . . . . . . (7-4)
where .RTR, .MDL, .DR, and .TD are assumed to be independent and
lognormally distributed with median values of unity, and lognormal standard
deviation parameters, .RTR, .DR, .TD, and .MDL, respectively. Since the four
component random variables are assumed to be statistically independent, the
lognormal standard deviation parameter, .TOT, describing total collapse
uncertainty, is given by:
FEMA P695 7: Performance Evaluation 7-9
2
MDL
2
TD
2
DR
2
TOT RTR . . . . . . . . . (7-5)
Where: .TOT = total system collapse uncertainty (0.275 - 0.950)
.RTR = record-to-record collapse uncertainty (0.20 - 0.40)
. .DR = design requirements-related collapse uncertainty
(0.10 – 0.50)
.TD = test data-related collapse uncertainty (0.10 – 0.50)
.MDL = modeling-related collapse uncertainty (0.10 – 0.50).
The performance evaluation process does not require explicit calculation of
the lognormal distribution given by Equation 7-3 and Equation 7-4.
Acceptance criteria, however, are based on the composite uncertainty, .TOT,
developed on the basis of Equation 7-5.
7.3.3 Effect of Uncertainty on Collapse Margin
Uncertainty influences the shape of a collapse fragility curve plotted from
incremental dynamic analysis results. Figure 7-2 shows two collapse
fragility curves reflecting two different levels of uncertainty. The dashed
curve “a” reflects a .RTR = 0.4, and the solid curve “b” reflects a .TOT = 0.65.
As indicated in the figure, additional uncertainty has the effect of flattening
the curve. While the median collapse intensity, CT Sˆ , is unchanged,
additional uncertainty causes a large increase in the probability of collapse at
the MCE intensity, SMT.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 1 MT 2 3 4 5 S
CT Sˆ
Figure 7-2 Collapse fragility curves considering (a) ground motion recordto-
record uncertainty, (b) modifications for total uncertainty.
Changes in the probability of collapse at the MCE intensity will affect the
collapse margin ratio, CMR. Figure 7-3 shows collapse fragility curves for
two hypothetical seismic-force-resisting systems that have different levels of
7-10 7: Performance Evaluation FEMA P695
collapse uncertainty. In this example, both systems have been designed for
the same seismic response coefficient, CS, and happen to have the same
median collapse intensity, CT Sˆ
. System No. 1, however, has a larger
uncertainty and a “flatter” collapse fragility curve. To achieve the same 10%
probability of collapse for MCE ground motions, a larger collapse margin
ratio is required for System No. 1 than is required for System No. 2 (i.e.,
CMR1 > CMR2). Thus, System No. 1 would have to be designed using a
smaller response modification coefficient, R, than System No. 2.
Figure 7-4 shows collapse fragility curves for another set of hypothetical
seismic-force-resisting systems with different levels of collapse uncertainty.
As in Figure 7-3, both systems are designed for the same seismic response
coefficient, CS, but in this case, the two systems are also designed for the
same response modification coefficient, R. The difference in uncertainty,
however, requires different collapse margin ratios to achieve the same
median probability of collapse. In order to utilize the same response
modification coefficient, R, System No. 1, with larger uncertainty and flatter
collapse fragility, is required to have a larger collapse margin ratio than
System No. 3 (i.e., CMR1 > CMR3).
Spectral Acceleration (g)
Collapse Probability
CS SMT,2
CMR2
SMT,1
1.5R1 CMR1
0.1
0.5
0.4
0.3
0.2
0.7
0.6
Collapse Fragility
System No. 1
Collapse Fragility
System No. 2
1.5R2
CT Sˆ
Spectral Acceleration (g)
Collapse Probability
CS SMT,2
CMR2
SMT,1
1.5R1 CMR1
0.1
0.5
0.4
0.3
0.2
0.7
0.6
Collapse Fragility
System No. 1
Collapse Fragility
System No. 2
1.5R2
CT Sˆ
Figure 7-3 Illustration of fragility curves and collapse margin ratios for two
hypothetical seismic-force-resisting systems – same median
collapse level.
FEMA P695 7: Performance Evaluation 7-11
Spectral Acceleration (g)
Collapse Probability
CS
CMR3
SMT
CMR1
1.5R3
1.5R1
0.1
0.5
0.4
0.3
0.2
0.7
0.6
Collapse Fragility
System No. 3
Collapse Fragility
System No. 1
CT,1 Sˆ
CT,3 Sˆ
Spectral Acceleration (g)
Collapse Probability
CS
CMR3
SMT
CMR1
1.5R3
1.5R1
0.1
0.5
0.4
0.3
0.2
0.7
0.6
Collapse Fragility
System No. 3
Collapse Fragility
System No. 3
Collapse Fragility
System No. 1
Collapse Fragility
System No. 1
CT,1 Sˆ
CT,3 Sˆ
Figure 7-4 Illustration of fragility curves and collapse margin ratios for two
hypothetical seismic-force-resisting systems – same R factor.
7.3.4 Total System Collapse Uncertainty
Total system collapse uncertainty is calculated based on Equation 7-5, and is
a function of record-to-record (RTR) uncertainty, design requirements-related
(DR) uncertainty, test data-related (TD) uncertainty, and modeling (MDL)
uncertainty.
Quality ratings for design requirements, test data, and nonlinear models are
translated into quantitative values of uncertainty based on the following
scale: (A) Superior, . = 0.10; (B) Good, . = 0.20; (C) Fair, . = 0.35; and (D)
Poor, . = 0.50. A broad range of subjective values of uncertainty are
associated with the four quality ratings (Superior, Good, Fair and Poor) to
reward “higher quality” systems that have more robust design requirements,
more comprehensive test data, and more reliable nonlinear analysis models
with lower values of total system collapse uncertainty.
Record-to-record uncertainty and, hence, total system collapse uncertainty,
depends on period-based ductility. While subjective in nature, uncertainty
values associated with quality ratings, when combined with record-to-record
uncertainty, yield reasonable values of total system collapse uncertainty.
Resulting values range from .TOT = 0.275 for the most certain of systems (all
Superior quality ratings and .RTR = 0.2), to .TOT = 0.950 for the least certain
of systems (all Poor quality ratings and .RTR = 0.4).
7-12 7: Performance Evaluation FEMA P695
Values of total system collapse uncertainty, .TOT, for index archetype models
with a period-based ductility, .T . 3 are provided in Table 7-2a through
Table 7-2d, based on record-to-record uncertainty, .RTR = 0.4. Each of these
four tables is specific to a different model quality rating of (A) Superior, (B)
Good, (C) Fair, or (D) Poor. Values in each table are based on Equation 7-5
and the applicable combination of quality ratings for test data and design
requirements.
Table 7-2a Total System Collapse Uncertainty (.TOT) for Model Quality (A)
Superior and Period-Based Ductility, .T
. 3
Quality of Test Data
Quality of Design Requirements
(A) Superior (B) Good (C) Fair (D) Poor
(A) Superior 0.425 0.475 0.550 0.650
(B) Good 0.475 0.500 0.575 0.675
(C) Fair 0.550 0.575 0.650 0.725
(D) Poor 0.650 0.675 0.725 0.825
Table 7-2b Total System Collapse Uncertainty (.TOT) for Model Quality (B)
Good and Period-Based Ductility, .T
. 3
Quality of Test Data
Quality of Design Requirements
(A) Superior (B) Good (C) Fair (D) Poor
(A) Superior 0.475 0.500 0.575 0.675
(B) Good 0.500 0.525 0.600 0.700
(C) Fair 0.575 0.600 0.675 0.750
(D) Poor 0.675 0.700 0.750 0.825
Table 7-2c Total System Collapse Uncertainty (.TOT) for Model Quality (C)
Fair and Period-Based Ductility, .T
. 3
Quality of Test Data
Quality of Design Requirements
(A) Superior (B) Good (C) Fair (D) Poor
(A) Superior 0.550 0.575 0.650 0.725
(B) Good 0.575 0.600 0.675 0.750
(C) Fair 0.650 0.675 0.725 0.800
(D) Poor 0.725 0.750 0.800 0.875
FEMA P695 7: Performance Evaluation 7-13
Table 7-2d Total System Collapse Uncertainty (.TOT) for Model Quality (D)
Poor and Period-Based Ductility, .T
. 3
Quality of Test Data
Quality of Design Requirements
(A) Superior (B) Good (C) Fair (D) Poor
(A) Superior 0.650 0.675 0.725 0.825
(B) Good 0.675 0.700 0.750 0.825
(C) Fair 0.725 0.750 0.800 0.875
(D) Poor 0.825 0.825 0.875 0.950
In general, most archetypes are expected to have a period-based ductility of
.T . 3, and Tables 7-2a through 7-2d will be used for collapse performance
evaluation of most systems. For index archetype models that have a periodbased
ductility of .T < 3, tables have not been provided. Values of record-torecord
uncertainty, .RTR, should be calculated using Equation 7-2, and total
collapse system uncertainty, .TOT, should be calculated using Equation 7-5,
rounded to the nearest 0.025. Values of design requirements uncertainty,
.DR, test data uncertainty, .TD, and model uncertainty, .MDL, should be based
on their respective quality ratings.
7.4 Acceptable Values of Adjusted Collapse Margin
Ratio
Acceptable values of adjusted collapse margin ratio are based on total system
collapse uncertainty, .TOT, and established values of acceptable probabilities
of collapse. They are based on the assumption that the distribution of
collapse level spectral intensities is lognormal, with a median value, CT Sˆ
, and
a lognormal standard deviation equal to the total system collapse uncertainty,
.TOT.
Table 7-3 provides acceptable values of adjusted collapse margin ratio,
ACMR10% and ACMR20%, based on total system collapse uncertainty and
values of acceptable collapse probability, taken as 10% and 20%,
respectively. Other values of collapse probability ranging from 5% - 25%
are shown for comparison and reference. Lower values of acceptable
collapse probability and higher levels of collapse uncertainty result in higher
required values of adjusted collapse margin ratio.
7-14 7: Performance Evaluation FEMA P695
Table 7-3 Acceptable Values of Adjusted Collapse Margin Ratio (ACMR10%
and ACMR20%)
Total System
Collapse
Uncertainty
Collapse Probability
5% 10%
(ACMR10%)
15% 20%
(ACMR20%)
25%
0.275 1.57 1.42 1.33 1.26 1.20
0.300 1.64 1.47 1.36 1.29 1.22
0.325 1.71 1.52 1.40 1.31 1.25
0.350 1.78 1.57 1.44 1.34 1.27
0.375 1.85 1.62 1.48 1.37 1.29
0.400 1.93 1.67 1.51 1.40 1.31
0.425 2.01 1.72 1.55 1.43 1.33
0.450 2.10 1.78 1.59 1.46 1.35
0.475 2.18 1.84 1.64 1.49 1.38
0.500 2.28 1.90 1.68 1.52 1.40
0.525 2.37 1.96 1.72 1.56 1.42
0.550 2.47 2.02 1.77 1.59 1.45
0.575 2.57 2.09 1.81 1.62 1.47
0.600 2.68 2.16 1.86 1.66 1.50
0.625 2.80 2.23 1.91 1.69 1.52
0.650 2.91 2.30 1.96 1.73 1.55
0.675 3.04 2.38 2.01 1.76 1.58
0.700 3.16 2.45 2.07 1.80 1.60
0.725 3.30 2.53 2.12 1.84 1.63
0.750 3.43 2.61 2.18 1.88 1.66
0.775 3.58 2.70 2.23 1.92 1.69
0.800 3.73 2.79 2.29 1.96 1.72
0.825 3.88 2.88 2.35 2.00 1.74
0.850 4.05 2.97 2.41 2.04 1.77
0.875 4.22 3.07 2.48 2.09 1.80
0.900 4.39 3.17 2.54 2.13 1.83
0.925 4.58 3.27 2.61 2.18 1.87
0.950 4.77 3.38 2.68 2.22 1.90
FEMA P695 7: Performance Evaluation 7-15
7.5 Evaluation of the Response Modification
Coefficient, R
Acceptable performance is defined by the following two basic collapse
prevention objectives:
. The probability of collapse for MCE ground motions is approximately
10%, or less, on average across a performance group.
. The probability of collapse for MCE ground motions is approximately
20%, or less, for each index archetype within a performance group.
Acceptable performance is achieved when, for each performance group,
adjusted collapse margin ratios, ACMR, for each index archetype meet the
following two criteria:
. the average value of adjusted collapse margin ratio for each performance
group exceeds ACMR10%:
ACMRi . ACMR10% (7-6)
. individual values of adjusted collapse margin ratio for each index
archetype within a performance group exceeds ACMR20%:
ACMR ACMR % i . 20 (7-7)
7.6 Evaluation of the Overstrength Factor, .O
The average value of archetype overstrength, ., is calculated for each
performance group. The value of the system overstrength factor, .O, for use
in design should not be taken as less than the largest average value of
calculated archetype overstrength, ., from any performance group. The
system overstrength factor, .O, should be conservatively increased to
account for variation in overstrength results of individual index archetypes,
and judgmentally rounded to half unit intervals (e.g., 1.5, 2.0, 2.5, and 3.0).
The system overstrength factor, .O, need not exceed 1.5 times the response
modification coefficient, R. A practical limit on the value of .O is about 3.0,
consistent with the largest value of this factor specified in Table 12.2-1 of
ASCE/SEI 7-05 for all current approved seismic-force-resisting systems.
Example applications (Chapter 9) show that values of archetype
overstrength, ., can be as large as . = 6.0 for certain configurations, and are
highly variable. Limiting system overstrength to .O = 3.0, as specified in
ASCE/SEI 7-05, was considered necessary for practical design
considerations.
7-16 7: Performance Evaluation FEMA P695
7.7 Evaluation of the Deflection Amplification
Factor, Cd
The deflection amplification factor, Cd, is based on the acceptable value of
the response modification factor, R, reduced by the damping factor, BI,
corresponding to the inherent damping of the system of interest:
I
d B
C . R (7-8)
where: Cd = deflection amplification factor
R ... = system response modification factor
BI = numerical coefficient as set forth in Table 18.6-1 of
ASCE/SEI 7-05 for effective damping, .I
, and period, T
.I
= component of effective damping of the structure due to the
inherent dissipation of energy by elements of the structure,
at or just below the effective yield displacement of the
seismic-force-resisting system, Section 18.6.2.1 of
ASCE/SEI 7-05.
In general, inherent damping may be assumed to be 5 percent of critical, and
a corresponding value of the damping coefficient, BI = 1.0 (Table 18.6-1,
ASCE/SEI 7-05). Thus, for most systems the value of Cd will be equal to the
value of R.
Equating the deflection amplification factor, Cd, to the R factor is based on
the “Newmark rule,” which assumes that inelastic displacement is
approximately equal to elastic displacement (at the roof). This is consistent
with research findings for systems with nominal (5% of critical) damping and
fundamental periods greater than the transition period, Ts. It is recognized
that for short-period systems (T < Ts) inelastic displacement generally
exceeds elastic displacement, but it was not considered appropriate to base
the deflection amplification factor on response of short-period systems,
unless the systems are displacement sensitive. Short-period, displacement
sensitive systems should incorporate the consequences of these larger
inelastic displacements.
FEMA P695 8: Documentation and Peer Review 8-1
Chapter 8
Documentation and Peer
Review
This chapter describes documentation and peer review requirements for a
proposed seismic-force-resisting system. It identifies recommended
qualifications, expertise, and responsibilities for personnel involved with the
development and review of a proposed system. It lists information that
should be included in a report documenting the development of a system, and
discusses requirements for review at each step of the developmental process.
8.1 Recommended Qualifications, Expertise and
Responsibilities for a System Development Team
In order to collect the necessary data and apply the procedures of this
Methodology, a system development team will need to have certain
qualifications, experience, and expertise. These include the ability to:
(1) adequately test materials, components and assemblies; (2) develop
comprehensive design and construction requirements; (3) develop archetype
designs; and (4) analyze archetype models.
A development team is responsible for following the procedures of this
Methodology in determining seismic performance factors for a proposed
system, defining the limits under which a system will be applicable,
documenting results, and obtaining approval of the peer review panel.
8.1.1 System Sponsor
The system sponsor is a person or organization that has conceived a new
seismic-force-resisting system and will benefit from its use. The system
sponsor is responsible for assembling a development team, selecting an
independent peer review panel, and submitting a proposed system for
approval and use.
8.1.2 Testing Qualifications, Expertise and Responsibilities
The testing laboratories engaged in the development of a proposed system
must have the capability to perform material, component, connection,
assembly, and system tests necessary for quantifying the material and
behavioral properties of the system. Testing laboratories used to conduct the
8-2 8: Documentation and Peer Review FEMA P695
experimental investigation program should comply with national or
international accreditation criteria, such as ISO/IEC 17025, General
Requirements for the Competence of Testing and Calibration Laboratories
(ISO/IEC, 2005). Testing laboratories that are not accredited may still be
used for an experimental investigation program, subject to the approval of
the peer review panel.
Testing facility staff should have the necessary expertise to establish and
execute an experimental program, conduct the tests, and mine existing
research from other relevant available test data.
8.1.3 Engineering and Construction Qualifications, Expertise
and Responsibilities
Member(s) of a development team must have sufficient experience and
expertise to develop comprehensive design and construction requirements,
and to perform trial system designs. This should include familiarity with
seismic design requirements specified in ASCE/SEI 7, and material-specific
reference standards for design and detailing requirements for other similar
systems. To be viable for use, a proposed system must be feasible to
construct. Familiarity with proposed construction techniques, or established
construction techniques for other similar systems is needed.
8.1.4 Analytical Qualifications, Expertise and
Responsibilities
Member(s) of a development team must have sufficient experience and
expertise to interpret test data and develop sophisticated nonlinear models
capable of simulating the potential failure modes and collapse behaviors of a
proposed system. This should include knowledge and experience in the
analytical approaches specified in the Methodology, knowledge of material,
component, connection, and overall system performance, and experience
with analysis software capable of simulating system response.
8.2 Documentation of System Development and
Results
The results of system development efforts must be thoroughly documented at
each step of the process for: (1) review and approval by the peer review
panel; (2) review and approval by an authority having jurisdiction over its
eventual use; and (3) use in design and construction.
Documentation of the development of seismic performance factors for a
proposed system should include, but is not necessarily limited to, the
following:
FEMA P695 8: Documentation and Peer Review 8-3
. Description of the intended system applications and expected
performance
. Limitations on system use
. Typical horizontal and vertical geometric configurations
. Clear and complete design requirements and specifications for the
system, providing enough information to quantify strength limit states,
proportion and detail components, analyze predicted response, and
confirm satisfactory behavior
. Summary of test data and other supporting evidence from an
experimental investigation program validating material properties and
component behavior, calibrating nonlinear analysis models, and
establishing performance acceptance criteria
. Description of index archetype configurations and extent of archetype
design space
. Identification of performance groups, applicable Seismic Design
Categories, and gravity load intensities
. Idealized model configurations, nonlinear modeling parameters,
documentation of analysis software, and information used in model
calibration
. Criteria for non-simulated collapse modes
. Summary of nonlinear model results, demand parameters, and response
quantities
. Quality ratings for design requirements, test data, and nonlinear models
. Summary of performance evaluation results, derived quantities, and
acceptance criteria
. Proposed seismic performance factors (R, .O, and Cd)
8.3 Peer Review Panel
Implementation of this Methodology involves much uncertainty, judgment
and potential for variation. Deciding on an appropriate level of detail to
adequately characterize performance of a proposed system should be
performed at each step in the process in collaboration with an independent
peer review panel.
It is recommended that a peer review panel consisting of knowledgeable
experts be retained for this purpose. The peer review panel should be
familiar with the procedures of this Methodology, should have sufficient
8-4 8: Documentation and Peer Review FEMA P695
knowledge to render an informed opinion on the developmental process, and
should include expertise in each of the following areas:
. Material, component, and assembly testing
. Engineering design and construction
. Nonlinear collapse simulation
Members of the peer review panel must be qualified to critically evaluate the
development of the proposed system including testing, design, and analysis.
If a unique computer code is developed by the development team, the peer
review panel should be capable of performing independent analyses of the
proposed system using other analysis platforms.
8.3.1 Peer Review Panel Selection
It is envisioned that the cost of the peer review panel will be borne by the
system sponsor. As such, it is expected that members of the peer review
panel could be selected by the system sponsor. An alternative arrangement
could be made in which the system sponsor submits funding to the authority
having jurisdiction, which then uses the funding to implement an
independent peer review process. Such an arrangement would be similar to
the outside plan check process currently used in some building departments.
It is intended that the peer review panel be an independent set of reviewers
who will advise and guide the development team at each step in the process.
It is recommended that other stakeholders, including authorities with
jurisdiction over the eventual use of the system in design and construction,
be consulted in the selection of peer review panel members, and in the
deliberation on their findings.
8.3.2 Peer Review Roles and Responsibilities
The peer review panel is responsible for reviewing and commenting on the
approach taken by the development team including the extent of the
experimental program, testing procedures, design requirements, development
of structural system archetypes, analytical approaches, extent of the nonlinear
analysis investigation, and the final selection of the proposed seismic
performance factors.
The peer review panel is responsible for reporting their opinion on the work
performed by the developmental team, their findings, recommendations, and
conclusions. All documentation from the peer review panel should be made
available for review by the authority having jurisdiction over approval of the
proposed system.
FEMA P695 8: Documentation and Peer Review 8-5
If there are any areas where concurrence between the peer review panel and
the development team was not reached, or where the peer review panel was
not satisfied with the approach or extent of the work performed, this
information should be made available as part of the peer review
documentation, and reflected in the total uncertainty used in calculating the
system acceptance criteria, and in determining the final values of proposed
seismic performance factors.
8.4 Submittal
It is expected that a system sponsor will wish to submit a proposed system to
an authority for approval and use. For national building codes and standards,
one such authority is the National Institute of Building Science’s Building
Seismic Safety Council (BSSC) Provisions Update Committee (PUC), which
has jurisdiction over the FEMA’s National Earthquake Hazards Reduction
Program (NEHRP) Recommended Provisions for Seismic Regulations for
New Buildings and Other Structures (NEHRP Recommended Provisions).
BSSC’s PUC, along with its technical subcommittees, is a nationally
recognized leader in reviewing and endorsing new seismic force-resisting
systems for ultimate inclusion in national building codes and standards.
In some cases, a proposed system could be submitted to the ASCE/SEI 7
standard development committee, but this committee would normally only
accept systems similar to systems that are already listed in the standard.
Systems can also be submitted directly to model building codes through the
code change process.
Another approach is to promote a new system through a relevant material
standard organization, such as the American Concrete Institute (ACI),
American Institute of Steel Construction (AISC), American Iron and Steel
Institute (AISI), or the American Forest & Paper Association (AF&PA).
Approval through one of these organizations, however, will still require
adoption by national building codes or standards before use.
If a proposed system is intended for a single project application, then
documentation should be submitted, along with drawings and calculations for
the single application, to the authority having jurisdiction over the site where
the system is being proposed for use.
FEMA P695 9: Example Applications 9-1
Chapter 9
Example Applications
This chapter presents examples illustrating the application of the
Methodology to reinforced concrete special moment frame, reinforced
concrete ordinary moment frame, and wood light-frame shear wall seismicforce-
resisting systems.
9.1 General
In the following sections three seismic-force-resisting systems are evaluated
using the methods outlined Chapters 3 through 7. The examples span
different system types, design requirements, test data, archetype models, and
analysis software. Models include both simulated and non-simulated
collapse modes. Each system is currently contained in Table 12.2-1 of
ASCE/SEI 7-05, and the examples utilize design requirements and test data
currently available for these approved systems. For new (proposed) systems,
design requirements will generally not exist, and would need to be
developed.
These examples illustrate the application of the Methodology, and one
example illustrates the process of iteratively modifying the system design
requirements so that the proposed structural system meets the prescribed
collapse performance objectives of the Methodology. These examples also
demonstrate consistency between the acceptance criteria of the Methodology
and the inherent safety against collapse intended by current seismic codes.
These examples were completed in parallel with the development of the
Methodology. As such, they are consistent with the procedures contained
herein, but are not necessarily in complete compliance with every
requirement. Where they occur, deviations are noted, and explanations are
provided as to how the example could be completed in accordance with the
Methodology.
In contrast to the process for developing structural system archetypes in
Chapter 4, these examples begin with the development of a representative
index archetype model, which is used as a basis for generating a set of
archetypical configurations that do not attempt to rigorously interrogate the
limits of what is permitted within the governing design requirements. As
such, they do not necessarily include archetypes that bound the full extent of
9-2 9: Example Applications FEMA P695
the design space. Additionally, the archetype designs used in these examples
do not account for potential overstrength caused by wind load requirements.
While representative examples from current code-approved systems have
been selected for developmental studies, the results are not intended to
propose specific changes to current building code requirements for any
currently approved system.
9.2 Example Application - Reinforced Concrete
Special Moment Frame System
9.2.1 Introduction
In this example, a reinforced concrete special moment frame system, as
defined by ACI 318-05 Building Code Requirements for Structural Concrete
(ACI, 2005), is considered as if it were a new system proposed for inclusion
in ASCE/SEI 7-05.
9.2.2 Overview and Approach
In this example, detailing requirements of ACI 318-05 are assumed to be
given. The system design requirements of ASCE/SEI 7-05 are used as the
framework, and seismic performance factors (SPFs) are determined by
iteration until the acceptance criteria of the Methodology are met. Seismic
performance factors under consideration in this example include the R factor,
Cd factor, and O0 factor.
All pertinent design requirements of ASCE/SEI 7-05, including drift limits
and minimum base shear requirements are assumed to apply initially. In the
Methodology, the user has full flexibility to define and modify any aspect of
the proposed system design requirements, as long as modifications are tested
within the index archetype configurations. This includes the R factor,
stiffness requirements, detailing requirements, capacity design requirements,
minimum base shear requirements, height limits, drift limits, and any other
requirements that control the design of the structural system.
The iterative assessment process begins with initial assumptions of R = 8,
Cd = 5.51, story drift limits of 2%, and minimum design base shear
requirements consistent with ASCE/SEI 7-05. Overstrength, O0, is not
assumed initially, but is determined from the computed lateral overstrength
1 Cd = 5.5 is used in this example, based on the value specified for reinforced
concrete special moment frames in ASCE/SEI 7-05. In actual applications of the
Methodology, Cd = R should be used unless Cd < R can be substantiated in
accordance with the criteria of Section 7.7.
FEMA P695 9: Example Applications 9-3
of the archetype designs. A set of structural system archetypes are developed
for reinforced concrete special moment frame buildings, nonlinear models
are developed to simulate structural collapse, models are analyzed to predict
the collapse capacities of each design, and the adjusted collapse margin ratios
are evaluated and compared to acceptance criteria.
After completing an assessment using the initial set of SPFs, certain
archetypes (taller building configurations) did not meet the acceptance
criteria, and were found to have inadequate collapse safety. Aspects of the
structural design requirements were modified, and the system was reassessed
and found to pass the acceptance criteria.
This example has been adapted from collaborative research on the
development of structural archetypes for reinforced concrete special moment
frames, calibration of nonlinear element models for collapse simulation,
simulation of structural response to collapse, spectral shape considerations,
and treatment of uncertainties (Haselton and Deierlein, 2007).
9.2.3 Structural System Information
Design Requirements
This example utilizes ACI 318-05 design requirements in place of the
requirements that would be developed for a newly proposed system. For the
purpose of assessing uncertainty according to Section 3.4, ACI 318-05
design requirements are categorized as (A) Superior since they represent
many years of development and include lessons learned from a number of
major earthquakes.
In the process of completing the assessment of the class of reinforced
concrete special moment frame buildings, it was found that often seemingly
subtle design requirements have important effects on the design and resulting
structural performance; the various design requirements often interact and
affect the design and performance differently than one might expect.
Therefore, for newly proposed systems, it is important that the set of design
requirements is well developed and clearly specified, and that the
requirements are applied in their totality when designing the archetype
structures.
Test Data
This example assessment relies on existing published test data in place of test
data that would be developed for a newly proposed system. Specifically,
column tests reported in Pacific Earthquake Engineering Research Center’s
Structural Performance Database developed by Berry, Parrish, and Eberhard
9-4 9: Example Applications FEMA P695
(PEER, 2006b; Berry et al., 2004) are utilized. To develop element models,
the data are utilized from cyclic and monotonic tests of 255 rectangular
columns failing in flexure and flexure-shear.
The quality of the test data is an important consideration when quantifying
the uncertainty in the overall collapse assessment process. The test data used
in this example cover a wide range of column design configurations and
contain both monotonic and cyclic loading protocols. Even so, many of the
loading protocols are not continued to deformations large enough to observe
loss of strength, and it is difficult to use such data to calibrate models for
structural collapse assessment. These test data also do not include beam
elements with attached slabs. Additionally, data include no systematic test
series that both (1) subject similar specimens to different loading protocols
(e.g., monotonic and cyclic) and (2) continue the loading to deformations
large enough for the capping behavior to be observed. Lastly, only column
element tests were utilized when used to calibrate the element model, while
sub-assemblage tests and full-scale tests were not used. Based on the
guidelines of Section 3.6 and considering the above observations, this test
data set is categorized as (B) Good.
9.2.4 Identification of Reinforced Concrete Special Moment
Frame Archetype Configurations
Figure 9-1 shows the two-dimensional three-bay multi-story frame that is
considered an appropriate index archetype model for reinforced concrete
frame buildings. This archetype model includes joint panel elements, beam
and column elements, elastic foundation springs, and a leaning column to
account for the P-delta effect from loads on the gravity system. This twodimensional
model, not accounting for torsional effects, is considered
acceptable because most reinforced concrete special moment frame buildings
that are regular in plan will not be highly sensitive to torsional effects, and
the goal is to verify the performance of a full class of buildings, rather than
one specific building with a unique torsional issue. This index archetype
model was used as a basis for postulating the index archetype configurations
covering the archetype design space for reinforced concrete special moment
frame buildings. Appendix C provides more background on the development
of this archetype configuration and model.
FEMA P695 9: Example Applications 9-5
Figure 9-1 Index archetype model for moment frame buildings (after
Haselton and Deierlein, 2007, Chapter 6).
Using the above index archetype model, a set of structural archetype designs
are developed to represent the archetype design space, following the design
configuration and performance group requirements of Chapter 4. Chapter 4
specifies consideration of up to 16 archetype performance groups for a
structural system whose performance can be adequately evaluated using two
basic configurations, as shown in Table 9-1. Two basic configurations are
considered in the example evaluation of the reinforced concrete special
moment frame system, archetypes with 20-foot and 30-foot bay widths,
respectively. While two configurations are sufficient to illustrate the
Methodology, additional configurations would likely be required to fully
investigate performance of the reinforced concrete special moment frame
system.
High and low gravity load intensities are represented by space frame and
perimeter frame systems, respectively. For the archetypes used in this
example, the ratio of gravity load tributary area to lateral load tributary area
is typically six times larger for space frame systems.
9-6 9: Example Applications FEMA P695
Table 9-1 Performance Groups for Evaluation of Reinforced Concrete
Special Moment Frame Archetypes
Performance Group Summary
Group No.
Grouping Criteria
Number of
Basic Archetypes
Config.
Design Load Level Period
Gravity Seismic Domain
PG-1
20-foot
Bay Width
High
(Space
Frame)
SDC Dmax
Short 2+11
PG-2 Long 4
PG-3
SDC Dmin
Short 0
PG-4 Long 12
PG-5
Low
(Perimeter
Frame)
SDC Dmax
Short 2+11
PG-6 Long 4
PG-7
SDC Dmin
Short 0
PG-8 Long 32
PG-9
30-foot
Bay Width
High
(Space
Frame)
SDC Dmax
Short 0
PG-10 Long 13
PG-11
SDC Dmin
Short
0
PG-12 Long
PG-13
Low
(Perimeter
Frame)
SDC Dmax
Short 0
PG-14 Long 13
PG-15
SDC Dmin
Short
0
PG-16 Long
1. Example includes only two archetypes for each short-period performance group
(PG-1 and PG-5); full implementation of the Methodology requires a total of 3
(2+1) archetypes in each performance group.
2. Example evaluates a selected number of low seismic (SDC Dmin) archetypes to
determine that high seismic (SDC Dmax) archetypes control the R factor.
3. Example evaluates two 30-foot bay width archetypes to determine that 30-foot
bay width archetypes do not control performance.
Archetypes are designed within a range of building heights (six heights
between 1-story and 20-stories, as expected for reinforced concrete frame
buildings with no walls), and then separated into the short-period and longperiod
performance groups. At a minimum, three buildings are needed for
each performance group, so if 16 complete performance groups were
evaluated, then at least 48 archetypes would need to be designed and
assessed. Instead of designing and assessing all 48 buildings, initial pilot
studies were used to find the more critical design cases: in this case, highseismic
(SDC D) designs with 20-foot bay spacing. By utilizing these pilot
studies and then focusing on the critical design cases, it was possible to
reduce the number of required archetypes from 48 to 18 (for a complete
FEMA P695 9: Example Applications 9-7
exercise of the Methodology, a few additional archetypes would be needed,
as described below). Appendix C provides a more detailed discussion on the
development of structural system archetypes for the reinforced concrete
special moment frame system.
Table 9-1 shows archetypes used in assessing the reinforced concrete special
moment frame structural system, and provides the rationale for why each
archetype was chosen. As indicated in the table, full implementation of the
Methodology would require two additional three-story buildings to be added
to PG-1 and PG-5 to meet the required minimum three archetypes per group.
The approach utilized here, focusing on the critical design cases, is not
required in the Methodology. The benefit of this approach is that it can
significantly reduce the number of required archetype designs (this example
needed only 20, instead of 48 archetypes) and allow a wider range of design
conditions to be considered in the assessment. When this approach is
utilized, the peer review panel should closely review choice of critical design
cases.
Table 9-2 shows the properties for each of the archetype designs used in this
evaluation. Seismic demands are represented by the maximum and minimum
seismic criteria of Seismic Design Category (SDC) D, in accordance with
Section 5.2.1: SDS = 1.0 g and SD1 = 0.60 g for SDC Dmax and SDS = 0.50 g
and SD1 = 0.20 g for SDC Dmin
2. The space frame buildings are denoted by
“S” and the perimeter frame buildings are denoted by “P.” The computed
value of the fundamental period, T, in Table 9-2, is based on Equation 5-5.
Each archetype was fully designed in accordance with the governing design
requirements. Additional information on index archetype designs is provided
in Appendix C. Figure 9-2 shows example design documentation for the
four-story, SDC Dmax design with 30-foot bay width of Archetype ID 1010.
This archetype will be used throughout this illustrative assessment, to clearly
show how the Methodology should be applied in assessing each archetype
design.
2 In this example, archetypes designed for low seismic (SDC Dmin) loads, assumed
SD1 = 0.167 g based on interim criteria, which differs slightly from the SD1 = 0.20 g
required by the Methodology.
9-8 9: Example Applications FEMA P695
Table 9-2 Reinforced Concrete Special Moment Frame Archetype
Structural Design Properties
Archetype
ID
No. of
Stories
Key Archetype Design Parameters
Framing
(Gravity
Loads)
Seismic Design Criteria SMT(T)
[g]
SDC R T [sec] T1 [sec] V/W [g]
Performance Group No. PG-5 (Short Period, 20' Bay Width Configuration)
2069 1 P Dmax 8 0.26 0.71 0.125 1.50
2064 2 P Dmax 8 0.45 0.66 0.125 1.50
-- 3 P Dmax 8 0.63 -- 0.119 1.43
Performance Group No. PG-6 (Long Period, 20' Bay Width Configuration)
1003 4 P Dmax 8 0.81 1.12 0.092 1.11
1011 8 P Dmax 8 1.49 1.71 0.050 0.60
5013 12 P Dmax 8 2.13 2.01 0.035 0.42
5020 20 P Dmax 8 3.36 2.63 0.022 0.27
Performance Group No. PG-1 (Short Period, 20' Bay Width Configuration)
2061 1 S Dmax 8 0.26 0.42 0.125 1.50
1001 2 S Dmax 8 0.45 0.63 0.125 1.50
-- 3 S Dmax 8 0.63 -- 0.119 1.43
Performance Group No. PG-3 (Long Period, 20' Bay Width Configuration)
1008 4 S Dmax 8 0.81 0.94 0.092 1.11
1012 8 S Dmax 8 1.49 1.80 0.050 0.60
5014 12 S Dmax 8 2.13 2.14 0.035 0.42
5021 20 S Dmax 8 3.36 2.36 0.022 0.27
Selected Archetypes - Performance Group Nos. PG-4 and PG-8 (20' Bay Width)
6011 8 P Dmin 8 1.60 3.00 0.013 0.15
6013 12 P Dmin 8 2.28 3.35 0.010 0.10
6020 20 P Dmin 8 3.60 4.08 0.010 0.065
6021 20 S Dmin 8 3.60 4.03 0.010 0.065
Selected Archetypes - Performance Group Nos. PG-10 and PG-14 (30' Bay Width)
1009 4 P-30 Dmax 8 1.03 1.16 0.092 1.03
1010 4 S-30 Dmax 8 1.03 0.86 0.092 1.03
FEMA P695 9: Example Applications 9-9
Floor 5
Floor 4
Floor 3
feet
Floor 2
feet
feet
Design base shear = g, k
f'c beams = ksi f'c,cols,upper = ksi
fy,rebar,nom. = ksi f'c,cols,lower = ksi
0.0065
4.0 4.0 4.0 4.0
0.0065 0.0065 0.0065
0.013 0.016 0.016 0.013
30 30 30 30
. =
.' =
. =
.' =
.sh =
s (in) =
0.013
0.0065
4.0
30 30 30 30
5.0
4.0 4.0
0.0025
0.0025
0.0065
30
4.0
30
0.0054
0.0110
30
30
0.0054
0.0110
0.016
0.0065
30
30
0.0054
0.0110
0.0025
5.0
0.0065
5.0
30 30
0.013 0.016
30
s (in) =
4.0
30
30
24
h (in) =
.tot =
30
13
15
30
30
4.0
0.0023
5.5
30 30
.sh =
.sh =
s (in) =
h (in) =
b (in) =
b (in) =
.tot =
.sh =
s (in) =
s (in) =
h (in) =
b (in) =
.tot =
h (in) =
b (in) =
.tot =
.sh =
24
30
0.0072
0.011
0.0116
0.0032
5.0
0.0032
5.0
24
30
30
30
0.0072
0.0116
0.011
0.0065
4.0
24
30
30
0.011
0.0065
4.0
24
5.0
30
0.0072
0.0116
0.0032
30
30
0.011
0.0065
30 30 30
4.0
0.0069
0.0132
0.0034
5.0
0.0065
30
30 30 30 30
0.011 0.016 0.016 0.011
0.0065 0.0065 0.0065 0.0065
0.0023
5.5
0.005
0.0100
4.0 4.0
30
30
0.005
0.0100
30
h (in) =
b (in) =
. =
.' =
.sh =
s (in) =
h (in) =
b (in) =
b (in) =
0.0069
0.0132
0.0034
30
30
0.005
0.0100
0.0023
5.0
30
4.0
30
5.5
0.0069
0.0132
.sh =
s (in) =
Story 1 Story 2
h (in) =
Story 3 Story 4
0.0034
5.0
30
30
24
30
60
0.092 193
5.0
5.0 5.0
(a)
Floor 4
Floor 3
Floor 2 1.32
1.16
0.08
1.21
0.50
1.99
1.39
1.13
1.24
0.52
0.52
1.16
0.05
1.14
1.24
1.64
0.11
0.74
1.58
1.27
1.31
1.38
1.17
1.17
1.6
fMn/Mu =
(fMn/Mu)neg =
(fMn/Mu)pos =
Mn,pos/Mn,neg =
1.30
fVn/Vmpr = 1.13
P/Agf'c = 0.08
fMn/Mu =
1.32
1.15
fVn/Vmpr = 0.11
1.14
1.21
Mn,pos/Mn,neg =
(fMn/Mu)neg =
(fMn/Mu)pos =
0.22
(fMn/Mu)neg =
fVn/Vmpr =
P/Agf'c =
1.64
fVn/Vmpr =
1.03
2.01
0.22
(fMn/Mu)pos =
2.02
fVn/Vmpr =
1.21
Mn,pos/Mn,neg =
1.15
3.06 3.06
1.03
1.21
1.17
1.27
1.31
1.30
1.22
1.49
0.50
1.16
1.29
fMn/Mu =
fVn/Vmpr =
0.05
1.37
1.17
P/Agf'c =
1.38
fMn/Mu =
fVn/Vmpr =
P/Agf'c =
SCWB = 0.74
Joint FVn/Vu = 1.58
1.16
0.11
1.16
1.13
0.11
1.18
1.43
0.52
1.16
1.29
1.49
1.59 2.61 2.61 1.59
1.49
3.16 3.18
1.99 1.31 1.31
1.6
1.17
1.17 0.03
1.27
1.14
1.17
0.54
1.17
0.54
0.06
1.27
1.02
0.63
1.16
1.21
1.16
1.28
1.92
2.78
0.54
1.30
0.54
1.30
1.16
1.17
0.63
1.16
1.22
1.16
1.16
1.17
0.63
0.16
0.03
1.14
1.17
1.15
0.50
1.16
Story 1 Story 2
1.16
0.16
Story 3 Story 4
1.02
0.06
2.80
0.54
1.92
Design
Drifts:
0.8%
1.0%
1.2%
1.4%
(b)
Figure 9-2 Design documentation for a four-story space frame archetype
with 30-foot bay spacing (Archetype ID 1010) (after Haselton
and Deierlein 2007, Chapter 6). The notation used in this
figure is defined in the list of symbols at the end of the
document.
9-10 9: Example Applications FEMA P695
9.2.5 Nonlinear Model Development
This section summarizes explicit modeling of structural collapse. A more
complete discussion on nonlinear modeling of reinforced concrete special
moment frame systems is provided in Appendices D and E.
Index Archetype Models
System-level modeling uses the three-bay multi-story frame configuration
shown in Figure 9-1. This model consists of elastic joint elements, plastic
hinge reinforced concrete beam and column elements, a leaning column to
account for P-delta effects, and elastic foundation springs to account for
foundation and soil flexibility.
Nonlinear Beam-Column Element Models
Even though many reinforced concrete element models exist, most cannot be
used to simulate structural collapse. Recent research by Ibarra, Medina, and
Krawinkler (2005) has resulted in an element model that is capable of
capturing the severe deterioration that precipitates sideway collapse
(Appendix E). Figure 9-3 shows the tri-linear monotonic backbone curve
and associated hysteretic behavior of this model. This model includes
important aspects, such as the “capping point,” where monotonic strength
loss begins, and the post-capping negative stiffness. These features enable
modeling of the strain-softening behavior associated with concrete crushing,
rebar buckling and fracture, or bond failure. In general, accurate simulation
of sidesway structural collapse requires modeling this post-capping behavior.
Researchers have also used a variety of other methods to simulate cyclic
response of reinforced concrete beam-columns, including creating fiber
models that can capture cracking behavior and the spread of plasticity
throughout the element (e.g., Filippou, 1999). The decision to use a lumpedplasticity
approach was based on the observation that currently available
fiber element models are not capable of simulating the strain-softening
associated with rebar buckling, and thus cannot reliably simulate collapse of
flexurally dominated reinforced concrete frames. Research is ongoing, and
this modeling limitation may be overcome in the near future. In future
applications of this Methodology, the choice of element model should be
carefully evaluated for any given structural system, with consideration of
available simulation technologies and their capabilities for simulating
structural collapse.
FEMA P695 9: Example Applications 9-11
0
0.2
0.4
0.6
0.8
1
1.2
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Chord Rotation (radians)
Normalized Moment (M/My)
Kc
.cap
p
l
Ks
My
Ke
.y .cap
Mc
.pc
.cap
tot
-1.5
-1
-0.5
0
0.5
1
1.5
-8 -6 -4 -2 0 2 4 6 8
Chord Rotation (radians)
Normalized Moment (M/My)
Non-Deteriorated
Backbone
Figure 9-3 Monotonic and cyclic behavior of component model used in
this study (after Haselton and Deierlein 2007, Chapter 4).
Figure 9-4 shows an example of the experimental and calibrated response of
a reinforced concrete column specimen. Appendix E presents a detailed
discussion of how this element model was calibrated using cyclic and
monotonic tests of 255 rectangular reinforced concrete elements. Such
calibration needs to be completed carefully, in order to avoid errors in the
collapse capacity prediction. Often various deterioration modes are
improperly mixed together in the calibration process, such as cyclic strength
deterioration versus in-cycle strength loss, and this can lead to substantial
errors in collapse predictions.
The calibration results for the 255 tests (Appendix E) were subsequently used
to create empirical equations to predict the element model parameters (as
shown in Figure 9-3), based on the element design information such as axial
load ratio and confinement ratio. These equations were used to predict the
modeling parameters for the archetype designs used in this example. For
illustration, Figure 9-5 shows the predicted modeling parameters for each
element of Archetype ID 1010 which is a four-story SDC Dmax design with
30-foot bay width.
9-12 9: Example Applications FEMA P695
-0.1 -0.05 0 0.05 0.1
-300
-200
-100
0
100
200
300
Shear Force (kN)
Column Drift (displacement/height)
Experimental Results
Model Prediction
Figure 9-4 Illustration of experimental and calibrated element response
(Saatciolgu and Grira, specimen BG-6) (after Haselton et al.,
2007). The solid black line shows the calibrated monotonic
backbone. The notation is defined in the list of symbols at the
end of this document.
Floor 5
Floor 4
Floor 3
Floor 2
Model periods (sec): T1 = T2= T3 =
fy,rebar,expected=
100
1.354E+08
0.080
108
-0.068
0.35
0.0453
-0.064
0.100
(Tcap,pl)neg (rad) =
(Tcap,pl)neg (rad) =
Tpc (rad) =
. =
My,pos,exp (k-in) =
100
0.15
EIstf, w/ Slab (k-in2/rad) =
0.100
0.063 0.056 0.056
(Tcap,pl)pos (rad) =
18812
My,n,slab,exp (k-in) =
EIstf/EIg =
EIstf, w/ Slab (k-in2/rad) =
(Tcap,pl)pos (rad) =
EIstf/EIg =
0.35
2.591E+08
-0.059
0.086
0.100
16800 16800 12270
-12282
0.065
0.35
0.045
-0.068
1.20
Story 1 Story 2 Story 3 Story 4
0.35
1.20
0.067
0.100
100
-16795
18812
0.35
0.038
-0.059
0.086
-16795
0.35
0.038
2.591E+08
0.038
-0.059
0.086
0.12
110
5842
112
0.060
7531
105
15072
0.100
0.066
1.354E+08
0.100
100
0.067
1.20 1.20
2.591E+08
0.035
-0.054
0.08
116
0.04
0.100
15072
-12282
105
7531
0.12
0.100
0.060
6106
-11024
0.35
0.0453
-0.064
0.100
100
5842
110
9770
0.35
13243
0.100
100
0.35
1.19
0.08
0.35
1.19
0.35
1.21
8583
0.069
119
0.02
0.35
0.04
0.35
116
1.354E+08
9602
0.35
0.045
116
0.04
0.35
0.066
1.20
2.591E+08
-16795
112
0.35
. =
0.100
0.063
113
0.06
1.20
0.100
0.15
0.35
1.19
110
0.08
0.35
1.19
My,n,slab,exp (k-in) =
Tpc (rad) =
My,pos,exp (k-in) =
0.08
Tpc (rad) =
. = 110
My,exp (k-in) =
EIstf/EIg =
Mc/My =
0.100
100
Tcap,pl (rad) =
Tpc (rad) =
. =
(P/Agf'c)exp =
Tcap,pl (rad) = 0.066
0.35
1.20
7531 113
0.06
0.100
13243
EIstf/EIg = 0.35
Mc/My = 1.20
. =
My,exp (k-in) =
My,exp (k-in) = 12270
0.35
Mc/My = 1.20
Tcap,pl (rad) = 0.066
EIstf/EIg =
Tpc (rad) =
. =
(Tcap,pl)pos (rad) =
EIstf, w/ Slab (k-in2/rad) =
9602
0.02
1.354E+08
0.069
5842
-12282
0.35
0.35
8583
119
0.100
1.21
My,exp (k-in) =
Mc/My =
Tcap,pl (rad) =
Tpc (rad) =
EIstf/EIg =
1.354E+08
6106
-11024
0.0453
0.35
0.100
0.35
0.035
-0.054
0.080
2.591E+08
108
108
7036 6106
-11024
0.35
0.035
-15481
2.591E+08
9770
0.35
1.20
0.100
(P/Agf'c)exp =
0.100
0.35
100
1.354E+08
-0.064
Tpc (rad) =
7036
-15481
116
(P/Agf'c)exp =
(P/Agf'c)exp =
7036
-0.054
0.04
-15481
0.065
. =
0.045
-0.068
0.100
0.100 0.100 0.100
0.080
112
My,pos,exp (k-in) =
(Tcap,pl)neg (rad) =
My,n,slab,exp (k-in) =
EIstf/EIg =
67 ksi
1.34
0.86 0.27 0.15
Mass tributary to one frame for lateral load (each floor) (k-s-s/in):
Figure 9-5 Structural modeling documentation for a four-story space frame
archetype with 30-foot bay spacing (Archetype ID 1010) (after
Haselton and Deierlein, 2007, Chapter 6).
Uncertainty due to Model Quality
For the purpose of assessing uncertainty, this modeling is rated as (B) Good,
according to the guidelines of Section 5.7. Reinforced concrete special
moment frame buildings are controlled by many detailing and capacity
FEMA P695 9: Example Applications 9-13
design requirements, which limit possible failure modes. The primary
expected failure mode is flexural hinging leading to sidesway collapse, which
the modeling approach can simulate reasonably well by capturing post-peak
degrading response under both monotonic and cyclic loading. The modeling
approach is able to directly simulate structural response up to collapse by
simulating all expected modes of damage that could lead to collapse, and is
well calibrated to large amounts of data. Even so, this model is not given the
(A) Superior rating because there is still room for improvement in the model:
the model was calibrated using column data and is not well-calibrated to
beam-slabs, and the model does not capture axial-flexural interaction in
columns. Additionally, for a complete assessment, the archetype design
space would likely need to be expanded to include a wider range of basic
configurations beyond 20-foot and 30-foot bay spacings.
9.2.6 Nonlinear Structural Analysis
The structural analysis software selected for completing this example is the
Open Systems for Earthquake Engineering Simulation (OpenSees, 2006),
which was developed by the Pacific Earthquake Engineering Research
(PEER) Center. This software includes the modeling aspects required for
collapse simulation of reinforced concrete special moment frame buildings,
such as the Ibarra et al. element model, joint models, a large deformation
geometric transformation, and several numerical algorithms for solving the
systems of equations associated with nonlinear dynamic and static analyses.
Nonlinear static (pushover) analysis is performed in accordance with Section
6.33, in order to compute the system overstrength factor, O0, and periodbased
ductility, .T, and to help verify the structural model. Figure 9-6 shows
an example of the pushover curve and story drift distributions for Archetype
ID 1010.
For the reinforced concrete special moment frame archetype, yielding occurs
at story drift ratio 0.005 and the effective yield roof displacement, .y,eff , is
computed according to Equation 6-7 as 0.0042hr. A maximum strength,
Vmax, of 635 kips is reached and followed by the onset of negative stiffness
which occurs at a story drift ratio of 0.040 and a roof displacement of
0.035hr. This is then followed with 20% strength loss at .u
= 0.056hr. Using
these values, the overstrength factor of Archetype ID 1010 can be computed
as O = 635k/193k = 3.3, and the period-based ductility, which will be used
3 In this example, pushover analysis is based on the lateral load pattern prescribed by
Equation 12.8-13 of ASCE/SEI 7-05.
9-14 9: Example Applications FEMA P695
later to adjust the CMR according to Section 7.2, can be computed as µT = .u
/.y,eff = 0.056hr/0.0042hr = 13.2.
To compute the collapse capacity for each archetype design, the incremental
dynamic analysis (IDA) approach is used (Section 6.4), with the Far-Field
ground motion set and ground motion scaling method presented in Section
6.2.3 (and Appendix A). Note that Section 6.4 does not require a full IDA
(as is shown in Figure 9-7) but only a simplified version, which has the goal
of quantifying the median collapse capacity of each archetype model.
0 0.02 0.04 0.06
0
100
200
300
400
500
600
700
Base Shear (kips)
Roof Drift Ratio
0 0.02 0.04 0.06 0.08 0.1
1
2
3
4
5
Floor Number
Interstory Drift Ratio
Figure 9-6 (a) Monotonic static pushover, and (b) peak story drift ratios at
three deformation levels during pushover for Archetype ID
1010. The pushover is based on the building code specified
lateral load distribution (ASCE, 2005) (after Haselton and
Deierlein, 2007, Chapter 6).
Figure 9-7 illustrates how the IDA method is used to compute the collapse
capacity of Archetype ID 1010. For each of the 44 ground motions of the
Far-Field Set the spectral acceleration at collapse (SCT) is computed. Next
the median collapse level ( CT Sˆ ) is computed to be 2.58 g. The collapse
margin ratio (CMR), defined as the ratio of CT Sˆ to the Maximum Considered
Earthquake (MCE) ground motion demand (SMT), is 2.50 for Archetype ID
(a)
(b)
V = 193k
1: RDR of 0.004
2: RDR of 0.035
3: RDR of 0.070
1 2
3
1 2 3
Vmax = 635k
0.8Vmax = 510k
.y,eff = 0.0042hr .u
= 0.056hr
Story Drift Ratio
O = 635k / 193k = 3.3
FEMA P695 9: Example Applications 9-15
1010. Although a full IDA was utilized in this example, it is not required to
quantify CMR, as discussed in Section 6.4.2.
In this example, it is assumed that reinforced concrete special moment frame
buildings collapse in a sideway mechanism, which can be directly simulated
using the structural analysis model. This assertion is made due to the many
detailing, continuity, and capacity design provisions preventing other
collapse modes (Appendix D). For structural systems where some collapse
modes are not simulated by the structural model, these additional modes
must be accounted for using component limit state checks for non-simulated
collapse modes (Section 5.5).
0 0.05 0.1 0.15
0
1
2
3
4
5
6
7
ST(T=1.03s) [g]
Maximum Interstory Drift Ratio
Figure 9-7 Incremental dynamic analysis to collapse, showing the
Maximum Considered Earthquake ground motion intensity
(SMT), median collapse capacity intensity ( ˆ
CT S ), and collapse
margin ratio (CMR) for Archetype ID 1010.
Table 9-3 summarizes the results of pushover and IDA analyses. These IDA
results show that the average CMR is 1.34 and 1.26 for perimeter frame
performance groups and is 2.01 and 1.56 for the space frame performance
groups. These values have not yet been adjusted for the beneficial effects of
spectral shape (Section 7.2). Allowable CMR values and acceptance criteria
are discussed later.
In addition, these results verify that the buildings designed in low-seismic
regions have higher CMRs (i.e., lower collapse risk) than those designed in
CMR = 2.58/1.04 = 2.5
CT Sˆ (T=1.03s) = 2.58g
SMT(T=1.03s) = 1.04g
x 2.5
Maximum Story Drift Ratio
9-16 9: Example Applications FEMA P695
high-seismic regions. Compared to a building with 20-foot bay width,
buildings with 30-foot bay width tend to have slightly higher CMR, due to
gravity load effects causing some increases in overstrength, and due to joint
shear provisions making columns slightly larger and more ductile.
Therefore, the remaining assessment focuses on 12 more critical archetype
buildings which are designed for high-seismic sites (SDC Dmax) and with 20-
foot bay width. This approach substantially reduces the number of required
structural designs and analyses.
Table 9-3 Summary of Collapse Results for Reinforced Concrete Special
Moment Frame Archetype Designs
Archetype
ID
Design Configuration Pushover and IDA Results
No. of
Stories
Framing
(Gravity
Loads)
Seismic
SDC Static . SMT [T]
(g)
SCT [T]
(g)
CMR
Performance Group No. PG-5 (Short Period, 20' Bay Width Configuration)
2069 1 P Dmax 1.6 1.5 1.77 1.18
2064 2 P Dmax 1.8 1.5 2.25 1.50
-- 3 P Dmax -- -- -- --
Performance Group No. PG-6 (Long Period, 20' Bay Width Configuration)
1003 4 P Dmax 1.6 1.11 1.79 1.61
1011 8 P Dmax 1.6 0.6 0.76 1.25
5013 12 P Dmax 1.7 0.42 0.51 1.22
5020 20 P Dmax 2.6 0.27 0.22 0.82
Mean of Performance Group: 1.9 NA NA 1.23
Performance Group No. PG-1 (Short Period, 20' Bay Width Configuration)
2061 1 S Dmax 4.0 1.5 2.94 1.96
1001 2 S Dmax 3.5 1.5 3.09 2.06
-- 3 S Dmax -- -- -- --
Performance Group No. PG-3 (Long Period, 20' Bay Width Configuration)
1008 4 S Dmax 2.7 1.11 1.97 1.78
1012 8 S Dmax 2.3 0.60 0.98 1.63
5014 12 S Dmax 2.8 0.42 0.67 1.59
5021 20 S Dmax 3.5 0.27 0.34 1.25
Mean of Performance Group: 2.8 NA NA 1.56
Selected Archetypes - Performance Group Nos. PG-4 and PG-8 (20' Bay Width)
6011 8 P Dmin 1.8 0.15 0.32 2.12
6013 12 P Dmin 1.8 0.1 0.21 2.00
6020 20 P Dmin 1.8 0.07 0.11 1.73
6021 20 S Dmin 3.4 0.07 0.24 3.70
Selected Archetypes - Performance Group Nos. PG-10 and PG-14 (30' Bay Width)
1009 4 P-30 Dmax 1.6 1.04 2.05 1.98
1010 4 S-30 Dmax 3.3 1.04 2.58 2.50
FEMA P695 9: Example Applications 9-17
9.2.7 Performance Evaluation
The previous section discussed how to simulate structural collapse, compute
median collapse level, ˆ
CT S , and compute the collapse margin ratio, CMR.
However, CMR does not account for the unique spectral shape of rare ground
motions. Chapter 7 discusses spectral shape and how it affects the predicted
collapse capacity, and provides simplified spectral shape factors, SSFs, that
are used to adjust the median collapse level , ˆ
CT S , to account for spectral
shape effects. Table 9-4 and Table 9-5 (taken from Chapter 7) show these
factors for Seismic Design Categories B, C, and Dmin and SDC Dmax,
respectively. The values in these tables depend on the fundamental period, T,
period-based ductility, .T, and the seismic design category. The tables show
that the SSF values range from 1.0 to 1.37 for Seismic Design Categories B,
C, and Dmin and from 1.00 to 1.61 for SDC Dmax.
The adjusted collapse margin ratio, ACMR, is computed by multiplying SSF
(from Table 9-4 or Table 9-5) and CMR (from Table 9-3). Later, Table 9-8
will show this margin adjustment for the reinforced concrete special moment
frame archetypes, and the resulting ACMR values.
Table 9-4 Spectral Shape Factor (SSF) for Archetypes Designed for Seismic
Design Categories B, C, or Dmin Seismic Criteria (from Table 7-
1a)
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.02 1.04 1.06 1.08 1.09 1.12 1.14
0.6 1.00 1.02 1.05 1.07 1.09 1.11 1.13 1.16
0.7 1.00 1.03 1.06 1.08 1.10 1.12 1.15 1.18
0.8 1.00 1.03 1.06 1.08 1.11 1.14 1.17 1.20
0.9 1.00 1.03 1.07 1.09 1.13 1.15 1.19 1.22
1.0 1.00 1.04 1.08 1.10 1.14 1.17 1.21 1.25
1.1 1.00 1.04 1.08 1.11 1.15 1.18 1.23 1.27
1.2 1.00 1.04 1.09 1.12 1.17 1.20 1.25 1.30
1.3 1.00 1.05 1.10 1.13 1.18 1.22 1.27 1.32
1.4 1.00 1.05 1.10 1.14 1.19 1.23 1.30 1.35
. 1.5 1.00 1.05 1.11 1.15 1.21 1.25 1.32 1.37
9-18 9: Example Applications FEMA P695
Table 9-5 Spectral Shape Factor (SSF) for Archetypes Designed using
Seismic Design Category Dmax Seismic Criteria (from Table 7-1b)
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.05 1.1 1.13 1.18 1.22 1.28 1.33
0.6 1.00 1.05 1.11 1.14 1.2 1.24 1.3 1.36
0.7 1.00 1.06 1.11 1.15 1.21 1.25 1.32 1.38
0.8 1.00 1.06 1.12 1.16 1.22 1.27 1.35 1.41
0.9 1.00 1.06 1.13 1.17 1.24 1.29 1.37 1.44
1.0 1.00 1.07 1.13 1.18 1.25 1.31 1.39 1.46
1.1 1.00 1.07 1.14 1.19 1.27 1.32 1.41 1.49
1.2 1.00 1.07 1.15 1.2 1.28 1.34 1.44 1.52
1.3 1.00 1.08 1.16 1.21 1.29 1.36 1.46 1.55
1.4 1.00 1.08 1.16 1.22 1.31 1.38 1.49 1.58
. 1.5 1.00 1.08 1.17 1.23 1.32 1.4 1.51 1.61
In addition to quantifying the ACMR by Equation 7-1, the composite
uncertainty, ßTOT, in collapse capacity is also needed. Table 9-6 (from Table
7-2b) shows composite uncertainties, which account for the variability
between ground motion records of a given intensity (defined as a constant
ßRTR = 0.40 for systems with µT =3), the uncertainty in the nonlinear structural
modeling ((B) Good), the quality of data used to calibrate the element models
((B) Good), and the quality of the structural system design requirements ((A)
Superior). For this example assessment, the composite uncertainty is ßTOT =
0.500 and is shown in bold.
Table 9-6 Total System Collapse Uncertainty (.TOT) for Model Quality (B)
Good and Period-Based Ductility, .T
. 3 (from Table 7-2b)
Quality of Test Data
Quality of Design Requirements
(A) Superior (B) Good (C) Fair (D) Poor
(A) Superior 0.475 0.500 0.575 0.675
(B) Good 0.500 0.525 0.600 0.700
(C) Fair 0.575 0.600 0.675 0.750
(D) Poor 0.675 0.700 0.750 0.825
The acceptable collapse margin ratio is determined from the composite
uncertainty and the acceptable conditional probability of collapse under the
MCE ground motions. Chapter 7 defines the collapse performance
objectives as: (1) a conditional collapse probability of 20% for each
archetype building, and (2) a conditional collapse probability of 10% for
each performance group. Table 9-7 (from Table 7-3) shows values of
acceptable ACMR computed assuming a lognormal distribution of collapse
capacity. Based on ßTOT = 0.500, the acceptable ACMR20% value is 1.52 for
FEMA P695 9: Example Applications 9-19
each individual archetype and the acceptable ACMR10% value is 1.90 for the
average of each performance group.
Table 9-7 Acceptable Values of Adjusted Collapse Margin Ratio (ACMR10%
and ACMR20%) (from Table 7-3)
Total System
Collapse
Uncertainty
Collapse Probability
5% 10%
(ACMR10%)
15% 20%
(ACMR20%)
25%
0.400 1.93 1.67 1.51 1.40 1.31
0.425 2.01 1.72 1.55 1.43 1.33
0.450 2.10 1.78 1.59 1.46 1.35
0.475 2.18 1.84 1.64 1.49 1.38
0.500 2.28 1.90 1.68 1.52 1.40
0.525 2.37 1.96 1.72 1.56 1.42
0.550 2.47 2.02 1.77 1.59 1.45
0.575 2.57 2.09 1.81 1.62 1.47
0.600 2.68 2.16 1.86 1.66 1.50
Table 9-8 presents the final results and acceptance criteria for each of the 18
archetype designs. This shows the collapse margin ratio (CMR) computed
directly from IDA, the SSF, and the final adjusted collapse margin ratio
(ACMR). The acceptable margins are then shown, and each archetype is
shown to either pass or fail the acceptance criteria.
The results in Table 9-8 show that high-seismic archetypes (SDC Dmax
designs) have lower ACMR values (as compared to SDC Dmin designs) and
control the collapse performance. Additionally, the 20-foot bay width
designs have slightly lower ACMR values. Therefore, this example focuses
on the performance groups containing archetypes designed for high seismic
(SDC Dmax) with 20-foot bay widths (PG-1, PG-3, PG-5, and PG-6). Based
on the results shown in Table 9-8, the perimeter frame performance groups
are more critical than the space frame performance groups, so PG-5 or PG-6
will govern the response modification factor, R.
9-20 9: Example Applications FEMA P695
Table 9-8 Summary of Final Collapse Margins and Comparison to
Acceptance Criteria for Reinforced Concrete Special Moment
Frame Archetypes
Arch.
ID
Design Configuration Computed Overstrength and Collapse
Margin Parameters
Acceptance
Check
No. of
Stories
Framing
(Gravity
Loads)
SDC Static
O CMR .T SSF ACMR Accept.
ACMR
Pass/
Fail
Performance Group No. PG-5 (Short Period, 20' Bay Width Configuration)
2069 1 P Dmax 1.6 1.18 14.0 1.33 1.57 1.52 Pass
2064 2 P Dmax 1.8 1.50 19.6 1.33 2.00 1.52 Pass
-- 3 P Dmax 1.7* -- -- -- 2.13* -- --
Mean of Performance Group: 1.7* 1.34 16.8 1.33 1.90* 1.90 Pass
Performance Group No. PG-6 (Long Period, 20' Bay Width Configuration)
1003 4 P Dmax 1.6 1.61 10.9 1.41 2.27 1.52 Pass
1011 8 P Dmax 1.6 1.25 9.8 1.61 2.01 1.52 Pass
5013 12 P Dmax 1.7 1.22 7.4 1.58 1.93 1.52 Pass
5020 20 P Dmax 2.6 0.82 4.1 1.40 1.15 1.52 Fail
Mean of Performance Group: 1.9 1.23 8.1 1.50 1.84 1.90 Fail
Performance Group No. PG-1 (Short Period, 20' Bay Width Configuration)
2061 1 S Dmax 4.0 1.96 16.1 1.33 2.61 1.52 Pass
1001 2 S Dmax 3.5 2.06 14.0 1.33 2.74 1.52 Pass
-- 3 S Dmax 3.11 -- -- -- 2.63* -- --
Mean of Performance Group: 3.51 2.01 15.0 1.33 2.66* 1.90 Pass
Performance Group No. PG-3 (Long Period, 20' Bay Width Configuration)
1008 4 S Dmax 2.7 1.78 11.3 1.41 2.51 1.52 Pass
1012 8 S Dmax 2.3 1.63 7.5 1.58 2.58 1.52 Pass
5014 12 S Dmax 2.8 1.59 8.6 1.61 2.56 1.52 Pass
5021 20 S Dmax 3.5 1.25 4.4 1.42 1.78 1.52 Pass
Mean of Performance Group: 2.8 1.56 8.0 1.51 2.36 1.90 Pass
Selected Archetypes - Performance Group Nos. PG-4 and PG-8 (20' Bay Width)
6011 8 P Dmin 1.8 2.12 3.0 1.21 2.56 1.52 Pass
6013 12 P Dmin 1.8 2.00 3.7 1.24 2.47 1.52 Pass
6020 20 P Dmin 1.8 1.73 2.8 1.20 2.08 1.52 Pass
6021 20 S Dmin 3.4 3.70 3.3 1.22 4.51 1.52 Pass
Selected Archetypes - Performance Group Nos. PG-10 and PG-14 (30' Bay Width)
1009 4 P-30 Dmax 1.6 1.98 13.4 1.41 2.79 1.52 Pass
1010 4 S-30 Dmax 3.3 2.50 13.2 1.41 3.53 1.52 Pass
1. For completeness, the reinforced concrete special moment frame example assumes values of
static overstrength and ACMR of missing 3-story archetypes (based on the average of
respective 2-and 4-story values).
The overstrength factor, .O, will likely not be governed by the same
performance group that governs the response modification factor, R. Table
9-8 shows that the space frame performance groups have higher overstrength,
FEMA P695 9: Example Applications 9-21
., values than perimeter frame groups, short-period space frames have
higher overstrength values than long-period space frames, and the four-story
30-foot bay width design has a higher overstrength than 20-foot bay width
designs (e.g., Archetype ID 1010 versus 1008). This suggests that either PG-
9 or PG-11 (defined in Table 9-1) would govern the overstrength factor, .O.,
but the archetypes in these performance groups were not designed and
assessed in this example application.
The results for the two performance groups that may govern the response
modification factor (PG-5 or PG-6), show that the majority of archetype
buildings have acceptable ACMR, but a trend is evident: for space and
perimeter frame buildings taller than four-stories, the ACMR decreases
substantially with increased building height. This causes the 20-story
perimeter frame archetype to have an unacceptable ACMR and causes the
average ACMR of the long period perimeter frame performance group (PG-6)
to also be unacceptable.
As currently defined’ the “newly proposed” reinforced concrete special
moment frame system does not meet the collapse performance objectives of
this Methodology, and needs adjusted design requirements in order to meet
the acceptance criteria. One simple alternative would be to limit the
proposed system to a maximum height of 12-stories (or 160 feet), or to
require a space frame system for buildings taller than 160 feet. This solution
is not ideal, so the next section looks more closely at other possible solutions.
9.2.8 Iteration: Adjustment of Design Requirements to Meet
Performance Goals
The reinforced concrete special moment frame system did not meet the
performance criteria with the initial set of design requirements. For the
initially assessed designs, the ACI 318-05 design requirements were used
along with R = 8, Cd = 5.5, a story drift limit of 2%, and minimum design
base shear requirements of ASCE/SEI 7-05. These requirements must now
be modified in some way that will improve the reinforced concrete special
moment frame collapse performance and cause the system to meet the
performance criteria of Section 7.4.
According to Table 9-8, the adjusted collapse margin ratio (ACMR) decreases
with increasing height. Haselton and Deierlein (2007) showed that this type
of poor performance is caused by the damage localizing more for taller
moment frames since damage localization is driven primarily by higher Pdelta
effects as the building height increases. In order to ensure better
collapse performance in taller reinforced concrete frame buildings, this issue
could be addressed in various ways. More conservative strong-column-
9-22 9: Example Applications FEMA P695
weak-beam ratios, i.e., larger than 1.2, could be developed for taller
buildings, to spread more uniformly over the height of the building. Instead,
strength requirements could also be increased for taller buildings, by using a
period-dependent R factor. Krawinkler and Zareian (2007) illustrate how the
R factor would need to change, as a function of building period, in order to
create uniform collapse probabilities for moment frame buildings of varying
height (more strength required for longer period frame buildings).
In this example, another simple approach is attempted and the minimum
design base shear is increased to solve this problem. This solution is not the
most direct way to solve the specifically identified problems of damage
localization and P-delta for taller moment-resisting frame buildings, but it is
a simple solution that works. Specifically, the ASCE/SEI 7-05 minimum
base shear requirement, Cs = 0.01 (ASCE, 2006a, Equation 12.8-5) is
replaced by Equation 9.5.5.2.1-3 of ASCE 7-02 (ASCE, 2002). The ASCE
7-02 equation is shown here as Equation 9-1.
C S I s DS . 0.044 (9-1)
This change to the design requirements impacts only the design of the 12-
and 20-story archetypes in SDC Dmax and the design of the 8-, 12-, and 20-
story archetypes in SDC Dmin.
Table 9-9 shows the design information for the redesigned buildings. A
comparison to Table 9-2 shows that the design base shear coefficient (V/W)
increased from 0.022 to 0.044 for the 20-story archetype in SDC Dmax and
increased from 0.010 to 0.017 for the 20-story archetype in SDC Dmin. The
base shear coefficient also increased, to a lesser extent, for the other
buildings shown in Table 9-9. The minimum base shear requirement is
governing the design of the taller frames.
The building designs were revised and the collapse assessments were
completed for the revised designs. Table 9-10 shows the updated collapse
performance results, with the italic lines showing the designs that were
affected by the change to the minimum base shear requirement. This shows
that each archetype building meets the performance requirement of ACMR .
1.52 (i.e., 20% conditional collapse probability) and the average ACMR .
1.90 for each performance group (i.e., 10% conditional collapse probability).
This shows that after modifying the minimum design base shear requirement,
the “newly proposed” reinforced concrete special moment frame system
attains the required collapse performance and could be added as a “new
system” in the building code provisions.
FEMA P695 9: Example Applications 9-23
Table 9-9 Structural Design Properties for Reinforced Concrete Special
Moment Frame Archetypes Redesigned Considering a Minimum
Base Shear Requirement
Archetype
ID
No. of
Stories
Key Archetype Design Parameters
Framing
(Gravity
Loads)
Seismic Design Criteria SMT (T)
[g]
SDC Reff
1 T [sec] T1 [sec] V/W [g]
Re-Designed Archetypes - Performance Group No. PG-6 (Long Period, 20' Bay Width)
1013 12 P Dmax 6.4 2.13 2.01 0.044 0.42
1020 20 P Dmax 4.1 3.36 2.63 0.044 0.27
Re-Designed Archetypes - Performance Group No. PG-3 (Long Period, 20' Bay Width)
1014 12 S Dmax 6.4 2.13 2.14 0.044 0.42
1021 20 S Dmax 4.1 3.36 2.36 0.044 0.27
Re-Designed Archetypes - Performance Group No. PG-4, PG-8 (Long Period, 20' Bays)
4011 8 P Dmin 5.8 1.6 3.00 0.017 0.15
4013 12 P Dmin 6.6 2.28 3.35 0.017 0.10
4020 20 P Dmin 4.2 3.6 4.08 0.017 0.065
4021 20 S Dmin 8.0 3.6 4.03 0.017 0.065
1. Effective value of R due to limits on the seismic coefficient, Cs.
Notice that the short-period perimeter frame high-seismic performance group
(PG-5) is now the governing group that controls the value of the response
modification factor, R. It is common that short-period structures have higher
strength demands than moderate period structures, and therefore can control
the collapse performance assessment; this has been documented in many
research publications, starting with Newmark and Hall (1973).
Table 9-10 shows that building archetypes were only assessed up to a height
of 20 stories. Taller buildings were not assessed because of the limitations of
the ground motion record set, which is applicable only to buildings with
elastic fundamental periods below 4.0 seconds (Chapter 6). Even so, Table
9-10 shows a trend that the collapse safety increases with building height.
As long as this trend is observed for buildings with fundamental periods
below 4.0 seconds, and the peer review committee believes that the trend is
stable and defensible, then a height limit would not need to be imposed for
this reinforced concrete special moment frame system.
9-24 9: Example Applications FEMA P695
Table 9-10 Summary of Final Collapse Margins and Comparison to
Acceptance Criteria for Archetypes Redesigned with an Updated
Minimum Base Shear Requirement
Arch.
ID
Design Configuration Computed Overstrength and Collapse
Margin Parameters
Acceptance
Check
No. of
Stories
Framing
(Gravity
Loads)
SDC Static
O CMR .T SSF ACMR Accept.
ACMR
Pass/
Fail
Performance Group No. PG-5 (Short Period, 20' Bay Width Configuration)
2069 1 P Dmax 1.6 1.18 14.0 1.33 1.57 1.52 Pass
2064 2 P Dmax 1.8 1.50 19.6 1.33 2.00 1.52 Pass
-- 3 P Dmax 1.7* -- -- -- 2.13* -- --
Mean of Performance Group: 1.7* 1.34 16.8 1.33 1.90* 1.90 Pass
Performance Group No. PG-6 (Long Period, 20' Bay Width Configuration)
1003 4 P Dmax 1.6 1.61 10.9 1.41 2.27 1.52 Pass
1011 8 P Dmax 1.6 1.25 9.8 1.61 2.01 1.52 Pass
1013 12 P Dmax 1.7 1.45 11.4 1.61 2.33 1.52 Pass
1020 20 P Dmax 1.6 1.66 5.6 1.49 2.47 1.52 Pass
Mean of Performance Group: 1.6 1.49 9.4 1.53 2.27 1.90 Pass
Performance Group No. PG-1 (Short Period, 20' Bay Width Configuration)
2061 1 S Dmax 4.0 1.96 16.1 1.33 2.61 1.52 Pass
1001 2 S Dmax 3.5 2.06 14.0 1.33 2.74 1.52 Pass
-- 3 S Dmax 3.11 -- -- -- 2.63* -- --
Mean of Performance Group: 3.51 2.01 15.0 1.33 2.66* 1.90 Pass
Performance Group No. PG-3 (Long Period, 20' Bay Width Configuration)
1008 4 S Dmax 2.7 1.78 11.3 1.41 2.51 1.52 Pass
1012 8 S Dmax 2.3 1.63 7.5 1.58 2.58 1.52 Pass
1014 12 S Dmax 2.1 1.59 7.7 1.60 2.54 1.52 Pass
1021 20 S Dmax 2.0 1.98 5.7 1.50 2.96 1.52 Pass
Mean of Performance Group: 2.3 1.75 8.1 1.52 2.65 1.90 Pass
Selected Archetypes - Performance Group Nos. PG-4 and PG-8 (20' Bay Width)
4011 8 P Dmin 1.8 1.93 3.6 1.23 2.38 1.52 Pass
4013 12 P Dmin 1.8 2.29 4.3 1.26 2.89 1.52 Pass
4020 20 P Dmin 1.8 2.36 3.9 1.24 2.94 1.52 Pass
4021 20 S Dmin 2.8 3.87 3.8 1.24 4.81 1.52 Pass
Selected Archetypes - Performance Group Nos. PG-10 and PG-14 (30' Bay Width)
1009 4 P-30 Dmax 1.6 1.98 13.4 1.41 2.79 1.52 Pass
1010 4 S-30 Dmax 3.3 2.5 13.2 1.41 3.53 1.52 Pass
1. For completeness, the reinforced concrete special moment frame example assumes values of
static overstrength and ACMR of missing 3-story archetypes (based on the average of respective 2-
and 4-story values).
FEMA P695 9: Example Applications 9-25
9.2.9 Evaluation of O0 Using Final Set of Archetype Designs
At this point, the overstrength, O0, value can be established for use in the
proposed design provisions. Chapter 7 specifies that the Oo value should not
be taken as less than the largest average value of overstrength, O, from any
performance group, and should be conservatively increased to account for
variations in individual O values. The final Oo value should be rounded to
the nearest 0.5, and limited to a maximum value of 3.0.
Section 9.2.7 explained that the governing performance group for the
overstrength factor, .O, will either be PG-9 or PG-11 (as defined in Table 9-
1). This portion of this example is not entirely complete, because the
archetypes in these performance groups were not designed and assessed.
Even so, the results from other performance groups are used to determine the
appropriate overstrength factor, .O.
The average O values for the high-seismic 20-foot bay width performance
groups are 1.7 and 1.6 for the perimeter frame performance groups (PG-5
and PG-6) and 3.8 and 2.3 for the space frame performance groups (PG-1
and PG-3). Even though all performance groups were not completed, these
results are enough to show that the upper-bound value of Oo = 3.0 is
warranted, due to the average O value of 3.8 observed for PG-1.
9.2.10 Summary Observations
This example shows that current seismic provisions for reinforced concrete
special moment frame systems in ACI 318-05 and ASCE/SEI 7-05 provide
an acceptable level of collapse safety in SDC D, with an important
modification of imposing the minimum base shear requirement from the
ASCE 7-02 provisions. In addition, it demonstrates that the Methodology is
reasonably well calibrated because recent building code design provisions
lead to acceptable collapse safety, as defined by the Methodology (Section
7.4). This example also illustrates how the Methodology could be used as a
tool for testing possible changes to building code requirements and
evaluating code change proposals.
9.3 Example Application - Reinforced Concrete
Ordinary Moment Frame System
9.3.1 Introduction
In this example, a reinforced concrete ordinary moment frame system, as
defined by ACI 318-05 and ASCE/SEI 7-05, is evaluated as if it were a new
system proposed for inclusion in ASCE/SEI 7-05.
9-26 9: Example Applications FEMA P695
This example illustrates the Methodology for limited ductility systems,
which are only permitted in lower seismic design categories, are typical of
construction in the Central and Eastern United States, and are designed for a
much lower ratio of lateral to gravity loads. Since these systems lack the
capacity design and ductile detailing provisions of special moment frames,
reinforced concrete ordinary moment frame systems are susceptible to
additional modes of damage, such as shear failure in columns, leading to
rapid strength and stiffness deterioration. Because there are many possible
failure modes in reinforced concrete ordinary moment frame systems, this
example incorporates limit state checks for collapse modes that are not
simulated directly in nonlinear analysis.
9.3.2 Overview and Approach
The structural system for this example is defined by the design and detailing
provisions of ASCE/SEI 7-05 and ACI 318-05, and is evaluated using the
methods of Chapters 3 through 7. A set of structural system archetypes are
developed for reinforced concrete ordinary moment frame buildings,
nonlinear models are developed to simulate structural collapse, models are
analyzed to predict the collapse capacities of each archetype, and adjusted
collapse margin ratios are evaluated and compared to acceptance criteria.
This example has been adapted from collaborative research on the
development of structural archetypes for reinforced concrete ordinary
moment frames, calibration of nonlinear element models for collapse
simulation, simulation of structural response to collapse, and treatment of
uncertainties (Liel and Deierlein, 2008).
9.3.3 Structural System Information
Design Requirements
This example utilizes ACI 318-05 design requirements, which are extremely
detailed and represent years of accumulated research and building code
development. For the purpose of assessing composite uncertainty, the design
requirements are categorized as (A) Superior to reflect the high degree of
confidence in the design equations for reinforced concrete ordinary moment
frames. This rating is in agreement with the reinforced concrete special
moment frame example.
Test Data
The element models used in this study are the same as those in the reinforced
concrete special moment frame study and based on published test data from
the Pacific Earthquake Engineering Research Center’s Structural
FEMA P695 9: Example Applications 9-27
Performance Database (PEER, 2006b, Berry et al., 2004). Although
extensive data is available for reinforced concrete elements, there are still
limitations as discussed in Section 9.2.3. Accordingly, for the purpose of
assessing the total uncertainty, the test data is categorized as (B) Good, in
agreement with the reinforced concrete special moment frame example.
Seismic Design Criteria
Reinforced concrete ordinary moment frames are permitted only in Seismic
Design Categories B and below. Following the requirements of Section
4.2.1, the highest allowable SDC is the focus of the performance evaluation.
Accordingly, reinforced concrete ordinary moment frames are evaluated at
the lower limit of SDC B, where SD1 = 0.067 g and SDS = 0.167 g (Bmin), and
the upper limit of SDC B, where SD1 = 0.133 g and SDS = 0.33g (Bmax). A
subset of archetype models is also evaluated at the limits of SDC C, where
reinforced concrete ordinary moment frames are not permitted, to illustrate
how the Methodology can be used to evaluate current code limits.
9.3.4 Identification of Reinforced Concrete Ordinary Moment
Frame Archetype Configurations
Figure 9-8 shows the two-dimensional three-bay multi-story frame that is
considered an appropriate index archetype model for reinforced concrete
ordinary moment frame buildings. This is the same general archetype model
that was used to evaluate the reinforced concrete special moment frame
system, and includes joint panel elements, beam and column elements, elastic
foundation springs, and a leaning column to account for P-delta effects due to
the seismic mass on the gravity system.
Figure 9-8 Index archetype model for reinforced concrete ordinary
moment frames.
Using the above index archetype model, a set of structural archetype designs
are developed to represent the archetype design space, following the design
configuration and performance group requirements of Chapter 4. Since this
9-28 9: Example Applications FEMA P695
example is intended primarily to demonstrate that the Methodology is
applicable to reinforced concrete ordinary moment frame systems, a partial
group of archetypes is considered and it is recognized that this partial group
does not interrogate all basic design configurations that would be permitted
by the code. Therefore, a complete FEMA P695 assessment of the
reinforced concrete ordinary moment frame system would include a wider
range of design configurations, in addition to the subset of archetypes
evaluated here.
Table 9-11 shows how archetypes are organized into performance groups,
according to the requirements of Section 4.3. These groups represent the two
basic configurations considered in this example, the range of allowable
gravity loads and design ground motion intensities, and the building period
range (short- and long-period systems).
Table 9-11 Performance Groups for Evaluation of Reinforced Concrete
Ordinary Moment Frames
Performance Group Summary
Group No.
Grouping Criteria
Number of
Basic Archetypes
Config.
Design Load Level Period
Gravity Seismic Domain
PG-1
20-foot
Bay Width
High
(Space
Frame)
SDC Bmax
Short
01
PG-2 Long
PG-3
SDC Bmin
Short
PG-4 Long
PG-5
Low
(Perimeter
Frame)
SDC Bmax
Short
PG-6 Long
PG-7
SDC Bmin
Short
PG-8 Long
PG-9
30-foot
Bay Width
High
(Space
Frame)
SDC Bmax
Short 02
PG-10 Long 4
PG-11
SDC Bmin
Short 02
PG-12 Long 4
PG-13
Low
(Perimeter
Frame)
SDC Bmax
Short 02
PG-14 Long 4
PG-15
SDC Bmin
Short 02
PG-16 Long 4
1. Performance of reinforced concrete ordinary moment frame archetypes with 20-foot beam
span not evaluated, because 30-foot spans judged to be representative for reinforced concrete
ordinary moment frames.
2. No short-period reinforced concrete ordinary moment frame archetypes are considered in
this example; for a complete reinforced concrete ordinary moment frame assessment, a 1-
story would need to be investigated.
FEMA P695 9: Example Applications 9-29
Table 9-11 shows that only four performance groups are utilized in this
example, accounting for the variations in seismic design level and gravity
load level. For reinforced concrete ordinary moment frame buildings, the
bay width is not varied, and the rationale for this is explained below. Since
reinforced concrete ordinary moment frames tend to be flexible systems, all
designs fall into the long-period performance groups. For a full application
of the Methodology, at least one one-story building would need to be
evaluated in each of the short-period performance groups (PG-9, PG-11, PG-
13, and PG-15).
Space frames, which are typical for OMF designs are “high gravity” systems.
“Low gravity” systems could represent either perimeter frames of a flat plate
system, or the perimeter framing of a one-way joist system. Figure 9-9 and
Figure 9-10 show the layout of the archetype space and perimeter frame
systems. Buildings have a bay spacing of 30-feet and cover a range of
building heights (2-, 4-, 8-, or 12-stories). The bay spacing is different from
the default of 20-feet used in the special moment frame example, to better
reflect the typical configurations of gravity-dominated reinforced concrete
ordinary moment frame designs. For space-frames, a transverse span of 35
feet was used to maximize the gravity load contribution, whereas a smaller
transverse span of 30 feet was used for the lightly loaded perimeter-frames.
Figure 9-9 High gravity (space frame) layout.
9-30 9: Example Applications FEMA P695
Figure 9-10 Low gravity (perimeter frame) layout. Interior (gravity) columns
are not shown for clarity.
Table 9-12 summarizes the design properties of the reinforced concrete
ordinary moment frame archetype designs needed to evaluate SDC B,
including the design base shear and code-calculated structural period. Each
of the archetypes was fully designed in accordance with the governing design
requirements (ASCE/SEI 7-05 and ACI 318-05). Additional information on
archetype designs is provided in Appendix C.
The design documentation provided is consistent with that provided for the
special moment frame models and an example for Archetype ID 9203 is
shown in Figure 9-11. Similar documentation has been maintained for all
other archetype designs.
FEMA P695 9: Example Applications 9-31
Table 9-12 Reinforced Concrete Ordinary Moment Frame Archetype Design
Properties, SDC B
Archetype
ID
No. of
Stories
Key Archetype Design Parameters
Framing
(Gravity
Loads)
Seismic Design Criteria
SMT(T)
[g]
SDC R T [sec] T1
[sec]
V/W
[g]
Performance Group No. PG-16 (Long Period)
9101 2 P Bmin 3 0.55 1.56 0.041 0.18
9103 4 P Bmin 3 0.99 2.81 0.023 0.10
9105 8 P Bmin 3 1.81 4.58 0.012 0.06
9107 12 P Bmin 3 2.59 5.80 0.010 0.04
Performance Group No. PG-12 (Long Period)
9102 2 S Bmin 3 0.55 0.85 0.041 0.18
9104 4 S Bmin 3 0.99 1.49 0.023 0.10
9106 8 S Bmin 3 1.81 2.53 0.012 0.06
9108 12 S Bmin 3 2.59 2.85 0.010 0.04
Performance Group No. PG-14 (Long Period)
9201 2 P Bmax 3 0.51 1.23 0.087 0.39
9203 4 P Bmax 3 0.93 1.93 0.048 0.22
9205 8 P Bmax 3 1.70 3.39 0.026 0.12
9207 12 P Bmax 3 2.44 4.43 0.018 0.08
Performance Group No. PG-10 (Long Period)
9202 2 S Bmax 3 0.51 0.81 0.087 0.39
9204 4 S Bmax 3 0.93 1.36 0.048 0.22
9206 8 S Bmax 3 1.70 2.35 0.026 0.12
9208 12 S Bmax 3 2.44 2.85 0.018 0.08
9-32 9: Example Applications FEMA P695
Floor 5
Floor 4
Floor 3
feet
Floor 2
feet
Grade beam column height (in) = Basement column height (in) =
feet
Design base shear = g, k
f'c beams = ksi f'c,cols,upper = ksi
fy,rebar,nom. = ksi f'c,cols,lower = ksi
32 0
0.0008 0.0008
0.0308 0.0345
11.5 11.5 11.5 11.5
0.0008 0.0008
0.0348 0.0308
24 24 24 24
. =
.' =
. =
.' =
.sh =
s (in) =
0.0190
0.0008
11.5
26 26 26 26
6.5
11.5 10.0
0.0013
0.0011
0.0009
28
10.0
24
0.0098
0.0155
28
24
0.0098
0.0155
0.0225
0.0008
28
24
0.0098
0.0155
0.0013
6.5
0.0009
8.5
26 26
0.0190 0.0225
24
s (in) =
11.5
24
26
28
h (in) =
.tot =
30
13
15
24
26
11.5
0.0011
8.0
24 24
.sh =
.sh =
s (in) =
h (in) =
b (in) =
b (in) =
.tot =
.sh =
s (in) =
s (in) =
h (in) =
b (in) =
.tot =
h (in) =
b (in) =
.tot =
.sh =
28
24
0.0038
0.0133
0.0078
0.0007
12.5
0.0007
12.5
28
24
26
24
0.0038
0.0078
0.0110
0.0008
11.5
28
26
24
0.0108
0.0008
11.5
28
12.5
24
0.0038
0.0078
0.0007
26
24
0.0133
0.0008
26 26 26
11.5
0.0053
0.0113
0.0007
12.5
0.0008
24
24 24 24 24
0.0140 0.0145 0.0145 0.0140
0.0008 0.0008 0.0008 0.0008
0.0011
8.0
0.0083
0.0143
11.5 11.5
28
24
0.0083
0.0143
24
h (in) =
b (in) =
. =
.' =
.sh =
s (in) =
h (in) =
b (in) =
b (in) =
28
24
0.0083
0.0143
11.5
26
9.5
0.0053
0.0113
0.0009 0.0007
0.0053
0.0113
.sh =
s (in) =
h (in) =
Story 1 Story 2 Story 3
26
Story 4
0.0007
12.5
28
24
28
24
12.5
60
0.048 301
5.0
5.0 5.0
(a)
Floor 4
Floor 3
Floor 2 1.35
1.91
0.07
1.16
0.64
1.26
1.16
1.18
1.15
0.59
0.59
2.16
0.05
1.18
1.16
0.71
0.10
0.61
1.54
0.92
0.67
0.94
0.84
1.16
1.40
fMn/Mu =
(fMn/Mu)neg =
(fMn/Mu)pos =
Mn,pos/Mn,neg =
0.99
fVn/Vmpr = 0.83
P/Agf'c = 0.07
fMn/Mu =
1.35
0.59
fVn/Vmpr =
0.10
1.18
1.15
Mn,pos/Mn,neg =
(fMn/Mu)neg =
(fMn/Mu)pos =
0.64
0.16
(fMn/Mu)neg =
fVn/Vmpr =
P/Agf'c =
0.7
fVn/Vmpr =
0.45
1.16
0.16
(fMn/Mu)pos =
1.15
fVn/Vmpr =
0.78
Mn,pos/Mn,neg =
0.59
1.15 1.16
0.45
0.78
0.84
0.92
0.67
0.99
1.28
1.52
0.64
2.16 0.78
fMn/Mu =
fVn/Vmpr =
0.05
1.16
1.16
P/Agf'c =
0.94
fMn/Mu =
fVn/Vmpr =
P/Agf'c =
SCWB = 0.61
Joint FVn/Vu = 1.54
0.53
0.08
0.53
0.83
0.08
1.25
1.43
0.59
2.03
0.78
0.96
1.15 1.16 1.16 1.15
0.96
1.16 1.16
1.26 1.16 1.16
1.40
3.28
3.28
0.02
0.70
1.17
1.17
0.48
3.28
0.48
0.04
0.71
0.72
0.49
1.22
1.45
8.15
1.56
1.53
1.17
0.40
1.30
0.40
1.30
1.18
1.16
0.49
8.15
1.27
8.15
1.18
1.16
0.49
0.12
0.02
1.17
1.15
1.18
1.78
0.12
1.91
Story 1 Story 2 Story 3 Story 4
0.72
0.04
1.15
0.48
1.55
1.6%
Design
Drifts:
0.9%
1.4%
1.8%
(b)
Figure 9-11 Design documentation for a 4-story reinforced concrete
ordinary moment frame archetype with perimeter frame.
FEMA P695 9: Example Applications 9-33
9.3.5 Nonlinear Model Development
The nonlinear models of reinforced concrete ordinary moment frames are
similar to those developed in the reinforced concrete special moment frame
system example; model development was discussed in Section 9.2.5 and is
also discussed in more exhaustive detail in Appendix E. For illustration,
Figure 9-12 shows the predicted modeling parameters for each element of
Archetype ID 9203. As expected, reinforced concrete beams and columns in
these structures have substantially less ductility than their special moment
frame counterparts, as reflected in the deformation capacity (.cap,pl and .pc)
and cyclic deterioration parameters (.). As with the special moment frame
example, these models were implemented in the OpenSees software
platform, developed by the Pacific Earthquake Engineering Research Center
(OpenSees, 2006).
Shear failure and subsequent loss of gravity-load bearing capacity in columns
is not explicitly included in the analysis models, but according to
requirements of Section 5.5, it is incorporated through post-processing as a
non-simulated failure mode. Shear-induced axial failure of columns is
difficult to simulate using available technologies and test data, and is
accounted for in post-processing for this reason; however, if possible, it
would be better to incorporate this failure mode directly into the nonlinear
structural simulation.
Floor 5
Floor 4
Floor 3
Floor 2
Model periods (sec): T1 = T2= T3 =
fy,rebar,expected=
0.030
0.033
(P/Agf'c)exp =
Tpc (rad) =
0.022
51
57
9270
0.08
0.030
0.35
1.20
11518
0.03
-14949
. =
0.030
7876
-14949
83
7876
0.35
0.026
0.035
0.35
55
1.974E+08
-0.021
-9061
5288
0.35
1.21
3648
-14949
1.974E+08
0.35
5825
54
0.033
53
0.014
-0.022
6557 72
7876
0.35
0.022
-0.029
0.06
-0.035
0.047
1.974E+08
My,exp (k-in) =
Mc/My =
Tcap,pl (rad) =
Tpc (rad) =
EIstf/EIg =
(P/Agf'c)exp =
1.974E+08
0.021
5065
-12223
0.35
0.035
1.21
. =
0.01
My,exp (k-in) =
My,exp (k-in) = 8924
0.35
Mc/My = 1.21
Tcap,pl (rad) = 0.020
EIstf/EIg =
(P/Agf'c)exp =
EIstf/EIg = 0.35
Mc/My = 1.20
Tcap,pl (rad) = 0.021
0.35
1.20
9270
52
0.04
0.032
My,exp (k-in) = 15407
48
Tcap,pl (rad) =
Tpc (rad) =
. =
(P/Agf'c)exp = Tpc (rad) =
. =
(Tcap,pl)pos (rad) =
EIstf, w/ Slab (k-in2/rad) =
95
EIstf/EIg =
Mc/My =
0.025
0.06
Tpc (rad) =
. = 51
0.11
0.35
1.19
15310
51
0.06
0.35
1.19
My,n,slab,exp (k-in) =
Tpc (rad) =
(Tcap,pl)neg (rad) =
EIstf, w/ Slab (k-in2/rad) =
95
My,pos,exp (k-in) =
0.35
. =
0.032
0.023
52
0.04
1.20
0.030
5825
0.021
54
0.01
0.35
1.21
0.35
53
1.974E+08
6557
0.035
55
0.014
0.035
0.35
5065
0.03
-0.021
0.035
55
5065
1.21
0.020
0.035
0.014
-9061
0.35
0.019
3648
-9061
0.35
0.014
0.35
1.974E+08
3648
0.35
55
5386
0.06
53
0.03
0.033
7608
-12223
-12223
0.030
0.021
1.20 1.20 1.21
1.974E+08
0.026
-0.035
53
0.03
0.020
0.35
1.20
0.35
1.20
1.974E+08
-16090
0.030
8924
0.033
-0.033
0.045
0.08
51
0.041
11518
79
0.022
9270
57
0.35
0.032
-0.041
0.053
-16090
0.35
0.032
1.974E+08
0.026
15491
Story 1 Story 2 Story 3 Story 4
0.019
0.35
0.014
-0.022
1.974E+08
0.035
55
0.020
7608
0.35
1.974E+08
-0.041
0.053
0.023
My,pos,exp (k-in) =
(Tcap,pl)neg (rad) =
My,n,slab,exp (k-in) =
EIstf/EIg =
-16090
0.35
0.020 0.020
(Tcap,pl)pos (rad) =
15491
My,n,slab,exp (k-in) =
EIstf/EIg =
EIstf, w/ Slab (k-in2/rad) =
(Tcap,pl)pos (rad) =
EIstf/EIg =
(Tcap,pl)neg (rad) =
Tpc (rad) =
. =
My,pos,exp (k-in) =
48
0.11
0.025
0.35
0.014
-0.021
0.035
55
1.974E+08
0.047
83
-0.022
67 ksi
4.03
1.93 0.60 0.33
Mass tributary to one frame for lateral load (each floor) (k-s-s/in):
Figure 9-12 Structural modeling documentation for a 4-story reinforced
concrete ordinary moment frame archetype (Archetype ID
9203).
9-34 9: Example Applications FEMA P695
9.3.6 Nonlinear Structural Analysis
Static Pushover Analysis
To compute the system overstrength, O, of each archetype design, a
monotonic static pushover analysis is utilized with the lateral load pattern
prescribed in ASCE/SEI 7-05. Figure 9-13 illustrates results of static
pushover analysis for Archetype ID 9203. For this reinforced concrete
ordinary moment frame archetype, the effective yield roof displacement, .y,eff
, is computed according to Equation 6-7 as 0.0047hr. The capping
displacement (the onset of negative stiffness) occurs at roof displacement of
about 0.0125hr and the displacement at 20% strength loss is .u
= 0.018hr.
The period-based ductility (which will later be used to adjust the CMR
according to Section 7.2), can be computed as µT = .u
/ .y,eff = 3.8. These
structures have substantially less ductility than their reinforced concrete
special moment frame counterparts (e.g., Figure 9-13 compared to Figure 9-
6).
Figure 9-13 Monotonic static pushover for reinforced concrete ordinary
moment frame archetype (Archetype ID 9203).
Results from pushover analysis are tabulated below in Table 9-13 for all
performance groups in SDC B. Pertinent results include computed static
overstrength (O) and period-based ductility (µT = .u
/ .y,eff). When nonsimulated
collapse modes are considered in the collapse assessment, .u
should account for the occurrence of the non-simulated collapse mode.
Nonlinear Dynamic Analysis and Simulation
The collapse capacity for each archetype design is computed according to
Section 6.4, using the incremental dynamic analysis (IDA) approach.
Ground motions (the Far-Field record set) and scaling procedures are based
on requirements of Sections 6.2.2 and 6.2.3, respectively. Figure 9-14
illustrates how the IDA method is used to compute the sidesway collapse
FEMA P695 9: Example Applications 9-35
margin ratio of Archetype ID 9203. It should be noted that a full IDA is not
required to quantify CMR, as discussed in Section 6.4.2.
Figure 9-14 Incremental dynamic analysis to collapse showing the Maximum
Considered Earthquake ground motion intensity (SMT), median
collapse capacity intensity (SCT), and sidesway collapse margin
ratio (CMRSS) for Archetype ID 9203.
Table 9-13 summarizes the IDA results for the selected 16 archetype designs.
These results reflect only the sidesway collapse mechanism, which is directly
simulated in the structural analysis model. The subscript “SS” denotes
consideration of sidesway collapse only and not the non-simulated collapse
modes. As such, CMRSS is not used for system evaluation, but is presented
here for illustrative purposes only.
Non-Simulated Collapse Modes
In this example, only sidesway collapse mechanisms based on strength and
stiffness degradation due to flexure and flexure-shear are simulated directly in
the analysis. Nonlinear models do not predict the occurrence of column shear
failure and subsequent rapid deterioration and loss of gravity-load bearing
capacity, which may occur because reinforced concrete ordinary moment
frame columns have light transverse reinforcement and are not subject to
capacity design requirements. A detailed discussion of possible failure modes
in reinforced concrete moment frames is included in Appendix D.
To account for failure modes not directly included in analysis modes,
according to Section 5.5, further post-processing of the nonlinear dynamic
results is required. Results of static pushover analysis also require postprocessing
so that the computed period-based ductility, µT, (which is used to
compute the SSF) accounts for non-simulated failure modes. The ultimate
0 0.02 0.04 0.06
0
0.2
0.4
0.6
0.8
1
1.2
1.4
Sa(T
1
) [g]
Maximum Interstory Drift Ratio
CMRSS = 2.36
SMT = 0.22g
ST(T1)
[g]
SCT = 0.51g
9-36 9: Example Applications FEMA P695
roof displacement, du, should be based on the roof displacement
corresponding to a 20% loss of base shear or the occurrence of the nonsimulated
collapse mode, whichever occurs at a smaller displacement.
Table 9-13 Summary of Pushover Analysis and IDA Sidesway Collapse
Results for Reinforced Concrete Ordinary Moment Frame
Archetype Designs, SDC B
Archetype
ID
Design Configuration Pushover and IDA Results
No. of
Stories
Framing
(Gravity
Loads)
Seismic
SDC Static . µT
1 SCT [T]
(g)
CMRSS
Performance Group No. PG-16 (Long Period)
9101 2 P Bmin 2.0 3.7 0.91 4.99
9103 4 P Bmin 1.8 3.0 0.33 3.30
9105 8 P Bmin 2.6 3.1 0.18 3.31
9107 12 P Bmin 2.3 2.5 0.14 3.68
Mean of Performance Group: 2.2 3.1 NA 3.82
Performance Group No. PG-12 (Long Period)
9102 2 S Bmin 6.6 3.0 1.40 7.69
9104 4 S Bmin 5.3 2.1 0.68 6.70
9106 8 S Bmin 6.0 3.0 0.39 7.05
9108 12 S Bmin 6.0 4.2 0.35 9.07
Mean of Performance Group: 6.0 3.1 NA 7.63
Performance Group No. PG-14 (Long Period)
9201 2 P Bmax 1.6 3.5 1.07 2.72
9203 4 P Bmax 1.6 3.8 0.51 2.36
9205 8 P Bmax 1.5 2.8 0.25 2.11
9207 12 P Bmax 1.7 3.0 0.18 2.22
Mean of Performance Group: 1.6 3.3 NA 2.36
Performance Group No. PG-10 (Long Period)
9202 2 S Bmax 2.9 3.3 1.36 3.47
9204 4 S Bmax 3.0 2.3 0.83 3.87
9206 8 S Bmax 3.1 3.0 0.41 3.49
9208 12 S Bmax 3.8 5.8 0.38 4.65
Mean of Performance Group: 3.2 3.6 NA 3.87
1. Due to time constraints, T values do not account for non-simulated collapse modes,
Even so, these must be included when computing period-based ductility.
Fragility functions (adopted from Aslani, 2005) are introduced to determine
the drift at which the column loses its ability to carry gravity loads following
shear failure. To determine if the shear-induced axial failure mode has
occurred, the column drift level (demand) is extracted from the analysis data
and compared to the median drift level associated with column axial failure
from the fragility functions (capacity). If the median drift level has been
exceeded in any column, the non-simulated collapse mode is assumed to
have occurred. This approach is likely conservative because is assumes that
FEMA P695 9: Example Applications 9-37
when the non-simulated collapse limit is exceeded in one element, it triggers
collapse of the entire building. In many cases, gravity loads can be
redistributed to nearby elements, and the axial failure of a single column will
not cause complete collapse of an entire structure. However, because the
structural simulation model cannot represent system behavior and
redistribution after the vertical collapse of a column, this is taken as the nonsimulated
collapse state.
Using the approach to non-simulated collapse modes in Section 5.5, the
collapse fragility is adjusted, increasing the probability of collapse to include
both the simulated and non-simulated failure modes. This change reduces
the collapse margin ratio. This reduction is more significant in space frame
structures that have higher column axial loads. The effect of non-simulated
failure modes on the computed collapse margin ratios for the 16 archetype
structures is shown in Table 9-14.
Table 9-14 Effect of Non-Simulated Collapse Modes on Computed Collapse
Margin Ratios for Reinforced Concrete Ordinary Moment Frame
Archetypes, SDC B
Archetype
ID
Design Configuration Collapse Margin Ratios
No. of
Stories
Framing
(Gravity
Loads)
Seismic
SDC
SCT [T]
(g) CMRSS CMRnonsimulated
Percent
Decrease
Performance Group No. PG-16 (Long Period, 30’ Bay Width)
9101 2 P Bmin 0.91 4.99 4.96 0.6%
9103 4 P Bmin 0.33 3.30 3.08 6.7%
9105 8 P Bmin 0.18 3.31 2.57 22.3%
9107 12 P Bmin 0.14 3.68 2.96 19.5%
Mean of Performance Group: NA 3.82 3.39 11.2%
Performance Group No. PG-12 (Long Period, 30’ Bay Width)
9102 2 S Bmin 1.40 7.69 3.98 48.2%
9104 4 S Bmin 0.68 6.70 2.79 58.4%
9106 8 S Bmin 0.39 7.05 4.36 38.2%
9108 12 S Bmin 0.35 9.07 4.19 53.8%
Mean of Performance Group: NA 7.63 3.83 49.8%
Performance Group No. PG-14 (Long Period, 30’ Bay Width)
9201 2 P Bmax 1.07 2.72 2.04 25.0%
9203 4 P Bmax 0.51 2.36 1.99 15.6%
9205 8 P Bmax 0.25 2.11 1.68 20.5%
9207 12 P Bmax 0.18 2.22 1.93 13.4%
Mean of Performance Group: NA 2.36 1.91 18.9%
Performance Group No. PG-10 (Long Period, 30’ Bay Width)
9202 2 S Bmax 1.36 3.47 1.79 48.5%
9204 4 S Bmax 0.83 3.87 2.08 46.3%
9206 8 S Bmax 0.41 3.49 2.48 28.9%
9208 12 S Bmax 0.38 4.65 1.95 58.0%
Mean of Performance Group: NA 3.87 2.08 46.4%
9-38 9: Example Applications FEMA P695
9.3.7 Performance Evaluation for SDC B
The collapse margin ratio is adjusted according to Section 7.2 to account for
the proper spectral shape of rare ground motions through the spectral shape
factor, SSF. According to Table 9-4 (from Section 7.2.2), the spectral shape
factor is computed based on the SDC, the period-based ductility, µT, and the
structural periods, T and T1; these parameter values are documented in Table
9-12 and later in Table 9-15. These structures have an average period-based
ductility, µT, of 3.2, resulting in an average SSF of 1.17, as shown below in
Table 9-15. Since the ordinary moment frame archetypes are designed for
SDC B, where the benefit of spectral shape is more limited, and they have
limited deformation capacity, the SSF values are smaller than for the special
moment frame example.
To assess the ordinary moment frame system, the composite uncertainty,
ßTOT, in collapse capacity is needed. As described above, the quality of test
data is rated (B) Good and the quality of structural system design
requirements is rated (A) Superior. According to Section 5.7, the uncertainty
in the archetype model is based on (1) the completeness of the set of index
archetypes and (2) how well the structural collapse behavior is captured
either by direct simulation or use of non-simulated component checks.
Regarding the completeness of the set of archetypes, it has already been
stated that this example is a partial example, and the archetype set is not
complete; however, for the purpose of quantifying model uncertainty in this
example, it is assumed that the set of archetypes is complete. Regarding the
collapse modeling, although the component model is calibrated to columns
that fail in flexure-shear, the structural simulation model may lose some
fidelity after the occurrence of shear failure, because shear failure is not
directly predicted. Based on this rationale, the archetype model quality is
rated as (C) Fair. Based on these individual values for the ordinary moment
frame example, the composite uncertainty determined to be ßTOT = 0.575,
according to Section 7.3.4.
The acceptable collapse margin ratio is determined from the composite
uncertainty and the acceptable conditional probability of collapse under the
MCE ground motions. Chapter 7 defines the collapse performance
objectives as: (1) a conditional collapse probability of 20% for each
archetype building, and (2) a conditional collapse probability of 10% for
each performance group. For the reinforced concrete ordinary moment frame
systems with a composite uncertainty of 0.575, this corresponds to a required
ACMR20% of 1.62 for each archetype building, with a required average
ACMR10% of 2.09 for each performance group. These values are taken from
Table 7-3 (also shown in Table 9-7).
FEMA P695 9: Example Applications 9-39
Table 9-15 presents the final results and acceptance criteria for the 16
reinforced concrete ordinary moment frame archetypes in SDC B. As shown
in Table 9-15, the reinforced concrete ordinary moment frame structural
system passes for each of the two performance groups in both Bmax and Bmin.
Although the average ACMR is close to the limit for PG-14, in some cases
the ACMRs are considerably above the required values. The large ACMRs
tend to occur for structures with substantial overstrength, which resulted
from the dominance of gravity loading in the design, especially where
seismic design forces are low compared to gravity loading (e.g., spaceframes
in SDC Bmin).
Table 9-15 Summary of Collapse Margins and Comparison to Acceptance
Criteria for Reinforced Concrete Ordinary Moment Frame
Archetypes, SDC B
Arch.
ID
Design Configuration Computed Overstrength and
Collapse Margin Parameters
Acceptance
Check
No. of
Stories
Framing
(Gravity
Loads)
SDC Static
O CMR .T
. SSF ACMR Accept.
ACMR
Pass/
Fail
Performance Group No. PG-16 (Long Period)
9101 2 P Bmin 2.0 4.96 3.7 1.09 5.40 1.62 Pass
9103 4 P Bmin 1.8 3.08 3.0 1.12 3.44 1.62 Pass
9105 8 P Bmin 2.6 2.57 3.1 1.21 3.10 1.62 Pass
9107 12 P Bmin 2.3 2.96 2.5 1.18 3.50 1.62 Pass
Mean of Performance Group: 2.2 3.39 3.1 1.15 3.86 2.09 Pass
Performance Group No. PG-12 (Long Period)
9102 2 S Bmin 6.6 3.98 3.0 1.08 4.29 1.62 Pass
9104 4 S Bmin 5.3 2.79 2.1 1.09 3.03 1.62 Pass
9106 8 S Bmin 6.0 4.36 3.0 1.21 5.26 1.62 Pass
9108 12 S Bmin 6.0 4.19 4.2 1.37 5.75 1.62 Pass
Mean of Performance Group: 6.0 3.83 3.1 1.19 4.58 2.09 Pass
Performance Group No. PG-14 (Long Period)
9201 2 P Bmax 1.6 2.04 3.5 1.09 2.21 1.62 Pass
9203 4 P Bmax 1.6 1.99 3.8 1.13 2.24 1.62 Pass
9205 8 P Bmax 1.5 1.68 2.8 1.20 2.01 1.62 Pass
9207 12 P Bmax 1.7 1.93 3.0 1.21 2.33 1.62 Pass
Mean of Performance Group: 1.6 1.91 3.3 1.15 2.20 2.09 Pass
Performance Group No. PG-10 (Long Period)
9202 2 S Bmax 2.9 1.79 3.3 1.08 1.94 1.62 Pass
9204 4 S Bmax 3.0 2.08 2.3 1.09 2.28 1.62 Pass
9206 8 S Bmax 3.1 2.48 3.0 1.21 2.99 1.62 Pass
9208 12 S Bmax 3.8 1.95 5.8 1.31 2.56 1.62 Pass
Mean of Performance Group: 3.2 2.08 3.6 1.17 2.44 2.09 Pass
1. Due to time constraints, T values do not account for non-simulated collapse modes.
Even so, these must be included when computing period-based ductility.
9-40 9: Example Applications FEMA P695
9.3.8 Performance Evaluation for SDC C
A smaller subset of archetypes was utilized to assess whether reinforced
concrete ordinary moment frame buildings designed in SDC C would meet
the collapse safety criteria of Section 7.4. The archetype designs and results
for Bmax are used in this assessment, since Bmax is identical to Cmin. In
addition, we consider four additional archetype reinforced concrete ordinary
moment frames, designed for Cmax, as shown in Table 9-16. These four
designs compose two partial performance groups for SDC Cmax; complete
performance groups would include at least three designs in each group.
Seismic performance is assessed using the same procedure as described for
the evaluation of reinforced concrete ordinary moment frames in SDC B.
Table 9-17 shows the computed overstrength, O, and period-based ductility,
µT, factors from static pushover analysis. The computed collapse margin
ratios, showing only the results that include non-simulated failure modes, are
also reported in Table 9-17. This table also compares the adjusted collapse
margin ratios to the acceptance criteria.
Table 9-16 Reinforced Concrete Ordinary Moment Frame Archetype Design
Properties for SDC C Seismic Criteria
Archetype
ID
No. of
Stories
Key Archetype Design Parameters
Framing
(Gravity
Loads)
Seismic Design Criteria SMT(T)
[g]
SDC R T [sec] T1 [sec] V/W [g]
SDC Cmin Performance Group (Long Period) - PG-14
9201 2 P Cmin 3 0.51 1.23 0.087 0.39
9203 4 P Cmin 3 0.93 1.93 0.048 0.22
9205 8 P Cmin 3 1.70 3.39 0.026 0.12
9207 12 P Cmin 3 2.44 4.43 0.018 0.08
SDC Cmin Performance Group (Long Period) - PG-10
9202 2 S Cmin 3 0.51 0.81 0.087 0.39
9204 4 S Cmin 3 0.93 1.36 0.048 0.22
9206 8 S Cmin 3 1.70 2.35 0.026 0.12
9208 12 S Cmin 3 2.44 2.85 0.018 0.08
SDC Cmax Performance Group (Long Period)
9303 4 P Cmax 3 0.87 1.51 0.077 0.34
9307 12 P Cmax 3 2.29 3.72 0.029 0.13
SDC Cmax Performance Group (Long Period)
9304 4 S Cmax 3 0.87 1.30 0.077 0.34
9308 12 S Cmax 3 2.29 2.57 0.029 0.13
FEMA P695 9: Example Applications 9-41
Table 9-17 Summary of Pushover Results, Collapse Margins, and
Comparison to Acceptance Criteria for Reinforced Concrete
Ordinary Moment Frame Archetypes, SDC C
Arch.
ID
Design Configuration Computed Overstrength and Collapse
Margin Parameters
Acceptance
Check
No. of
Stories
Framing
(Gravity
Loads)
SDC Static
O CMR .T
. SSF ACMR Accept.
ACMR
Pass/
Fail
SDC Cmin Performance Group (Long Period) - PG-14
9201 2 P Cmin 1.6 2.04 3.5 1.09 2.21 1.62 Pass
9203 4 P Cmin 1.6 1.99 3.8 1.13 2.24 1.62 Pass
9205 8 P Cmin 1.5 1.68 2.8 1.20 2.01 1.62 Pass
9207 12 P Cmin 1.7 1.93 3.0 1.21 2.33 1.62 Pass
Mean of Performance Group: 1.6 1.91 3.3 1.15 2.20 2.09 Pass
SDC Cmin Performance Group (Long Period) - PG-10
9202 2 S Cmin 2.9 1.79 3.3 1.08 1.94 1.62 Pass
9204 4 S Cmin 3.0 2.08 2.3 1.09 2.28 1.62 Pass
9206 8 S Cmin 3.1 2.48 3.0 1.21 2.99 1.62 Pass
9208 12 S Cmin 3.8 1.95 5.8 1.31 2.56 1.62 Pass
Mean of Performance Group: 3.2 2.08 3.6 1.17 2.44 2.09 Pass
SDC Cmax Performance Group (Long Period)
9303 4 P Cmax 1.5 1.55 3.6 1.14 1.76 1.62 Pass
9307 12 P Cmax 1.4 1.03 2.1 1.16 1.19 1.62 Fail
Mean of Performance Group: 1.5 1.29 2.9 1.15 1.48 2.09 Fail
SDC Cmax Performance Group (Long Period)
9304 4 S Cmax 2.1 1.97 3.3 1.13 2.23 1.62 Pass
9308 12 S Cmax 2.7 1.58 4.7 1.27 2.01 1.62 Pass
Mean of Performance Group: 2.4 1.78 4.0 1.20 2.12 2.09 Fail
1. Due to time constraints, T values do not account for non-simulated collapse modes.
Even so, these must be included when computing period-based ductility.
As shown in Table 9-17, archetype structures with a perimeter configuration
fail the acceptance criteria for Cmax seismic criteria for both the average
ACMR of the long-period performance group, as well as individual
Archetype ID 9307 (i.e., taller archetype). These results indicate that the
Methodology would not allow the use of reinforced concrete ordinary
moment frames from SDC C, consistent with current Code restrictions. In
order to be permitted in SDC C a lower R factor or other changes in design
requirements would be necessary.
9.3.9 Evaluation of Oo Using Set of Archetype Designs
Development of the overstrength factor, Oo, is based on SDC B archetype
designs, since this is the highest SDC that is currently allowed for this
9-42 9: Example Applications FEMA P695
system. If the system were being approved for use in SDC C, then the SDC
C archetypes would instead be used for establishing Oo.
The first step is to compute the overstrength values (O) for each individual
archetype building. There is a relatively wide range of overstrength observed
for the set of archetype designs (O ranges from 1.5 to 6.6, as reported in
Table 9-17). The PG-14 archetypes (SDC Bmax, perimeter frame) that govern
the R factor have computed O values ranging from 1.5 to 1.7, with an average
of 1.6. The PG-10 archetypes (SDC Bmax, space frame) have higher
computed overstrength values, between 2.9 and 3.8, with an average of 3.2.
The PG-12 archetypes (SDC Bmin, space frame) have the largest overstrength
values, ranging up to 6.6, with an average of 6.0. Due to the average values
being greater than 3.0 for one or more of the performance groups, the upperbound
value of Oo = 3.0 is recommended for reinforced concrete ordinary
moment frames, based on the requirements of Section 7.6.
9.3.10 Summary Observations
This example shows that current seismic provisions for reinforced concrete
ordinary moment frame systems provide an acceptable level of collapse
safety for SDC B, but not for SDC C. These results are consistent with the
provisions for use of reinforced concrete ordinary moment frame in
ASCE/SEI 7-05. Levels of collapse safety observed for reinforced concrete
ordinary moment frames in SDC Bmax are comparable to those for reinforced
concrete special moment frames in SDC Dmax, and in Bmin these systems have
a large margin against collapse. Note that in some cases the collapse safety
of actual reinforced concrete ordinary moment frame buildings may be
higher than calculated in this example, particularly when there are a large
number of gravity-designed columns which increase structural strength and
stiffness. Even so, for the purpose of establishing seismic design
requirements according to this Methodology, Chapter 4 requires that only
elements that are designed as part of the seismic-force-resisting system, and
are accordingly governed by seismic design requirements, be included in this
assessment.
To account for non-simulated failure modes as described in Section 5.5,
component limit state checks are incorporated through post-processing of
dynamic analysis results. In some cases, incorporation of the non-simulated
failure modes significantly affects the collapse margin ratio, demonstrating
the importance of carefully considering and including all critical failure
modes either explicitly in the simulation models or in non-simulated limit
state checks. Use of non-simulated failure modes to account for collapse due
FEMA P695 9: Example Applications 9-43
to column loss of gravity-load bearing capacity may be conservative, because
it does not allow for load redistribution.
9.4 Example Application - Wood Light-Frame System
9.4.1 Introduction
In this example, a wood light-frame system with structural panel sheathing is
considered as if it were a new system proposed for inclusion in ASCE/SEI
7-05.
9.4.2 Overview and Approach
Wood light-frame system design requirements of ASCE/SEI 7-05 are used as
the framework. A set of structural archetypes are developed for wood lightframe
buildings, nonlinear models are developed to simulate structural
collapse, models are analyzed to predict the collapse capacities of each
design, and the adjusted collapse margin ratio, ACMR, is evaluated and
compared to acceptance criteria.
Seismic performance factors (SPFs) are determined by iteration until the
acceptance criteria of the Methodology are met. This example begins with
an initial value of R = 6 and checks if such designs pass the acceptance
criteria of Section 7.4. This value is different from the current value of R =
6.5 for wood light-frame shear wall systems with wood structural panel
sheathing in ASCE/SEI 7-05. It has been rounded to the nearest whole
number for simplicity, and because developmental studies have shown that
there is no discernable difference in collapse performance of structures
design for fractional R factors (e.g., R = 6 versus R = 6.5). The O0 factor is
not assumed initially, but is determined from the actual overstrength factors,
O, calculated for the archetype designs.
9.4.3 Structural System Information
Design Requirements
This example utilizes design requirements for engineered wood light-frame
buildings included in ASCE/SEI 7-05, in place of the requirements that
would need to be developed for a newly proposed system. For the purpose of
assessing uncertainty, the ASCE/SEI 7-05 design requirements are
categorized as (A) Superior since they represent many years of development,
include lessons learned from a number of major earthquakes, and consider
recent results obtained from large research programs on wood light-frame
systems, such as the FEMA-funded CUREE-Caltech Woodframe Project and
the NSF/NEES-funded NEESWood Project.
9-44 9: Example Applications FEMA P695
Test Data
This example relies on existing published sheathing-to-framing connection
test data and wood shear wall assembly test data. Specifically, this example
relies on information developed during the CUREE-Caltech Woodframe
Project (Fonseca et al., 2002; Folz and Filiatrault, 2001), the NEESWood
Project (Ekiert and Hong, 2006), the CoLA wood shear wall test program
(CoLA, 2001), and data provided directly by the wood industry (Line et al.,
2008).
The quality of the test data is an important consideration when quantifying
the uncertainty in the overall collapse assessment process. Cyclic test data
were provided by the wood industry for each of the archetypes used later in
this example. In addition, more data were used by the authors to calibrate
and validate the numerical model; these include monotonic and cyclic tests
which cover a wide range of wood sheathing types and thicknesses (e.g.,
oriented strand board and plywood), framing grades, species, and connector
types (e.g., common vs. box nails). All loading protocols were continued to
deformations large enough for the capping strength to be observed, which
allows better calibration of models for structural collapse assessment.
Nevertheless, some uncertainties still exist with these test data sets including:
(1) premature failures in some of the CUREE data set caused by specimens
with smaller connector edge distances than specified; (2) the use of the
Sequential Phased Displacement, SPD, loading protocol in the CoLA tests
that tends to cause premature specimen failure by connectors fatigue, which
is seldom observed after real earthquakes; (3) the inherent large variability
associated with the material properties of wood; and (4) a lack of duplicate
tests of the same specimen. Therefore, for the purpose of assessing
uncertainty, this test data set is categorized as (B) Good.
9.4.4 Identification of Wood Light-Frame Archetype
Configurations
The archetypes are established according to the requirements of Chapter 4,
and separated into performance groups according to Section 7.4. The first
step in archetype development is to establish the possible building design
configurations. Figure 9-15 shows the two different building configurations
that are assumed to be representative for the purpose of defining the twodimensional
archetypes for wood light-frame shear wall systems with wood
structural panel sheathing. The first configuration is representative of
residential buildings, while the second configuration is associated with
office, retail, educational, and warehouse/light-manufacturing buildings.
FEMA P695 9: Example Applications 9-45
Figure 9-15 Building configurations considered for the definitions of wood
light-frame archetype buildings.
Table 9-18 lists the range of design parameters considered for the
development of the two-dimensional archetype wall models. According to
Section 5.3, two-dimensional archetype wall models, not accounting for
torsional effects, are considered acceptable because the intended use of the
Methodology is to verify the performance of a full class of buildings, rather
than one specific building with a unique torsional issue. According to the
requirements of Section 4.2.3, nonstructural wall finishes, such as stucco and
gypsum wallboard, were not considered in the modeling of the archetypes.
These finishes are excluded because they are not defined as part of the lateral
structural system, and therefore are not governed by the seismic design
Residential building dimensions
Commercial/educational building dimensions
9-46 9: Example Applications FEMA P695
provisions. Depending on their type, wall finishes may greatly influence the
seismic response of wood buildings. The Methodology would allow such
elements to be included in the structural model, if one defines them as part of
the lateral structural system, and design provisions are included to govern
their design.
Table 9-18 Range of Variables Considered for the Definition of Wood Light-
Frame Archetype Buildings
Variable Range
Number of stories 1 to 5
Seismic Design Categories (SDC) Dmax and Dmin
Story height 10 ft
Interior and exterior nonstructural wall finishes Not considered
Wood shear wall pier aspect ratios High/Low
Following the guidelines of Section 4.3, low aspect ratio (1:1 to 1.43:1) and
high aspect ratio (2.70:1 to 3.33:1) walls were used as the two basic
configurations in the archetype designs. This was done to evaluate the
influence of the aspect ratio strength adjustment factor contained in
ASCE/SEI 7-05, which effectively increases the strength of high aspect ratio
wood shear walls.
Table 9-19 shows the performance groups (PG) used to evaluate the wood
light-frame buildings, consistent with the requirements of Section 4.3.1. To
represent these ranges of design parameters, 48 archetypes could have been
used to evaluate the system (three designs for each of the 16 performance
groups shown in Table 9-19). However, Table 9-19 shows that 16
archetypes were found to be sufficient. The notes in the table explain why
these specific archetypes were selected, including the rationale for why these
16 can be used in place of the full set of 48. These 16 wood archetypes were
divided among five performance groups: (1) three low aspect ratio wall
short-period archetypes designed for SDC Dmax (PG-1); (2) five SDC Dmax -
high aspect ratio wall short-period archetypes in SDC Dmax (PG-9); (3) one
low aspect ratio shear wall long-period archetype designed for SDC Dmin
(PG-4); and (4) seven SDC Dmin - high aspect ratio shear wall systems, which
are divided into four short-period buildings (PG-11) and three long-period
buildings (PG-12). It is believed that this ensemble of 16 archetypes covers
the current design space for wood light-frame buildings fairly well, but
additional configurations would be required for a complete application of the
Methodology. Appendix C provides detailed descriptions of the 16
archetype models developed for wood light-frame buildings.
FEMA P695 9: Example Applications 9-47
Table 9-19 Performance Groups Used in the Evaluation of Wood Light-
Frame Buildings
Performance Group Summary
Group No.
Grouping Criteria
Number of
Basic Archetypes
Config.
Design Load Level Period
Gravity Seismic Domain
PG-1
Low Wall
Aspect
Ratio
High
(Nominal)
SDC Dmax
Short 3
PG-2 Long 01
PG-3
SDC Dmin
Short 0
PG-4 Long 12
PG-5
Low
(NA)
SDC Dmax
Short
03
PG-6 Long
PG-7
SDC Dmin
Short
PG-8 Long
PG-9
High
Wall
Aspect
Ratio
High
(Nominal)
SDC Dmax
Short 5
PG-10 Long 01
PG-11
SDC Dmin
Short 4
PG-12 Long 3
PG-13
Low
(NA)
SDC Dmax
Short
PG-14 Long 03
PG-15
SDC Dmin
Short
PG-16 Long
1. No long-period SDC Dmax wood-frame archetypes, because representative
designs never exceed T = 0.6 s.
2. Only one archetype in low-aspect/SDC Dmin/long-period performance group,
because only one representative design exceeds T = 0.4 s.
3. No archetypes because light wood-frame archetype design and performance not
influenced significantly by gravity loads (i.e., nominal gravity loads used for all
designs).
Table 9-20 reports the properties of each of these 16 archetypes. Seismic
demands are based on the ground motion intensities of Seismic Design
Category D. The archetypes are designed for maximum and minimum
seismic criteria of Section 5.2.1: SDS = 1.0 g and SD1 = 0.6 g for SDC Dmax,
and SDS = 0.50 g and SD1 = 0.20 g for SDC Dmin
4. The MCE ground motion
spectral response accelerations, SMT, shown in Table 9-20 are based on Table
6-1. In accordance with Section 5.2.4, the periods reported in Table 9-20 are
the fundamental period of the archetypes based on Section 12.8.2 of
ASCE/SEI 7-05 ( u a T . C T ) with a lower bound limit of 0.25 sec.
4 In this example, archetypes designed for low seismic (SDC Dmin) loads, are based
on SDS = 0.38 g and SD1 = 0.167 g, based on interim criteria which differ slightly
from the final values required by the Methodology.
9-48 9: Example Applications FEMA P695
Table 9-20 Wood Light-Frame Archetype Structural Design Properties
Arch.
ID
No. of
Stories
Key Archetype Design Parameters
Building
Configuration
Wall
Aspect
Ratio
Seismic Design Criteria SMT(T)
[g]
SDC T [sec] T1 [sec] V/W [g]
Performance Group No. PG-1 (Short Period, Low Aspect Ratio)
1 1 Commercial Low Dmax 0.25 0.40 0.167 1.50
5 2 Commercial Low Dmax 0.26 0.46 0.167 1.50
9 3 Commercial Low Dmax 0.36 0.58 0.167 1.50
Performance Group No. PG-9 (Short Period, High Aspect Ratio)
2 1 1&2 Family High Dmax 0.25 0.29 0.167 1.50
6 2 1&2 Family High Dmax 0.26 0.37 0.167 1.50
10 3 Multi-Family High Dmax 0.36 0.44 0.167 1.50
13 4 Multi-Family High Dmax 0.45 0.53 0.167 1.50
15 5 Multi-Family High Dmax 0.53 0.62 0.167 1.50
Partial Performance Group No. PG-4 (Long Period, Low Aspect Ratio)
11 3 Commercial Low Dmin 0.41 0.93 0.063 0.75
Performance Group No. PG-11 (Short Period, High Aspect Ratio)
3 1 Commercial High Dmin 0.25 0.50 0.063 0.75
4 1 1&2 Family High Dmin 0.25 0.41 0.063 0.75
7 2 Commercial High Dmin 0.30 0.61 0.063 0.75
8 2 1&2 Family High Dmin 0.30 0.62 0.063 0.75
Performance Group No. PG-12 (Long Period, High Aspect Ratio)
12 3 Multi-Family High Dmin 0.41 0.69 0.063 0.75
14 4 Multi-Family High Dmin 0.51 0.81 0.063 0.75
16 5 Multi-Family High Dmin 0.60 0.91 0.063 0.75
9.4.5 Nonlinear Model Development
Structural modeling of the wood light-frame archetypes is based on a
“pancake” approach (Isoda et al., 2001). This system-level modeling
approach is capable of simulating the three-dimensional seismic response of
a wood light-frame building through a degenerated two-dimensional planar
analysis. The computer program SAWS: Seismic Analysis of Woodframe
Structures, developed within the CUREE-Caltech Woodframe Project (Folz
and Filiatrault, 2004a, b), was used to analyze the wood light-frame
archetype models. Because this example does not involve any buildings with
torsional irregularities, only a two-dimensional model is utilized by fixing the
rotational degree-of-freedom in the SAWS model.
In the SAWS model, the building structure is composed of rigid horizontal
diaphragms and nonlinear lateral load resisting shear wall elements. The
pinched, strength and stiffness degrading hysteretic behavior of each wood
shear wall in the building is characterized using an associated numerical
FEMA P695 9: Example Applications 9-49
model (Folz and Filiatrault, 2001) that predicts the load-displacement
response of the full wall assemblies under general quasi-static cyclic loading,
based on sheathing-to-framing connection cyclic test data. Alternatively,
cyclic test results from full-scale walls can also be used directly to
characterize their hysteretic response. In the SAWS model, the hysteretic
behavior of each wall panel is represented by an equivalent nonlinear shear
spring element. As shown in Figure 9-16, the hysteretic behavior of this
shear spring includes pinching, as well as stiffness and strength degradation,
and is governed by 10 different physically identifiable parameters (Folz and
Filiatrault 2004a, b). The predictive capabilities of the SAWS program have
been demonstrated by comparing its predictions with the results of shake
table tests performed on full-scale wood light-frame buildings (Folz and
Filiatrault, 2004b; White and Ventura, 2007).
Figure 9-16 Hysteretic model of shear spring element included in SAWS
program (after Folz and Filiatrault, 2004a, b).
Table 9-21 shows the sheathing-to-framing connector hysteretic parameters
used to construct the equivalent nonlinear shear spring elements of each of
the walls contained in the archetype models. The hysteretic model used for
these sheathing-to-framing connectors is the same model used for the entire
wall panel assemblies, which is shown in Figure 9-16. Figure 9-17 shows the
monotonic backbone curves of 8’ x 8’ shear wall specimens generated using
the hysteretic parameters for 8d common nail on 7/16” OSB sheathing shown
in Table 9-21, compared with average test data provided by the wood
industry (Line et al., 2008). This wood industry data is based on the cyclic
envelope averaged from two identical test specimens. Very good agreement
is observed between the numerical predictions and the test data shown here,
9-50 9: Example Applications FEMA P695
which typically continues up to displacements near the onset of strength loss.
The capping displacement and the post-capping behavior of the analytical
model are based on additional cyclic test data that is not shown here, which
was continued to larger displacements to exhibit strength deterioration.
Table 9-21 Sheathing-to-Framing Connector Hysteretic Parameters Used to
Construct Shear Elements for Wood Light-Frame Archetype
Models
Connector
Type
K0
(lbs/in) r1 r2 r3 r4
Fo
(lbs)
FI
(lbs)
.u
(in)
.. ..
7/16" OSB - 8d
common nails
6,643 0.026 -0.039 1.0 0.008 228 32 0.51 0.7 1.2
19/32" Plywood -
10d common nails
7,777 0.031 -0.056 1.1 0.007 235 39 0.49 0.6 1.2
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Deflection (in.)
Load (lbf)
8d at 6"
8d at 2"
8d at 3"
8d at 4"
Numerical Model
Test Data
Figure 9-17 Comparison of monotonic backbone curves of 8’ x 8’ shear wall
specimens generated using the hysteretic parameters for 8d
common nail on 7/16” OSB sheathing shown in Table 9-21 with
average cyclic envelope test data provided by the wood industry
(Line et al.,2008).
Uncertainty due to Model Quality
For the purpose of assessing model uncertainty, according to Section 5.7, the
archetype designs are assumed to be well representative of the archetype
design space, even though a complete assessment may include more basic
structural configurations. The structural modeling approach for the wood
light-frame archetypes captures the primary shear deterioration modes of the
shear walls that precipitate sidesway collapse. However, not all behavioral
aspects are captured by this system-level modeling, such as axial-flexural
FEMA P695 9: Example Applications 9-51
interaction effects of the wall elements, the uplift of narrow wall ends, and
the slippage of sill and top plates. These effects are secondary for walls with
low aspect ratios, which deform mainly in a shear mode, but are important
for archetypes incorporating walls with high aspect ratios. Therefore, the
structural model for the archetypes incorporating low-aspect ratio walls is
rated as (B) Good, while the same structural model for the archetypes
incorporating high-aspect ratio walls is rated as (D) Poor.
9.4.6 Nonlinear Structural Analyses
To compute the system overstrength, O, and to help verify the structural
model, monotonic static pushover analysis is used with an inverted-triangular
lateral load pattern; this approach differs slightly from the final requirements
of Section 6.3. Figure 9-18 shows an example of the pushover curve for the
two-story Archetype ID 5. For the wood light-frame archetype, the design
LRFD seismic coefficient is V/W = 0.167. Capping (the onset of negative
stiffness) occurs for a seismic coefficient of 0.417 and at a roof drift ratio of
0.0229. Therefore, O is calculated to be 2.49 for this archetype model
Figure 9-18 Monotonic static pushover curve and computation of . for twostory
wood light-frame archetype (Archetype ID 5).
Following Section 6.4, to compute the collapse capacity of each wood lightframe
archetype design, the incremental dynamic analysis (IDA) approach is
used with the Far-Field record set and ground motion scaling method
specified in Section 6.2. The intensity of the ground motion causing collapse
of the wood light-frame archetype models is defined as the point on the
intensity-drift IDA plot having a nearly horizontal slope but without
exceeding a peak story drift of 7% in any wall of a model. This collapse
story drift limit was selected based on recent collapse shake table testing
9-52 9: Example Applications FEMA P695
conducted on full-scale two-story wood buildings in Japan (Isoda et al.,
2007). The resulting collapse capacities should not be highly sensitive to this
choice of 7% drift, since the IDA curves are relatively flat at such large drifts
(see Figure 9-19 below).
Figure 9-19 and Figure 9-20 illustrate how the IDA method is used to
compute the collapse margin ratio, CMR, for the two-story Archetype ID 5.
The spectral acceleration at collapse is computed for each of the 44 ground
motions of the Far-Field Set, as shown in Figure 9-19. The collapse fragility
curve can then be constructed from the IDA plots, as shown in Figure 9-20.
The collapse level earthquake spectral acceleration (spectral acceleration
causing collapse in 50% of the analyses) is SCT(T = 0.26 sec) = 2.23 g for this
example. The collapse margin ratio, CMR, of 1.49 is then computed as the
ratio of SCT and the MCE spectral acceleration value at T = 0.26 sec, which is
SMT =1.50 g for this building and SDC.
It should be noted that a full IDA is not required to quantify CMR, as
discussed in Section 6.4.2.
Figure 9-19 Results of Incremental Dynamic Analysis to Collapse for
Two-Story Wood Light-Frame Archetype ID 5.
Figure 9-19 Results of incremental dynamic analysis to collapse for the twostory
wood light-frame archetype (Archetype ID 5).
FEMA P695 9: Example Applications 9-53
0
0.2
0.4
0.6
0.8
1
0 1 2 3 4 5 6 7
Median Sa (T = 0.19 sec) (g)
Collapse Probability
SMT(T = 0.26 s) = 1.50 g
SCT(T = 0.26 s) = 2.23 g
X1.49
CMR = 2.23/1.50 = 1.49
0.26
Figure 9-20 Collapse fragility curve for the two-story wood light-frame
archetype (Archetype ID 5).
Static pushover analyses were conducted and the IDA method was applied to
each of the 16 wood light-frame archetype designs, and Table 9-22
summarizes the results of these analyses. These IDA results indicate that the
average collapse margin ratio is 1.43 for the SDC Dmax, short period – low
aspect ratio archetypes (PG-1), 1.90 for the SDC Dmax, short period – high
aspect ratio archetypes (PG-9), 2.64 for the SDC Dmin, long period – low
aspect ratio archetypes (partial PG-4), 2.57 for the SDC Dmin, short period –
high aspect ratio archetypes (PG-11) and 2.82 for the SDC Dmin, long period
– high aspect ratio archetypes (PG-12). These margin values, however, have
not yet been adjusted for the beneficial effects of spectral shape (according to
Section 7.2). Allowable collapse margins and acceptance criteria are
discussed later.
The results shown in Table 9-22 show that the wood light-frame archetypes
designed for minimum seismic criteria (SDC Dmin) have higher collapse
margin ratios (lower collapse risk) compared with the archetypes designed
for maximum seismic criteria (SDC Dmax). It is believed that this result
originates from the longer vibration periods of archetypes designed for lower
levels of seismic load, since the longer periods reduce seismic demands.
Also, archetypes incorporating walls with high aspect ratios have higher
collapse margin ratios than archetypes with low aspect ratio walls. This is
the result of the ASCE/SEI 7-05 strength reduction factor applied to walls
with high aspect ratios, which cause an increase in required number of nails
to reach a given design strength. This increased nailing density causes an
increase in the shear capacity of the walls with high aspect ratios, but the
model does not account for the associated increase in flexural deformations.
9-54 9: Example Applications FEMA P695
Table 9-22 Summary of Collapse Results for Wood Light-Frame Archetype
Designs
Archetype
ID
Design Configuration Pushover and IDA Results
No. of
Stories
Building
Configuration
Wall
Aspect
Ratio
Static
.
SMT[T]
(g)
SCT T]
(g)
CMR
Performance Group No. PG-1 (Short Period, Low Aspect Ratio)
1 1 Commercial Low 2.0 1.50 2.01 1.34
5 2 Commercial Low 2.5 1.50 2.23 1.49
9 3 Commercial Low 2.0 1.50 2.18 1.45
Mean of Performance Group: 2.2 NA NA 1.43
Performance Group No. PG-9 (Short Period, High Aspect Ratio)
2 1 1&2 Family High 4.1 1.50 2.90 1.94
6 2 1&2 Family High 3.8 1.50 3.20 2.14
10 3 Multi-Family High 3.7 1.50 2.87 1.91
13 4 Multi-Family High 2.9 1.50 2.60 1.73
15 5 Multi-Family High 2.6 1.50 2.67 1.78
Mean of Performance Group: 3.4 NA NA 1.90
Partial Performance Group No. PG-4 (Long Period, Low Aspect Ratio)
11 3 Commercial Low 2.1 0.75 1.98 2.64
Performance Group No. PG-11 (Short Period, High Aspect Ratio)
3 1 Commercial High 3.6 0.75 1.71 2.28
4 1 1&2 Family High 5.4 0.75 2.09 2.78
7 2 Commercial High 4.0 0.75 1.95 2.60
8 2 1&2 Family High 3.5 0.75 1.95 2.60
Mean of Performance Group: 4.1 NA NA 2.57
Performance Group No. PG-12 (Long Period, High Aspect Ratio)
12 3 Multi-Family High 4.0 0.75 2.34 3.12
14 4 Multi-Family High 3.4 0.75 2.09 2.78
16 5 Multi-Family High 3.3 0.75 1.92 2.56
Mean of Performance Group: 3.6 NA NA 2.82
9.4.7 Performance Evaluation
Collapse margin ratios computed above do not account for the unique
spectral shape of rare ground motions. According to Section 7.2, spectral
shape adjustment factors, SSF, must be applied to the CMR results to account
for spectral shape effects. In accordance with Section 7.2.2, the SSF can be
computed for each archetype based on the SDC and the archetypes’ periodbased
ductility, µT, obtained from the pushover curve. Figure 9-21 shows an
example of calculating µT from the pushover curve for the two-story
Archetype ID 5. The period-based ductility, µT, of 7.1 is then computed as
the ratio of the ultimate roof displacement (defined as the displacement at
FEMA P695 9: Example Applications 9-55
80% of the capping strength in the descending branch of the pushover curve)
of .u
= 0.0303hr, to the equivalent yield roof displacement of .y,eff = 0.0043hr.
Figure 9-21 Monotonic static pushover curve and computation of .u
/.y,eff for
the two-story wood light-frame Archetype ID 5.
The adjusted collapse margin ratio, ACMR, is then computed for each wood
light-frame archetype as the multiple of the SSF (from Table 7-1b for SDC
D) and CMR (from Table 9-22). Table 9-23 shows the resulting adjusted
collapse margin ratios for the wood light-frame archetypes.
To calculate acceptable values of the adjusted collapse margin ratio, the total
system uncertainty is needed. Section 7.3.4 provides guidance for this
calculation. Table 7-2 shows these composite uncertainties, which account
for the variability between ground motion records of a given intensity
(defined as a constant .RTR = 0.40), the uncertainty in the nonlinear structural
modeling, the quality of the test data used to calibrate the element models,
and the quality of the structural system design requirements. For this
example assessment, the composite uncertainty was based on a (B) Good
model quality for archetypes with low aspect ratio walls and a (D) Poor for
archetypes with high aspect ratio walls, (A) Superior quality of design
requirements and (B) Good quality of test data. Thus, .... = 0.500 for
archetype buildings incorporating low aspect ratio walls (Table 7-2b) and
.... = 0.675 for archetype buildings incorporating high aspect ratio walls
(Table 7-2d).
An acceptable collapse margin ratio must now be selected based on a
composite uncertainty, .......and a target collapse probability. Section 7.1.2
defines the collapse performance objectives as: (1) a conditional collapse
9-56 9: Example Applications FEMA P695
probability of 20% for all individual wood light-frame archetypes, and (2) a
conditional collapse probability of 10% for the average of each of the
performance groups of wood light-frame archetypes. . Table 7-3 presents
acceptable values of adjusted collapse margin ratio computed assuming a
lognormal distribution of collapse capacity. For archetypes incorporating
low aspect ratio walls, this corresponds to an acceptable collapse margin
ratio ACMR20% of 1.52 for every wood light-frame archetype and an
ACMR10% of 1.90 for each performance group. For archetype buildings
incorporating high aspect ratio walls, this corresponds to an acceptable
collapse margin ratio ACMR20% of 1.76 for every wood light-frame archetype
and an ACMR10% of 2.38 for each performance group.
Table 9-23 Adjusted Collapse Margin Ratios and Acceptable Collapse
Margin Ratios for Wood Light-Frame Archetype Designs
Arch.
ID
Design Configuration
Computed Overstrength and Collapse
Margin Parameters
Acceptance
Check
No. of
Stories
Building
Config.
Wall
Asp.
Ratio
Static
O CMR .T SSF ACMR Accept.
ACMR Pass/ Fail
Performance Group No. PG-1 (Short Period, Low Aspect Ratio)
1 1 Comm. Low 2.0 1.34 9.9 1.33 1.78 1.52 Pass
5 2 Comm. Low 2.5 1.49 7.1 1.31 1.95 1.52 Pass
9 3 Comm. Low 2.0 1.45 12.4 1.33 1.93 1.52 Pass
Mean of Performance Group: 2.2 1.43 9.8 1.32 1.89 1.90 Pass
Performance Group No. PG-9 (Short Period, High Aspect Ratio)
2 1 1&2-F. High 4.1 1.94 9.9 1.33 2.57 1.76 Pass
6 2 1&2-F. High 3.8 2.14 9.6 1.33 2.84 1.76 Pass
10 3 Multi-F. High 3.7 1.91 7.9 1.33 2.54 1.76 Pass
13 4 Multi-F. High 2.9 1.73 5.8 1.28 2.21 1.76 Pass
15 5 Multi-F. High 2.6 1.78 5.4 1.27 2.26 1.76 Pass
Mean of Performance Group: 3.4 1.90 7.7 1.31 2.48 2.38 Pass
Partial Performance Group No. PG-4 (Long Period, Low Aspect Ratio)
11 3 Comm. Low 2.1 2.64 7.0 1.13 2.98 1.52 Pass
Performance Group No. PG-11 (Short Period, High Aspect Ratio)
3 1 Comm. High 3.6 2.28 9.9 1.14 2.58 1.76 Pass
4 1 1&2-F. High 5.4 2.78 9.9 1.14 3.16 1.76 Pass
7 2 Comm. High 4.0 2.60 7.7 1.13 2.95 1.76 Pass
8 2 1&2-F. High 3.5 2.60 7.7 1.13 2.94 1.76 Pass
Mean of Performance Group: 4.1 2.57 8.8 1.13 2.91 2.38 Pass
Performance Group No. PG-12 (Long Period, High Aspect Ratio)
12 3 Multi-F. High 4.0 3.12 7.1 1.13 3.51 1.76 Pass
14 4 Multi-F. High 3.4 2.78 6.2 1.12 3.12 1.76 Pass
16 5 Multi-F. High 3.3 2.56 5.7 1.13 2.90 1.76 Pass
Mean of Performance Group: 3.6 2.82 6.3 1.13 3.18 2.38 Pass
FEMA P695 9: Example Applications 9-57
Table 9-23 summarizes the final results and acceptance criteria for each of
the 16 wood light-frame archetypes. The table presents the collapse margin
ratios computed directly from the collapse fragility curves, CMR, the periodbased
ductility, µT, the Spectral Shape Factor, SSF, and the adjusted collapse
margin ratio, ACMR. The acceptable ACMRs are shown and each archetype
is shown to either pass or fail the acceptance criteria. Average ACMRs are
also shown for the four complete performance groups of archetypes.
The results shown in Table 9-23 show that all individual archetypes pass the
ACMR20% criteria and the averages of each performance group pass the
ACMR10% criteria. Therefore, if wood light-frame buildings were a “newly
proposed” seismic-force-resisting system with R = 6, it would meet the
collapse performance objectives of the Methodology, and would be approved
as a new system.
The results in Table 9-23 also show that performance groups of archetypes
designed for maximum seismic loads (PG-1 and PG-9) have lower adjusted
collapse margins ratios than other groups and govern determination of the R
factor. Another observation is that archetypes incorporating high-aspect ratio
walls have higher collapse margin ratios than those with low-aspect ratio
walls. Even so, acceptable ACMR values are also higher for the high-aspect
ratio wall archetypes, due to higher composite uncertainty.
9.4.8 Calculation of O0 using Set of Archetype Designs
This section determines the value of the overstrength factor, O0, which would
be used in the design provisions for the “newly-proposed” wood light-frame
system. Table 9-23 shows the calculated O values for each of the archetypes,
with a range of values from 2.0 to 5.4. The average values for each
performance group are 2.2, 3.4, 4.1, and 3.6, with the largest value of 4.1
being for the high aspect ratio walls in short-period buildings designed for
low-seismic demands (PG-11).
According to Section 7.6, the largest possible Oo = 3.0 is warranted, due the
average values being greater than 3.0 for three of the performance groups.
9.4.9 Summary Observations
This example shows that current seismic provisions for engineered wood
light-frame construction included in ASCE/SEI 7-05 (with use of R = 6
rather than R = 6.5) to provide an acceptable level of collapse safety. Note
that the collapse safety of actual engineering wood light-frame construction
is most likely higher than calculated in this example because of the beneficial
effects of interior and exterior wall finishes. In accordance with Section
9-58 9: Example Applications FEMA P695
4.2.3, wall finishes were not included in this example because they are not
currently defined as part of the lateral structural system, and therefore are not
governed by the seismic design provisions.
9.5 Example Applications - Summary Observations
and Conclusions
9.5.1 Short-Period Structures
For both reinforced concrete special moment frame and wood light-frame
systems, the short-period archetypes (e.g., T < 0.6 s for SDC Dmax designs)
were those that had the lowest level of collapse performance. For both of
these systems, the short-period performance groups just meet acceptance
criteria, and would fail if the Methodology acceptance criteria were made
stricter. Thus, it is observed that short-period systems need additional
strength (or some other modification that improves the performance) to
achieve a level of collapse performance equivalent to systems with longer
fundamental periods. This finding is not new, but rather has been reported in
research, beginning with Newmark and Hall in 1973. Similar findings have
since been reported by a large number of researchers based on analysis of
single degree of freedom systems (e.g., Lai and Biggs, 1980; Elghadamsi and
Mohraz, 1987; Riddell et al., 1989; Nassar and Krawinkler, 1991; Vidic et
al., 1992; Miranda and Bertero, 1994), and for simple multiple degree of
freedom systems (Takeda et al., 1998; Krawinkler and Zareian, 2007).
The example applications of this Chapter have verified that strength
requirements should be higher for short-period systems, if consistent collapse
performance is desired for all systems regardless of fundamental period.
These strength requirements suggest the use of a period-dependent R factor
as is proposed in many of the referenced papers and reports on this topic.
Currently, the ASCE/SEI 7-05 document utilizes period-independent R
factors. Future work should look more closely at the question of perioddependent
R factors and whether or not they should be considered for use in
future versions of ASCE/SEI 7-05.
9.5.2 Tall Moment Frame Structures
The reinforced concrete special moment frame system example in Section
9.2 found that perimeter frame buildings taller than 12-stories high designed
on the basis of ASCE/SEI 7-05 do not meet the collapse performance
objectives of this Methodology, with the collapse safety worsening with
increasing building height. Tall buildings have more damage localization
and higher P-delta effects, causing this observed trend in performance.
FEMA P695 9: Example Applications 9-59
The issue of worsening collapse safety with increasing building height could
be addressed in various ways. Larger column to beam strength ratios could
be developed for taller buildings, more restrictive drift limits could be
imposed, a period-dependent R factor could be used, or other approaches
could be taken. In the example, minimum base shear requirement of ASCE
7-02 was reintroduced into the design requirements, successfully reversing
the trends and creating increasing collapse safety with increasing building
height.
This information was made available to the ASCE 7 Seismic Committee and
a special code change proposal was passed in 2007 (Supplement No. 2),
amending the minimum base shear requirements of ASCE/SEI 7-05 to
correct this potential deficiency.
9.5.3 Collapse Performance for Different Seismic Design
Categories
Example applications generally found lower collapse safety for buildings
designed in seismic design categories with stronger ground motion intensity.
For example, the ACMR is typically lower for a building designed for SDC
Dmax, as compared to a building design for SDC Dmin. This trend is primarily
caused by the increasing effects of gravity loads for lower levels of seismic
demand, which increases the overstrength of the structural system, and in
turn increases the collapse capacity of the system.
This finding suggests that the R factor will be governed by the SDC with the
strongest ground motion for which the system is proposed. Based on this
observation, the Methodology requires that this SDC with the strongest
ground motion be used when verifying the R factor. It is expected that such
R factors will be conservative for other Seismic Design Categories, but this
trend should be confirmed in the archetype investigation.
FEMA P695 10: Supporting Studies 10-1
Chapter 10
Supporting Studies
This chapter describes additional studies performed in support of the
development of the Methodology. These studies supplement the illustrative
examples presented in Chapter 9, and serve to examine selected aspects of
the Methodology as applied to different seismic-force-resisting systems.
10.1 General
Two supporting studies are presented. One study evaluates a 4-story steel
special moment frame system. This study illustrates the use of component
limit state checks to evaluate failure modes that are not explicitly simulated
in the nonlinear dynamic analysis. It also demonstrates the application of the
Methodology to steel moment frame systems.
A second study assesses the collapse performance of seismically-isolated
systems. This study illustrates application of the Methodology to isolated
structural systems, which have fundamentally different dynamic response
characteristics, design requirements and collapse failure modes than those of
conventional, fixed-base structures. This study also demonstrates the
potential use of the Methodology as a tool for validating and improving
current design requirements, in this case requirements for isolated structures.
10.2 Assessment of Non-Simulated Failure Modes in a
Steel Special Moment Frame System
10.2.1 Overview and Approach
The purpose of this study is to illustrate how component limit state checks
can be used to evaluate failure modes that are not explicitly simulated in the
nonlinear dynamic analysis. This study follows the approach for nonsimulated
collapse modes described in Chapter 5.
The procedure for evaluating non-simulated collapse modes is illustrated
through the evaluation of a steel special moment frame structure, designed
using pre-qualified Reduced Beam Section (RBS) connection details in
accordance with current design standards, ASCE/SEI 7-05 and ANSI/AISC
341-05 Seismic Provisions for Structural Steel Buildings (AISC, 2005). This
study focuses on assessment of a single steel special moment frame building,
10-2 10: Supporting Studies FEMA P695
which in concept could be one of many index archetype configurations
serving to describe the archetype design space. In order to evaluate the entire
class of steel special moment frames, the procedures applied to this
individual building would be extended to the full set of index archetype
models.
The primary collapse mechanism of this steel special moment frame occurs
through hinging in the RBS regions of the beams and the columns, which can
lead to sidesway collapse under large deformations. While gradual
deterioration of the inelastic hinges associated with yielding and local
buckling is simulated in the analyses, sudden strength and stiffness
degradation associated with ductile fractures are not explicitly modeled. In
this study, ductile fracture is not simulated because of software limitations.
The use of separate non-simulated limit state checks is supported by a
number of related factors. First, through the use of pre-qualified RBS
connections, the initiation of ductile fracture is unlikely to occur until large
inelastic rotations have been reached and sidesway collapse has occurred or
nearly occurred. Hence, the simplified limit state check for fracture is not
expected to dominate the results. Second, available test data suggests that
the location where ductile fracture may occur and the deformations at which
ductile fracture may occur are highly variable, and simulation models would
need to define correlations relating fracture probabilities at multiple
connections. In this particular structure, the large columns tend to enforce
equal rotations across a story. Therefore, even if ductile fracture were
modeled, fractures would tend to form simultaneously across a given story
and the collapse results obtained would be similar to those obtained using
non-simulated limit state checks, unless correlations were explicitly
incorporated in the analysis. There is limited data to support estimation of
correlations.
In order to ensure that the collapse assessment process represents the
behavior of the structural system of interest, the choice to incorporate a
particular failure mode using a limit state check, in lieu of direct simulation,
should be based on careful consideration of factors like those described
above. Where non-simulated failure modes dominate the results, or where
their exclusion jeopardizes simulation accuracy before the non-simulated
limit state is reached, the appropriateness of the nonlinear model should be
re-examined.
FEMA P695 10: Supporting Studies 10-3
10.2.2 Structural System Information
The steel special moment frame archetype analyzed in this study is one of
four perimeter moment frames that comprise the seismic-force-resisting
system of a four-story building illustrated in Figure 10-1. The four-bay fourstory
frame provides lateral support to a floor area of 10,800 sq. ft. per floor
and gravity support to a tributary area of 1,800 sq. ft. The seismic weight
(mass) is equal to 940 kips on the second, third and fourth floors and 1,045
kips on the roof, for a total of 3,865 kips per frame.
The building is designed for a high seismic site located in Seismic Design
Category (SDC) D, based on T = 0.94 seconds and a Maximum Considered
Earthquake (MCE) spectral demand, SMT, of 0.96 g (corresponding to SDC
Dmax). The structure has a design base shear, V = 0.08W. Designed in
accordance with ASCE/SEI 7-05 and ANSI/AISC 341-05, beam sizes range
from W24 to W30, and are governed by minimum stiffness requirements
(drift limits). The RBS sections have 45% flange reduction. W24 columns
are sized to satisfy the connection panel zone strength requirements without
the use of web doubler plates. As such, they automatically satisfy other
requirements, including the strong-column weak-beam (SCWB) requirement.
As a result, the actual SCWB ratio is about 2.5 times larger than the required
minimum. This large column overstrength reflects a possible design decision
that is representative of current practice in California; however, it implies
that this study will not necessarily demonstrate the lower-bound performance
of code-conforming steel special moment frames.
Design requirements for this system are well-established, based on
experience in past earthquakes, reflecting a high degree of confidence and
completeness. For the purpose of assessing system uncertainty, the design
requirements are rated (A) Superior.
W30x108
W30x108
x - location of RBS
W24x162
W24x162
W24x207
W24x207
W24x84
W24x84
30 ft.
13 ft.
W30x108
W30x108
x - location of RBS
W24x162
W24x162
W24x207
W24x207
W24x84
W24x84
30 ft.
13 ft.
Figure 10-1 Index archetype model of 4-story steel special moment frame
seismic-force-resisting system
10-4 10: Supporting Studies FEMA P695
10.2.3 Nonlinear Analysis Model
This structure is judged to have primarily two collapse modes: (1) sidesway
collapse associated with beam and column hinging; and (2) collapse
triggered by ductile fracture in one or more RBS connections. Nonlinear
dynamic analyses were conducted using the OpenSees (OpenSees, 2006)
software, employing elements with concentrated inelastic springs to capture
flexural hinging in beams and columns and an inelastic (finite size) joint
model for the beam-column panel zone.
Inelastic springs in beams and columns were modeled using the peakoriented
Ibarra element model (Ibarra et al., 2005), which can capture cyclic
deterioration and in-cycle negative stiffness in elements as the structure
collapses. The monotonic backbone used to model the W24x162 columns is
illustrated in Figure 10-2.
0
0.2
0.4
0.6
0.8
1
1.2
0 0.05 0.1 0.15 0.2 0.25 0.3
M / My
. (rad)
My
.y
Mc
.p .pc
K
. – cyclic deterioration
parameter
Figure 10-2 Monotonic backbone showing calibrated concentrated plasticity
model for a typical column (W24x162).1
Model parameters for beams and columns (e.g., plastic rotation capacity,
cyclic deterioration parameters) were calibrated to experimental test data
(Lignos and Krawinkler, 2007) and reported in Table 10-1. For each beam
and column, the model yield point is defined by the plastic moment capacity
of the section, Mp, calculated with expected values for the steel yield
strength, i.e., 1.1 x Fy. Column initial rotational stiffness is based on
Young’s modulus for steel and the cross-sectional stiffness. Calculation of
beam stiffness includes the contribution of the composite floor slab, and the
1For more information, model parameters are defined in Appendix E. The
discussion in Appendix E deals with the same Ibarra model, but as applied to
modeling of reinforced concrete beams and columns.
FEMA P695 10: Supporting Studies 10-5
additional flexibility in beams due to the RBS section is neglected. RBS
sections are modeled as adjacent to the panel zone, neglecting offsets from
the column face. Beam properties are modeled as asymmetric, depending on
the loading direction.
Table 10-1 Model Parameters for Column and Beam Plastic Hinges in
4-Story Steel Special Moment Frame
Section My (kip-in) Mc/My +(-) .p +(-) .pc
+(-) . K
W24x162 2.5 x104 1.05 0.025 0.35 330 0.4
W24x207 3.3 x104 1.05 0.03 0.30 440 0.4
W24x84 8.2 x103 1.1(1.05) 0.025(0.020) 0.17 380 0.4
W30x108 1.3 x104 1.1(1.05) 0.022(0.016) 0.15 260 0.5
As described previously, these backbones do not predict ductile fracture of
RBS sections, so that failure mode is evaluated through a non-simulated limit
state check. The joint panel zone yield point and hardening parameters are
based on Equation 9-1 in ANSI/AISC 341-05 and Krawinkler (1971, 1978).
The panel zone spring is modeled as non-deteriorating with a bilinear
kinematic hardening model.
Other modeling assumptions are consistent with the requirements of Chapter
5 and Chapter 6. Expected dead and live loads (1.05D + 0.25L) are applied
to the structure and used in the computation of the seismic mass. The
contribution of the gravity frame is not included in the analysis model,
though the leaning P-delta column accounts for gravity loads not tributary to
the seismic-force-resisting system. Foundation flexibility is neglected and
the foundation is modeled as a fixed-base.
The model is rated (B) Good in accordance with Table 5-3. The model is
rated as to how well it captures the behavior of the system up to the point at
which the non-simulated collapse mode (ductile fracture) occurs. This rating
is the same as that given to the reinforced concrete special moment frame
models of Chapter 9, using much of the same rationale. The model is judged
to have a high degree of accuracy and robustness, but does not account for
effects of overturning on column behavior.
The available test data is also rated (B) Good in accordance with Table 3-2.
There is significant test data for steel columns, which has been conducted by
a number of different researchers. However, as with the reinforced concrete
component test data certain critical configurations are missing, such as tests
of steel beams with reinforced concrete slabs.
10-6 10: Supporting Studies FEMA P695
10.2.4 Procedure for Collapse Performance Assessment,
Incorporating Non-Simulated Failure Modes
Accounting for non-simulated failure modes in assessment of collapse
margin ratio first requires the identification and calibration of appropriate
limit-state models. These limit state models are used to evaluate analysis
results to see if the non-simulated limit state was exceeded. The CMR is then
computed to account for both simulated and non-simulated failure modes,
and adjusted for spectral shape effects with the SSF. The resulting ACMR is
compared to the acceptance criteria in Chapter 7.
Identify Non-Simulated Collapse Modes
Properties of beam-column plastic hinges in the analysis model for steel
special moment frames are calibrated to predict hinging and gradual
deterioration associated with yielding and local buckling. However,
experimental data (e.g., Engelhardt et al., 1998; Ricles et al., 2004; Lignos
and Krawinkler, 2007) suggest that the steel frame may also experience
ductile fracture in RBS sections, or possibly at the joint between the beam
and column. For example, in testing done as part of the SAC Steel Project,
Engelhardt et al. (1998) reported fractures in qualifying connections at
inelastic rotations between 0.05 and 0.07 radians. It should be emphasized
that the fracture being considered here is triggered by ductile crack initiation,
and occurs after significant inelastic yielding has occurred, in contrast to the
connection fractures observed in steel frame structures during the 1994
Northridge earthquake.
Ductile-fracture-induced collapse is conservatively assumed to take place if
ductile fracture occurs in any RBS. This collapse limit state is chosen
because the steel special moment frame model is able to capture critical
aspects of system behavior until the point at which ductile fracture occurs in
the first RBS. If the analysis was continued after a ductile fracture is
experienced, the fidelity of the results would be in doubt. Since the strong
column sidesway mechanism in this frame imposes similar peak rotations in
all the RBS hinges, the assumption of equating the first instance of a
fractured connection with fracture-induced collapse may not be too
unreasonable (overly conservative) for this particular example.
Generally, it would be desirable to incorporate fracture deterioration directly
in the nonlinear model. Since it is not directly incorporated in this analysis, a
conservative judgment is made about what constitutes collapse for the
fracture limit state. A typical representation of the deformations at fracture is
shown in Figure 10-3.
FEMA P695 10: Supporting Studies 10-7
Develop Component Fragility
Calculation of the non-simulated collapse mode requires the definition of a
fracture fragility function. In this case, the fragility function relates the
probability of ductile fracture in the RBS hinge to a response parameter such
as the maximum plastic hinge rotation that has occurred in the plastic hinge.
Figure 10-4 shows the resulting fragility function, P[Fracture|.p], which is
based on available data (Lignos and Krawinkler, 2007) and engineering
judgment.
Figure 10-3 Illustration of fracture behavior (Engelhardt et al., 1998).
Figure 10-4 Component fragility function, describing probability of ductile
fracture occurring as a function of the plastic rotation, .p.
The fragility function for fracture is assumed to follow a lognormal
distribution, and has a median capacity of p
ˆ.
= 0.063 radians and a
logarithmic standard deviation of .F = 0.35. The dispersion, .F, reflects both
test data statistics (from 10 tests) and judgment as to the additional variability
that may be encountered in actual buildings. Assuming that the parameters
10-8 10: Supporting Studies FEMA P695
associated with fracture are the same throughout the building, the fragility
function is applicable to every RBS in the building.
Identify Collapse Limit based on Component Fragility
The collapse limit point for the non-simulated collapse mode is defined by
the median value of the fragility function associated with component failure.
Therefore, if the plastic rotation in a RBS exceeds the median of 0.063
radians, that component is assumed to be fractured and the non-simulated
limit state is triggered.
Note that the collapse limit state in this formulation ignores the effect of
dispersion, .F, in the collapse fragility. The magnitude of the impact of .F
on the total collapse uncertainty, .TOT, depends on the relative dominance of
simulated (sidesway) and non-simulated failure modes. However, the other
sources of uncertainty considered in the Methodology (e.g., .RTR, .DR, .TD,
.MDL) dominate the uncertainty in the collapse fragility. For explanation of a
rigorous approach accounting for the effects of .F , see Aslani (2005).
Nonlinear Static (Pushover) Analysis
The static pushover analysis results for the steel special moment frame are
illustrated in Figure 10-5. The maximum base shear, Vmax, is 1170 kips,
compared to a design base shear of approximately 345 kips, for an
overstrength, O = 3.4. This large overstrength relative to the design lateral
forces is due to the very strong columns sized to avoid doubler plates.
The period-based ductility, given by µT = du/dy,eff, is obtained from the
pushover analysis as described in Section 6.3. The ultimate roof
displacement, du, is taken as either the roof displacement corresponding to a
20% loss in base shear, or the roof displacement at which the non-simulated
(ductile fracture) failure mode occurs. Since the strong columns in this
structure impose approximately uniform distribution of story drift, the roof
displacement at which the non-simulated (fracture) failure mode occurs is
approximately 0.063hr, corresponding to a roof displacement of 39.3 inches.
For comparison, du associated with a 20% loss in base shear from pushover
analysis, is approximately 0.075hr, corresponding to a roof displacement of
approximately 46.8 inches. The effective yield displacement, dy,eff, is
computed from Equation 6-7 as 0.009hr, or 5.6 inches, for this structure.
Therefore, µT is 7.0.
FEMA P695 10: Supporting Studies 10-9
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
0
200
400
600
800
1000
1200
Base Shear (kips)
d Roof Drift Ratio du = 0.063hr y,eff = 0.009hr
Vmax
~0.8Vmax
du ˜ 0.075hr
(non-simulated modes
not considered)
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
0
200
400
600
800
1000
1200
Base Shear (kips)
d Roof Drift Ratio du = 0.063hr y,eff = 0.009hr
Vmax
~0.8Vmax
du ˜ 0.075hr
(non-simulated modes
not considered)
Figure 10-5 Results of nonlinear static (pushover) analysis of a steel special
moment frame, illustrating computation of period-based
ductility from non-simulated collapse modes.
Assess Collapse Performance
To assess collapse performance, response of the 4-story steel special moment
frame is calculated using the Far-Field record set and nonlinear dynamic
analysis, as described in Chapters 5 through 7. If only simulated collapse
modes are considered, a median collapse capacity, CT Sˆ
, of 2.36 g is obtained.
Due to the strength of the columns relative to the beams, the sidesway
collapse mode is a full four-story mechanism with hinges at all RBS regions
and at the fixed column bases for all ground motions.
The collapse fragility is computed based on both simulated and nonsimulated
collapse modes. This procedure is illustrated graphically in Figure
10-6, using representative curves from incremental dynamic analysis for this
structure (each curve contains the dynamic analysis results for one ground
motion, scaled until collapse). The non-simulated fracture mode occurs if the
plastic rotations at any RBS in the building exceed the median capacity
defined by the component fragility function. In the figure, sidesway collapse
points based on results from simulation, SCT(SC), are shown with short bold
pointers. Non-simulated (fracture-induced) collapse points, SCT(NSC), are
shown with pointers from hatched circles corresponding to maximum story
drift at fracture ( p
ˆ.
= 0.063 radians).
For the purposes of this figure, the component collapse limit of p
ˆ.
= 0.063 is
shown as approximately equal to the story drift ratio. This assumption is
10-10 10: Supporting Studies FEMA P695
made for illustration purposes only, and is not actually used in computing the
occurrence of non-simulated failure modes. Of the three curves shown in
Figure 10-6, the lowest reaches simulated sidesway collapse and the nonsimulated
collapse limit state at approximately the same level of ground
motion intensity. The other curves reach the non-simulated collapse point
under less intense ground motions than suggested by sidesway collapse.
Investigation of the occurrence of simulated and non-simulated failure modes
is repeated for all ground motions. In computing the collapse fragility, the
more critical of these two limit states is taken as the governing collapse point
for each ground motion record.
Figure 10-6 Selected simulation results for the steel special moment frame
illustrating the identification of non-simulated collapse modes.
The governing collapse point for each ground motion record is
identified with an asterisk (*).
Given ratings of (B) Good for modeling, (A) Superior for design
requirements, and (B) Good for test data, the total system collapse
uncertainty is .TOT = 0.500. Since µT . 3 for this building, this value includes
record-to-record uncertainty, .RTR = 0.40, in accordance with Section 7.3.4.
It is noted that for systems driven by very brittle non-simulated collapse
modes, .RTR can potentially be reduced in accordance with Equation 7-2.
This reduction should be exercised with caution. The approach to nonsimulated
failure modes is based only on the median component limit state,
neglecting the underlying uncertainty associated with the occurrence of the
component failure mode, .F. If record-to-record variability is significantly
reduced with Equation 7-2, the total uncertainty in the collapse fragility
could be too low, and non-conservative.
FEMA P695 10: Supporting Studies 10-11
The combined collapse fragility is illustrated in Figure 10-7. In this figure,
the horizontal axis fragility parameter, SCT, is normalized by MCE demand,
SMT to permit direct comparison of the collapse margin ratio, CMR, for the
structure with and without consideration of non-simulated fracture-induced
failure modes. For this structure, the net result of including the fractureinduced
collapse is a 32% reduction in the collapse margin ratio, CMR, from
2.5 for the simulated sidesway-only case, to 1.9. The conditional probability
of collapse at the MCE increases from 8% to 14%. (Note: these margins and
the collapse probabilities do not include the spectral shape factor, which is
considered in the evaluation of acceptance criteria.)
0
0.25
0.5
0.75
1
0 1 2 3 4 5
P[Collapse]
SCT/SMT
Collapse Fragility (Simulated & Non-Simulated
Collapse)
Simulated (Sidesway) Collapse Fragility
Reduction in CMR due to
non-simulated failure
modes
Increase in probability of collapse
due to non-simulated failure modes
Figure 10-7 Comparison of steel special moment frame collapse fragilities for
sidesway-only and combined simulated and non-simulated
(sidesway and fracture-induced) collapse (not adjusted for
spectral shape).
Compare to Acceptance Criteria
Finally, the combined fragility data, reflecting the likelihood of both
simulated and non-simulated collapse, should be compared to the acceptance
criteria in Chapter 7. For this structure, the spectral shape factor, SSF, of
1.41 is determined from Table 7-1b with T = 0.94 seconds and du/dy,eff = 7.0,
which gives an ACMR of 2.7. The acceptable ACMR is obtained from Table
7-3, with..TOT = 0.500. The ACMR is compared to an acceptable collapse
margin ratio, ACMR10%, of 1.90 (based on the 10% probability of collapse
limit), and easily satisfies the acceptance criteria.
10-12 10: Supporting Studies FEMA P695
If the Methodology were applied to a complete set of steel special moment
frames, including different performance groups for high and low gravity
loads, maximum and minimum seismic criteria and short-period and long
period systems, each individual archetype would be compared to ACMR20%,
and the average of each performance group would be compared to ACMR10%
from Table 7-3.
10.3 Collapse Evaluation of Seismically Isolated
Structures
10.3.1 Introduction
Seismic isolation (commonly known as base isolation) is a technology that is
intended to protect facility function and provide substantially greater damage
control than conventional, fixed-base, structures for moderate and strong
earthquake ground motions. For extreme ground motions, seismically
isolated structures are expected to be at least as safe against collapse as their
conventional counterparts. To ensure adequate performance, ASCE/SEI
7-05 requires explicit evaluation of the design of every isolated structure
under Maximum Considered Earthquake (MCE) ground motions and
comprehensive testing of prototype isolator units to verify design properties
and demonstrate stability under MCE loads.
The provisions of ASCE/SEI 7-05 require the seismic force-resisting system
of the structure above the isolation system (superstructure) to be designed for
response modification factors, RI, that are smaller than the R factors
permitted for conventional structures. Reduced values of the response
modification factor are intended to keep the superstructure “essentially
elastic” for design earthquake ground motions. To protect against potential
brittle failure for extreme ground motions, ASCE/SEI 7-05 requires the
superstructure to have the same ductile capacity as that required for a
conventional structure of the same type in the seismic design category of
interest. The provisions of ASCE/SEI 7-05 are generally considered to be
conservative with respect to design of the superstructure, although the degree
of conservatism, if any, is not known.
Objectives
This study is intended to demonstrate the application of the Methodology to
isolated structures, which have fundamentally different dynamic response
characteristics, performance properties and collapse failure modes than those
of conventional, fixed-base structures. Special issues include the following:
Period Definition. The fundamental-mode “effective” period of an
isolated structure is based on secant stiffness at the response
FEMA P695 10: Supporting Studies 10-13
amplitude of interest, rather than “elastic” stiffness used to define the
period, T, of conventional structures.
Record-to-Record (RTR) Variability. Record-to-record variability
may be smaller than 0.4 because base isolated structures typically do
not undergo large period elongation before collapse. Are the
recommendations for reduced record-to-record (RTR) variability in
Chapter 7 suitable for isolated structures?
Test Data and Modeling Uncertainty. Can the collapse margin ratio
(CMR) of isolated structures be evaluated using the same uncertainty
associated with test data and modeling of the superstructure as
considered for a conventional structure with the same type of
seismic-force-resisting system?
Spectral Shape Factor (SSF). Can the spectral shape factor (SSF)
used to adjust the CMR of isolated structures be calculated using the
same methods as those specified for conventional structures
(Appendix B)?
This study is also intended to illustrate how the Methodology can be used as
a tool for assessing the validity of current design requirements, in this case
Chapter 17 of ASCE/SEI 7-05, and to develop improved code provisions. In
order to evaluate the design requirements for isolated structures, this study
specifically explores the sensitivity of collapse performance to the following
key design properties of isolated structures:
Superstructure Strength. How does collapse performance vary for
superstructures that have different design strength levels (e.g.,
superstructures designed for different effective values of the RI
factor)? Effective RI values yielding higher and lower design base
shears than the code-specified values are considered.
Superstructure Ductility. How does superstructure ductility
influence collapse performance (e.g., performance of special moment
frame superstructures as compared to that of ordinary moment frame
superstructures)?
Moat Wall Clearance. How does collapse performance vary for
isolated structures that have different amounts of clearance between
the isolated structure and the moat wall?
10-14 10: Supporting Studies FEMA P695
Scope and Approach
The scope of this study is necessarily limited and relies on archetypical
models available from other examples developed in this project to represent
the superstructures of isolated archetypes. Specifically, superstructures are
based on the 2-dimensional, archetypical models of 4-story reinforced
concrete special moment frame and ordinary moment frame systems, from
the example applications included in Chapter 9. Design of isolator systems
in this study varies from typical isolator design in that isolator properties are
designed to satisfy ASCE/SEI 7-05 design requirements between the
isolation system and the superstructure, given the reinforced concrete frame
superstructures developed in the Chapter 9 examples.
The archetypical models of isolated structures incorporate force-deflection
properties of isolation systems typical of actual projects that use either (1)
elastomeric, rubber bearings (RB), or (2) sliding, friction-pendulum (FP)
bearings. These two isolation system types are designed and archetypical
models of the isolated structure evaluated for maximum and minimum SDC
D ground motions (SDC Dmax and SDC Dmin). SDC Dmax ground motions are
typical of those used for most seismic isolation projects (e.g., projects in high
seismic regions of coastal California).
This study includes discussion of background information necessary for
proper application of the Methodology to isolated structures and related
development of isolated archetypes. Archetype configurations, nonlinear
analysis techniques, and collapse performance methods are described with
reference to specific differences in applying the methodology to isolated
structures. Collapse evaluation results are reported for archetypes that
comply with the design requirements of ASCE/SEI 7-05 (referred to herein
as Code-Conforming archetypes) and for archetypes that deviate from current
requirements (Non-Code-Conforming archetypes). The Non-Code-
Conforming archetypes demonstrate potential use of the Methodology as a
tool for code development, by evaluating collapse performance for archetype
models that have weaker (or stronger) superstructures, less ductility, or
different moat clearances than those specified in ASCE/SEI 7-05.
10.3.2 Isolator and Structural System Information
Archetypes of isolated structures must be designed using established design
requirements, and modeling of isolated archetypes must be supported by
appropriate test data (Chapter 3). Further, the quality of the design
requirements and test data must be rated for establishing system collapse
uncertainty (Chapter 7).
FEMA P695 10: Supporting Studies 10-15
Design Requirements
Archetypes of isolated structures are designed according to the provisions of
ASCE/SEI 7-05 and related superstructure design codes, including ACI 318-
05, except as some specific provisions are modified, or ignored, to evaluate
the effects of reduced superstructure strength, or limited ductility, or moat
clearance in the Non-Code-Conforming archetypes. Chapter 17 of
ASCE/SEI 7-05 requires thorough and rigorous design of the isolated
structure, including explicit evaluation of the isolation system for MCE
ground motions, and peer review.
For the purpose of assessing the composite uncertainty in the Methodology,
the isolation system and superstructure design requirements are rated as (A)
Superior, as they are thorough, detailed and vetted through the building code
process.
Test Data
The requirements for test data relate both to testing of superstructure
components (i.e., reinforced concrete beams, columns and connections), and
testing of prototype isolator units. The test data related to reinforced
concrete elements is rated (B) Good, as discussed in the Chapter 9 examples.
While there is a large amount of test data on reinforced concrete components,
there are still several areas where test data are not complete (i.e. tests of
beams with slabs, tests to very large deformations).
Test data related to isolation units is both qualitatively and quantitatively
different. For the purposes of modeling conventional structures, test data is
taken from a variety of different researchers regarding components which are
similar, but not identical, to the components being modeled in the structure.
In contrast, the provisions ASCE/SEI 7-05 require prototype testing of
isolator units for the purpose of establishing and validating the design
properties of the isolation system and verifying stability for MCE response.
These tests are specific to the isolation system installed in a particular
building, and follow detailed requirements for force-deflection response
outlined in the design requirements. As a result, there is substantially smaller
uncertainty related to the test data in an isolated system than the
superstructure.
A rating of (B) Good is assigned to the uncertainty assessment for test data
for these systems. This rating is associated with uncertainty in test data
related to superstructure modeling and potential isolator failure modes at
loads and displacements greater than those required for isolator prototype
testing by Section 17.8 of ASCE/SEI 7-05.
10-16 10: Supporting Studies FEMA P695
10.3.3 Modeling Isolated Structure Archetypes
This section identifies the basic configuration, systems and elements of the
isolated structures of this study and provides a general overview of the index
archetype models used to evaluate collapse performance of these systems.
Section 10.3.4 describes specific design properties for the isolation system,
moat wall clearance, and the superstructure of each model.
The seismic-force-resisting system of an isolated structure includes: (1) the
isolation system; and (2) the seismic-force-resisting system of the
superstructure above the isolation system. The isolation system includes
individual isolator units (e.g., elastomeric or sliding bearings), structural
elements that transfer seismic force between elements of the isolation system
(e.g., beams just above isolators) and connections to other structural
elements. Energy dissipation devices (dampers) are sometimes used to
supplement damping of isolator units, but such devices are not considered in
this study.
Isolated structures are typically low-rise or mid-rise buildings that have
relatively stiff superstructures. This study assumes that the seismic-forceresisting
system of the superstructure is a reinforced-concrete moment frame
system, in order to make use of already developed models for Chapter 9 that
explicitly capture sidesway collapse. A relatively short height (4 stories) is
used to assure adequate stiffness of the superstructure.
Isolated structures typically have a “moat” around all or part of the perimeter
of the building at the ground floor level. The moat is usually covered by
architectural components (e.g., cover plates) that permit access to the
building, but do not inhibit lateral earthquake displacement of the isolated
structure.
Impact with the moat wall can cause collapse. Accordingly, Section 17.2.5.2
of ASCE/SEI 7-05 requires a minimum building separation to “retaining
walls and other fixed obstructions” not less than the total maximum (MCE)
displacement of the isolation system. The intent of this provision is to limit
the likelihood of such impacts, even for strong ground motions. However,
Section 17.2.4.5 of ASCE/SEI 7-05 recognizes that providing clearance for
MCE ground motions may not be practical for some systems and permits
isolation system design to incorporate a “displacement restraint” that would
limit MCE displacement, provided certain criteria are met. These criteria
include the requirement that “the structure above the isolation system is
checked for stability and ductility demand of the maximum considered
earthquake.” This study shows how the Methodology can be used to perform
this check.
FEMA P695 10: Supporting Studies 10-17
Index Archetype Models
Models of isolated systems consist of a superstructure model, an isolator
model, and a moat wall model. Each of these components must be capable of
capturing the inelastic effects in the structure up until the point at which the
structure collapses. Analysis models are two-dimensional, and neglect
possible torsional effects. A schematic diagram of the index archetype model
for isolated systems is illustrated in Figure 10-8.
W includes weight of ground floor
(floor just above isolators)
assumed to be the same as 1st
floor of fixed-base archetype
W includes weight of ground floor
(floor just above isolators)
assumed to be the same as 1st
floor of fixed-base archetype
Moat wall spring
Bilinear spring
representing base
isolator
Superstructure
Model
Grade beams (linear
W, includes weight of elastic)
ground floor
W includes weight of ground floor
(floor just above isolators)
assumed to be the same as 1st
floor of fixed-base archetype
W includes weight of ground floor
(floor just above isolators)
assumed to be the same as 1st
floor of fixed-base archetype
Moat wall spring
Bilinear spring
representing base
isolator
Superstructure
Model
Grade beams (linear
W, includes weight of elastic)
ground floor
Figure 10-8 Index archetype model for isolated systems.
The superstructure is modeled using the same assumptions as described in
the Chapter 9 examples. The model incorporates material nonlinearities in
beams, columns and beam-column joints, as well as deterioration of strength
and stiffness as the structure becomes damaged. As before, the
superstructure model includes a leaning (P-delta) column to account for the
effect of the seismic mass on the gravity system.
Isolator Modeling
The isolation system bearings (isolators) are modeled using a bilinear spring
between the foundation and the ground-floor of the superstructure, as shown
in Figure 10-8. The bilinear spring is assumed to be a non-degrading, fully
hysteretic element. Some isolators exhibit significant changes in properties
with repeated cycles of loading as a result of heating, but explicit modeling
of such behavior is currently in its infancy and is beyond the scope of this
study. Bilinear springs are commonly used in practice and provide
sufficiently accurate estimates of nonlinear response when their stiffness and
damping properties are selected to match those of the isolators (Kircher,
2006).
10-18 10: Supporting Studies FEMA P695
Figure 10-9 illustrates the modeled force-displacement response of nominal,
upper-bound and lower-bound bilinear springs used to represent isolators in
this study. In general, isolators are modeled with nominal spring properties.
However, in certain cases, isolators are modeled with upper-bound and
lower-bound spring properties to evaluate the effects of these properties on
collapse. Upper-bound and lower-bound properties are also needed for
isolation system design. Section 17.5 of ASCE/SEI 7-05 requires lowerbound
properties for calculation of isolation system displacements, and
upper-bound properties for calculation of design forces. The range of upperbound
and lower-bound bilinear spring properties used to model isolators is
based on the prototype testing acceptance criteria of Section 17.8.4 of
ASCE/SEI 7-05.
D
D
A
Ay
A
Displacement (in.)
Force/W (g) Upper-Bound
Nominal
Lower-Bound
Hysteresis Loop
(Nominal Spring)
Dy
D
D
A
Ay
A
Displacement (in.)
Force/W (g) Upper-Bound
Nominal
Lower-Bound
Hysteresis Loop
(Nominal Spring)
Dy
Figure 10-9 Example nominal, upper-bound and lower-bound bilinear
springs and hysteretic properties used to model the isolation
system.
Design of the isolation system is defined by the two control points, the
“yield” point (Dy, Ay) and the post-yield point (D, A), located somewhere on
the yielded portion of the curve. The resulting bilinear response is illustrated
in Figure 10-9. The yield point represents the dynamic friction level of
sliding bearings (e.g., FP bearings) and is closely related to the (normalized)
characteristic strength of elastomeric bearings (e.g., lead-rubber or highdamping
rubber bearings). The post-yield point is used simply to define the
slope of yielded system (and the isolation system is assumed capable of
displacing without failure beyond this point). The properties are assumed to
be symmetric for positive and negative displacements.
Amplitude-dependent values of effective stiffness, keff, effective period, Teff
(in seconds), and effective damping, .eff, of the isolation system may be
calculated by the following equations:
FEMA P695 10: Supporting Studies 10-19
D
k AW eff . (10-1)
gA
T D eff . 2. (10-2)
D A
A D D A y y
eff
.
.
.
. 2
(10-3)
Equations 10-1, 10-2, and 10-3 are consistent with the definitions of effective
stiffness, keff, and effective damping, .eff, of Section 17.8.5 of ASCE/SEI 7-
05. In Section 17.8.5, these equations are used to determine the forcedeflection
characteristics of the isolation system from the tests of prototype
isolators.
ASCE/SEI 7-05 defines two amplitude-dependent fundamental-mode periods
for isolated structures based on secant stiffness, TD, (effective period at the
design displacement), and TM, (effective period at the MCE displacement).
Values of TD and TM are typically close together. Isolated systems are
initially stiff (see Figure 10-9), but yield early and become very flexible, so
this representation of period is different from conventional fixed-base
structures, which use an estimation of initial stiffness for an “elastic” period.
The MCE fundamental-mode period, TM, as defined by ASCE/SEI 7-05, is
used for evaluation of seismic collapse performance of isolated structures in
this study:
k g
T W
Mmin
M . 2. (10-4)
where:
TM = effective period, in seconds, of the seismically isolated structure at
the maximum displacement in the direction under consideration, as
prescribed by Equation 17.5-4 of ASCE/SEI 7-05,
W = effective seismic weight of the structure above the isolation interface,
as defined in Section 17.5.3.4 of ASCE/SEI 7-05, and
kMmin = minimum effective stiffness, in kips/in, of the isolation system at the
maximum displacement in the horizontal direction under
consideration, as prescribed by Equation 17.8-6 of ASCE/SEI 7-05.
10-20 10: Supporting Studies FEMA P695
Moat Wall Modeling
Nonlinear springs are used to represent the effects of impact with the moat
wall when the seismic demand on the isolated system exceeds the clearance
provided. The moat wall is represented by 5 symmetrical gap springs
implemented in parallel as illustrated in Figure 10-10.
2
1
3
4
5
2
1
3
4
5
Moat Wall Gap
Moat Wall Gap
Isolation System Displacement (in.)
10
-10
20
-40 -30 -20
30 40
-Vmax
Vmax
Moat Wall Force
Composite Spring
Energy Loss
(Wall Crushing)
2
1
3
4
5
2
1
3
4
5
Moat Wall Gap
Moat Wall Gap
Isolation System Displacement (in.)
10
-10
20
-40 -30 -20
30 40
-Vmax
Vmax
Moat Wall Force
Composite Spring
Energy Loss
(Wall Crushing)
2
1
3
4
5
2
1
3
4
5
Moat Wall Gap
Moat Wall Gap
Isolation System Displacement (in.)
10
-10
20
-40 -30 -20
30 40
-Vmax
Vmax
Moat Wall Force
Composite Spring
Energy Loss
(Wall Crushing)
Figure 10-10 Five individual springs and effective composite spring used to
model moat wall resistance for an example 30-inch moat wall
gap.
Gap springs have zero force until the isolated structure reaches the moat wall,
and then begin to resist further displacement of the isolated structure. The
five springs engage sequentially to effect increasing stiffness and nonlinear
resistance as the structure pushes into the moat wall. Gap springs are
modeled as inelastic elements to account for energy loss due to localized
crushing at the structure-wall interface. Gap spring properties are defined
relative to the strength of the superstructure such that moat wall force is
equal to the strength of the superstructure, Vmax, at approximately 4 inches of
moat wall displacement. (Note. A similar result could have been obtained
using a more complicated force-displacement relationship and a single gap
spring, depending on available models in software.)
Model Limitations
It should be noted that the models used in this study, and illustrated in Figure
10-9 and Figure 10-10, are relatively simplistic representations of isolated
systems and neglect many complex aspects of isolator performance. Bilinear
properties of the isolator neglect stiffening that may occur at large
displacements, and reduction in characteristic strength that can occur for
repeated cycles of loading. The isolation system model also neglects
potential uplift at isolators (when overturning loads exceed gravity loads and
isolator tension capacity). Local uplift of isolators is a potential failure
mode, but permitted by Section 17.2.4.7 of ASCE/SEI 7-05 provided that
“resulting deflections do not cause overstress or instability of isolator units or
FEMA P695 10: Supporting Studies 10-21
other structure elements.” Softening and damping loss due to heat effects in
elastomeric bearings are not considered. Despite these limitations, the
models represented in Figures 10-8, 10-9, and 10-10 are expected to give
reasonable predictions of dynamic response and collapse performance, as
well as relative differences in performance associated with variability in
design parameters.
Uncertainty due to Model Quality
Model quality is rated as (B) Good for the purpose of assessing the
composite uncertainty in the performance predictions of index archetype
models with a reinforced concrete special moment frame superstructure, and
(C) Fair for index archetype models with a reinforced concrete ordinary
moment frame superstructure. These model quality ratings are the same as
those assigned to the reinforced concrete frame examples in Chapter 9.
Since design of isolated structures require a nonlinear, building-specific
model of the superstructure, the model of an isolated structure represents an
individual building rather than one archetype within a performance group, as
in the Chapter 9 examples. In addition, the isolation system filters out some
higher-mode effects that contribute to collapse variability in fixed-base
structures. Therefore, it may be possible for an isolated system to earn a
better model rating than the constituent superstructure. However, in this
study, the simplified modeling of isolators, such as neglecting cyclic changes
to bearing properties, does not warrant better model quality ratings.
10.3.4 Design Properties of Isolated Structure Archetypes
Specific design properties of index archetype models in this study include
isolation system properties, moat wall clearance and superstructure
properties. The archetypes are selected to probe the effects of the critical
design parameters on collapse performance. The goal of considering a wide
variety of archetype configurations is to assess the validity of current code
provisions. As such, archetypes are developed for both “Code-Conforming”
systems that comply with ASCE/SEI 7-05 requirements, and “Non-Code-
Conforming” systems that deviate from ASCE/SEI 7-05 requirements in
terms of superstructure strength or ductility.
Isolation System Design Properties
Design properties of the isolation system are developed using the equations
and design requirements of the Equivalent Lateral Force Procedure, Section
17.5 of ASCE/SEI 7-05. These equations provide a convenient basis for
design and are commonly used for preliminary design and review of isolated
structures.
10-22 10: Supporting Studies FEMA P695
This study considers archetype isolation systems that are representative of
systems with either elastomeric rubber bearings (RB) or sliding friction
pendulum (FP) bearings. Isolation systems are designed for either SDC Dmax
or SDC Dmin seismic criteria (Site Class D). The response characteristics of
isolation systems with either RB or FP bearings are sufficiently similar for
strong (SDC Dmax) ground motions to permit modeling both systems with the
same set of bilinear springs properties (i.e., a single set of “generic”
properties is used to represent both systems). Such is not the case for
moderate (SDC Dmin) ground motions, and different spring properties are
used to model isolation systems with RB and FP bearings.
Nominal isolation system design properties for SDC Dmax are given in Table
10-2 for the generic (GEN) system, and the RB system and FP system for
SDC Dmin. Upper and lower-bound spring properties are also given for the
generic system. Only nominal properties are shown for the RB and FP
systems in SDC Dmin (although MCE design parameters still utilize upperbound
and lower-bound properties, as required by Section 17.5 of ASCE/SEI
7-05).
Table 10-2 also provides design values of the effective period, TM, effective
damping, .M, and total maximum displacement, DTM, for each system. For
each archetype isolation system, design values of the yield and the post-yield
control points are selected such that the corresponding values of maximum
(MCE) displacement, DM, effective period, TM, effective stiffness, and
effective damping, .M, meet the following criteria:
. Values of DM, TM and .M comply fully with the equations and
requirements of Section 17.5 of ASCE/SEI 7-05.
. Values of effective stiffness and damping are consistent with actual
isolation system properties (i.e. for isolations systems with either
elastomeric or sliding bearings)2.
. Values of effective stiffness and damping in the isolator reduce response
such that forces required for design of the superstructure are
approximately the same as the design base shear required for a
conventional fixed-base system of the same type and configuration, so
that models of code-compliant systems from Chapter 9 can be used for
the superstructure (without re-design).
2 The effective stiffness properties of rubber bearings represent a “low modulus”
rubber compound and are assumed to be the same as those of friction pendulum
bearings to limit the number of models. The corresponding effective period, TM, is
somewhat atypical of rubber bearing systems which generally have an effective
period less than or equal to 3 seconds.
FEMA P695 10: Supporting Studies 10-23
Calculation of DTM was based on the Equivalent Lateral Procedure of Section
17.5 of ASCE/SEI 7-05. Total maximum displacement includes an
additional 15 percent of torsional displacement, consistent with Equation
17.5-6 of ASCE/SEI 7-05 (assuming a square configuration of the building in
plan):
TM M D . 1.15D (10-5)
Table 10-2 Isolation System Design Properties
Isolator
Properties
Force-Deflection Curve MCE Design Parameters
Yield Point Post-Yield TM
(sec)
ßM
(% crit.)
DTM
Type Range D (in.) y (in) Ay (g) D (in) A (g)
Generic Elastomeric or Sliding Systems - Dmax Designs
GEN Nominal 0.5 0.05 23.3 0.225
3.47 10.5% 29.3
GENUB
Upper-
Bound
0.5 0.06 21.1 0.23
GENLB
Lower-
Bound
0.5 0.04 25.5 0.22
Rubber (RB) or Friction Pendulum (FP) Systems - Dmin Designs
RB Nominal 1.5 0.04 6.5 0.081 3.18 12.5% 8.5
FP Nominal 0.1 0.04 4.2 0.072 2.78 28% 5.7
Isolation System Clearance
The performance of isolated structures may also be dependent on the
clearance of the isolated system and Table 10-3 summarizes the moat wall
clearance (gap) distances used in the archetype isolation systems. For the
generic (GEN) isolation system (SDC Dmax design), five different moat wall
distances are used to investigate the effects of this parameter on collapse
performance.
Moat wall clearance is based on a fraction of the total maximum
displacement, DTM, plus a little extra displacement for fit-up tolerance.
Unless dynamic analysis can justify a smaller value, 1.0DTM is the minimum
clearance permitted by Section 17.5 of ASCE/SEI 7-05. For the generic
system, moat wall gap displacements of 0.6DTM and 0.8DTM test the
consequences of restricting isolation system displacement, and clearances of
1.2DTM and 1.4DTM evaluate the benefits of having extra clearance.
Moat wall clearance is influenced by site conditions (e.g., sloping site),
building configuration and architectural features, but economic
10-24 10: Supporting Studies FEMA P695
considerations usually dictate design at or near the minimum required
displacement, so that moat wall clearances between approximately 0.8DTM
and 1.0DTM are typical (when the configuration has a moat wall). A moat
wall clearance of 1.4DTM, or greater (42 inches in this study), is not common.
ASCE/SEI 7-05 does not permit moat wall clearance less than 0.8DTM unless
the superstructure is explicitly evaluated for stability at MCE demand (which
is not typically done).
Table 10-3 Summary of Moat Wall Clearance (Gap) Distances
Isolation System Properties Moat Wall Gap Distance (inches)
Type
Displacement (in.) Approximate Fraction of Code Minimum
DTM Fit-up 0.6 DTM 0.8 DTM 1.0 DTM 1.2 DTM 1.4 DTM
Generic Elastomeric or Sliding Systems - Dmax Designs
GEN 29.3 0.7 18 24 30 36 42
GEN-UB 29.3 0.7 30
GEN-LB 29.3 0.7 30
Rubber (RB) or Friction Pendulum (FP) Systems - Dmin Designs
RB 8.5 0.5 9
FP 5.7 0.3 6
Superstructure Design Properties
Isolated structure archetypes are grouped as Code-Conforming and Non-
Code-Conforming archetypes on the basis of whether the superstructure
meets code requirements for strength, ductility, and detailing. Code-
Conforming archetypes include systems with reinforced concrete special
moment frame superstructures that conform to all the design requirements of
Chapter 17 of ASCE/SEI 7-05. Non-Code-Conforming archetypes include
systems with superstructures that do not conform, either in terms of design
strength, such as reinforced concrete special moment frame superstructures
designed for less than the minimum required base shear, or in terms of
ductility, such as reinforced-concrete ordinary moment frame superstructures
not permitted for use as a SDC D system. In a few cases, the Non-Code-
Conforming superstructures exceed code requirements.
Table 10-4 summarizes design properties for the 3 superstructures of Code-
Conforming archetypes used in this study:
(C1) – A reinforced concrete special moment frame (perimeter frame)
system, designed for base shear, Vs = 0.092W (RI = 2.0).
FEMA P695 10: Supporting Studies 10-25
(C2) – A reinforced concrete special moment frame (space frame) system,
designed for base shear, Vs = 0.092W (RI = 2.0).
(C3) – A reinforced concrete special moment frame (space frame) system,
designed for base shear, Vs = 0.077W.
The first two systems (C1, C2) are superstructures of isolated archetypes
designed for SDC Dmax seismic criteria, and the last system (C3) is the
superstructure of isolated archetypes designed for SDC Dmin seismic criteria.
The base shear of the last system is governed by the limit of Section 17.5.4.3
of ASCE/SEI 7-05, which requires the base shear, Vs, not be less than 1.5
times either the “yield level” of an elastomeric system or the “breakaway”
friction level of a sliding system. In this case, the base shear (Vs = 0.077W) is
approximately equal to 1.5 times 0.05, the upper-bound yield level of
systems designed for SDC Dmin seismic criteria with either RB or FP
bearings.
Table 10-4 Isolated Structure Design Properties for Code-Conforming
Archetypes
Arch.
ID
Isolated Structure Archetype Design Properties
Superstructure Isolation System Isolated Structure
Vs/W .. Vmax/W Type
Gap
(in.)
TM
(sec.)
SMT
(g)
Reinforced Concrete Special Moment Frame Perimeter Systems
Evaluated at Dmax
C1-1 0.092 1.6 0.15 GEN 18 3.47 0.26
C1-2 0.092 1.6 0.15 GEN 24 3.47 0.26
C1-3 0.092 1.6 0.15 GEN 30 3.47 0.26
C1-4 0.092 1.6 0.15 GEN 36 3.47 0.26
C1-5 0.092 1.6 0.15 GEN 42 3.47 0.26
Reinforced Concrete Special Moment Frame Space Frame Systems
Evaluated at Dmax
C2-1 0.092 3.3 0.30 GEN 18 3.47 0.26
C2-2 0.092 3.3 0.30 GEN 24 3.47 0.26
C2-3 0.092 3.3 0.30 GEN 30 3.47 0.26
C2-4 0.092 3.3 0.30 GEN 36 3.47 0.26
C2-5 0.092 3.3 0.30 GEN 42 3.47 0.26
C2-3U 0.092 3.3 0.30 GEN-UB 30 3.47 0.26
C2-3B 0.092 3.3 0.30 GEN-LB 30 3.47 0.26
Reinforced Concrete Special Moment Frame Space Frame Systems
Evaluated at Dmin
C3-1 0.077 3.7 0.28 RB 9 3.18 0.09
C3-2 0.077 3.7 0.28 FP 6 2.78 0.11
10-26 10: Supporting Studies FEMA P695
The Code-Conforming archetypes include systems that meet all code
requirements for the superstructures (C1, C2 and C3). Isolation system
design meets all code requirements with the exception that moat wall gap
distances in C1-1, C1-2, C2-1 and C2-2, would typically not meet code
requirements unless special stability analyses were performed.
Table 10-5 summarizes the design properties of the Non-Code-Conforming
Archetypes. These isolated structures evaluate the effects of modifying code
requirements for (a) superstructure strength and (b) superstructure ductility.
Variation in superstructure strength includes some structures, such as NC1,
that exceed code strength requirements, while others, such as NC2, do not
meet code strength requirements.
Table 10-5 Isolated Structure Design Properties for Non-Code-Conforming
Archetypes
Arch.
ID
Isolated Structure Archetype Design Properties
Superstructure Isolation System Isolated Structure
Vs/W .. Vmax/
W
Gap
(in.)
Type TM
(sec.)
SMT
(sec.)
Reinforced Concrete Special Moment Frame Space Frame Systems
Evaluated at Dmax
NC1-1 0.164 2.8 0.46 18 GEN 3.47 0.26
NC1-2 0.164 2.8 0.46 24 GEN 3.47 0.26
NC1-3 0.164 2.8 0.46 30 GEN 3.47 0.26
NC1-4 0.164 2.8 0.46 42 GEN 3.47 0.26
NC2-1 0.046 5.2 0.24 18 GEN 3.47 0.26
NC2-2 0.046 5.2 0.24 24 GEN 3.47 0.26
NC2-3 0.046 5.2 0.24 30 GEN 3.47 0.26
NC2-4 0.046 5.2 0.24 42 GEN 3.47 0.26
Reinforced Concrete Ordinary Moment Frame Space Frame Systems
Evaluated at Dmax
NC3-1 0.246 1.9 0.47 30 GEN 3.47 0.26
NC3-2 0.246 1.9 0.47 42 GEN 3.47 0.26
NC4-1 0.164 1.8 0.30 30 GEN 3.47 0.26
NC4-2 0.164 1.8 0.30 42 GEN 3.47 0.26
NC5-1 0.092 1.9 0.17 30 GEN 3.47 0.26
NC5-2 0.092 1.9 0.17 42 GEN 3.47 0.26
While reinforced concrete ordinary moment frame systems from Chapter 9
are used for the purpose of investigating the effects of superstructure
ductility, it should be noted that ASCE/SEI 7-05 does not permit isolated
systems with a reinforced concrete ordinary moment frame superstructure in
FEMA P695 10: Supporting Studies 10-27
regions of high seismicity. The 2006 IBC does, however, allow isolation
systems with either a steel ordinary moment frame or steel OCBF (braced
frame) in regions of high seismicity when designed in accordance with AISC
341-05 and RI = 1.0.
10.3.5 Nonlinear Static Analysis for Period-Based Ductility,
SSFs, Record-to-Record Variability and Overstrength
The Methodology requires a nonlinear static (pushover) analysis to determine
the overstrength of the archetype and to evaluate values of period-based
ductility (µT = du/dy,eff) for determining the spectral shape factor, SSF, and, in
certain instances, record-to-record variability, ßRTR. Although the
overstrength parameter, O0, is not used in isolation design, pushover analysis
provides a useful tool for evaluating the strength of the superstructure
relative to the level of lateral force in the isolation system.
Pushover Analysis of Isolated Systems
Nonlinear static analysis is performed on isolated structures with the
isolation system free to displace. Pushover forces are based on a uniform
pattern of lateral load emulating the approximate pattern of uniform lateral
displacement of the isolated structure (at displacements up to significant
yielding of the superstructure). Figure 10-11 illustrates results of nonlinear
static analysis for an isolated structure and for the same superstructure on a
fixed-base.3 Approximately the same ultimate strength is obtained from the
two analyses, but the isolated structure effectively shares system ductility
between displacement of the isolation system and displacement of the
superstructure.
It is also noted that pushover analysis of isolated systems is performed
without the moat wall springs (as the sudden increase in stiffness associated
with the moat wall is inconsistent with the assumptions used in developing
the relationship between µT and SSF in Appendix B).
Superstructure Overstrength Properties
A static pushover analysis of each of the archetype superstructures listed in
Table 10-3 and Table 10-4 was performed to determine the actual maximum
strength, Vmax, and to compare actual strength with design strength, Vs.
Values of normalized design strength (Vs/W) and normalized maximum
strength (Vmax/W) are plotted in Figure 10-12.
3 Nonlinear static analysis of the fixed-base structure is based on the lateral load
pattern prescribed by Equation 12.8-13 of ASCE/SEI 7-05.
10-28 10: Supporting Studies FEMA P695
0 10 20 30 40 50 60 70 80
0
100
200
300
400
500
600
Base Shear (kips)
Roof Displacement (inches)
Fixed-base
structure
Isolated structure
Isolators yield
Displacement
primarily in
isolation
system
Ultimate base shear is approximately
the same in isolated and fixed-base
structure
Figure 10-11 Pushover curves of a non-code-conforming isolated structure
(NC1) and the same superstructure on a fixed base.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
0 0.05 0.1 0.15 0.2 0.25
Normalized Design Shear (V s /W) Normalized Strength (Vmax/W)
Space RC SMF
Perimeter RC SMF
Space RC OMF
Figure 10-12 Normalized design shear (Vs/W) and maximum strength (Vmax/W)
of superstructures.
As Figure 10-12 shows, the maximum strength of the superstructure is
system dependent and does not decrease in proportion to design base shear.
When designed with a base shear of 0.092W, the perimeter reinforced
concrete special moment frame has an ultimate strength 0.15W, the space
reinforced concrete special moment frame has an ultimate strength of 0.30W
and the space reinforced concrete ordinary moment frame has an ultimate
strength of 0.17W. Space frames typically have higher overstrength (relative
FEMA P695 10: Supporting Studies 10-29
to the design lateral load) than perimeter frames due to the higher
contribution of gravity loads in space frame designs. Reinforced concrete
ordinary moment frame systems typically have lower overstrength because of
a lack of capacity-design provisions. The relative strength of the
superstructure and isolation system can have a significant influence on the
performance of isolated structures under extreme loading.
System Period-Based Ductility and Spectral Shape Factor (SSF)
Pushover analysis of the isolated system (including both isolator and
superstructure) is used for determining period-based ductility, µT = du/dy,eff,
for computation of the SSF. The calculation of µT is illustrated in Figure
10-13.
0 10 20 30 40 50 60 70 80
0
100
200
300
400
500
600
Base Shear (kips)
Roof Displacement (inches)
Vmax
0.8Vmax
d du = 79.4 in y,eff = 46.8 in
0 10 20 30 40 50 60 70 80
0
100
200
300
400
500
600
Base Shear (kips)
Roof Displacement (inches)
Vmax
0.8Vmax
du d = 79.4 in y,eff = 46.8 in
Figure 10-13 Illustration of calculation of period-based ductility, µT, for
isolated system NC-1, where µT = du/ dy,eff.
Calculation of the effective yield roof drift displacement, dy,eff, for base
isolated systems is given by the formula:
2
2
max
, 4 y eff M g T
W
V
..
.
..
. .
.
. (10-6)
which is based on Equation 6-7 for values of the modal coefficient, C0 = 1.0
and fundamental period, T = T1 = TM. The modal coefficient, C0, is defined
by Equation 6-8 and generally has values near unity for isolated structures
10-30 10: Supporting Studies FEMA P695
with roof drift displacement dominated by lateral displacement of the
isolation system displacement. In the limiting case of roof drift displacement
equal to isolation system displacement (i.e., no drift in the superstructure),
the ultimate roof drift displacement, du, would be equal to the effective yield
roof drift displacement, dy,eff.
Due to inherent flexibility in isolated systems, period-based ductility is
smaller for base isolated systems than for fixed base systems because base
isolated systems have a larger effective yield roof drift displacement, dy,eff,
relative to the ultimate roof displacement, du. For fixed-base special moment
frames in SDC Dmax, µT is approximately equal to 11, whereas for isolated
special moment frames, µT ranges between 1.4 and 2.5. The period-based
ductility values obtained in this study are relatively large due to the
flexibility of reinforced concrete frame superstructures. Typically, isolated
buildings would not have as much displacement in the superstructure.
Record-To-Record Uncertainty (.RTR)
For most conventional (fixed-base) structures, the period-based ductility, µT
is greater than or equal to 3.0, and record-to-record variability, .RTR, is equal
to 0.40. Isolated systems have limited period-based ductility (i.e., µT = 3.0),
and Equation 7-2 is used to calculate ßRTR for these systems. Values of .RTR
calculated for base isolated systems using Equation 7-2 range from 0.24 to
0.35, with an average value of 0.27.
Values of SSF and .RTR for isolated systems were computed in accordance
with Table 7-1 and Equation 7-2, respectively, which were derived for
conventional fixed-base structures. The resulting values of SSF and .RTR
showed good agreement with values of spectral shape adjustment and recordto-
record variability obtained directly from collapse assessment of isolated
systems.
10.3.6 Collapse Evaluation Results
Evaluation Process and Acceptance Criteria
The base isolated systems listed in Tables 10-4 and 10-5 are analyzed and
evaluated, as described in Chapter 5 through Chapter 7. Nonlinear dynamic
analyses are used with the Far-Field record set to determine the median
spectral acceleration at which the structure collapses ( CT Sˆ
). The collapse
margin ratio, CMR, is computed as the ratio of the median collapse
capacity, CT Sˆ
, and the MCE demand, SMT. These results are reported in
Tables 10-6 and 10-7.
FEMA P695 10: Supporting Studies 10-31
Note that for base isolated reinforced concrete ordinary moment frames,
CMR accounts for both the sidesway simulated collapse modes and nonsimulated
column shear failure modes not included in simulation models.
Collapse due to column shear failure is predicted when the column loses its
vertical-load-carrying capacity, on the basis of component fragility functions
as described in Chapter 9. Section 10.2 describes the procedure for
incorporating non-simulated failure modes.
Acceptance criteria are based on the adjusted collapse margin ratio (ACMR)
which is the CMR modified by the spectral shape factor, SSF, to account for
the unique spectral shape of rare ground motions. The computed values of
period-based ductility, µT, for base isolated systems are reported in Table
10-6 and Table 10-7. The spectral shape factor is determined from Table 7-1
as a function of µT and building period (T > 1.5 seconds for all the isolated
systems).
The composite (total) uncertainty, .TOT, associated with collapse must be
assessed in order to compare the ACMR to the acceptance criteria. Isolated
archetypes with ductile (reinforced concrete special moment frame)
superstructures are assigned ratings of (B) Good for modeling, (B) Good for
test data, and (A) Superior for design requirements. The isolated archetypes
with non-ductile (reinforced concrete ordinary moment frame)
superstructures have ratings of (C) Fair for modeling, (B) Good for test data,
and (A) Superior for design requirements.
For isolated structures, which have limited period-based ductility (i.e., µT <
3.0), record-to-record variability, .RTR , is calculated using Equation 7-2.
Values of record-to-record variability range from 0.2 to 0.3 for the isolated
archetypes of this study. These smaller values are supported by a special
study on record-to-record variability described in Appendix A.
Record-to-record variability is combined with other sources of uncertainty
(i.e., modeling, test data and design requirements, respectively) to determine
total composite uncertainty, .TOT, in accordance with Equation 7-5. For the
isolated systems considered,..TOT ranges from 0.375 to 0.475. The
acceptable ACMR in Chapter 7 will depend on the total uncertainty for each
isolated system archetype.
Since design of isolated structures is based on building-specific testing and
evaluation, archetype systems are not grouped into performance groups in
this study. Instead, each archetype is evaluated individually in comparison
with the acceptable ACMR. In judging acceptability of each isolation system
archetype, the computed ACMR is compared to the acceptable ACMR
10-32 10: Supporting Studies FEMA P695
associated with a 10% probability of collapse, ACMR10%. The use of the 10%
criteria for evaluating the performance of an individual building is consistent
with recommendations contained in Appendix F.
Collapse Results for Code-Conforming Archetypes
Collapse results for the Code-Conforming base isolation system archetypes
are tabulated in Table 10-6a, Table 10-6b, and Table 10-6c.
Table 10-6a Collapse Results for Code-Conforming Archetypes: Various Gap
Sizes
Arch.
No.
Gap
Size
(in.)
Computed Collapse Margin Ratio Acceptable ACMR
CMR µT SSF ACMR .TOT ACMR10%
Perimeter Reinforced Concrete Special Moment Frame Systems
Evaluated at Dmax
C1-1 18 1.54 1.94 1.23 1.89 0.425 1.72
C1-2 24 1.66 1.94 1.23 2.04 0.425 1.72
C1-3 30 1.67 1.94 1.23 2.05 0.425 1.72
C1-4 36 1.70 1.94 1.23 2.09 0.425 1.72
C1-5 42 1.70 1.94 1.23 2.09 0.425 1.72
Space Reinforced Concrete Special Moment Frame Systems
Evaluated at Dmax
C2-1 18 1.92 1.57 1.18 2.27 0.400 1.67
C2-2 24 2.16 1.57 1.18 2.55 0.400 1.67
C2-3 30 2.19 1.57 1.18 2.58 0.400 1.67
C2-4 36 2.40 1.57 1.18 2.83 0.400 1.67
C2-5 42 2.52 1.57 1.18 2.97 0.400 1.67
Table 10-6b Collapse Results for Code-Conforming Archetypes: Nominal
(GEN), Upper-Bound (GEN-UB) and Lower-Bound (GEN-LB)
Isolator Properties
Arch.
No.
Isolator
Prop's.
Computed Collapse Margin Ratio Acceptable ACMR
CMR µT. SSF ACMR .TOT ACMR10%
Reinforced Concrete Special Moment Frame Space Frame Systems
Evaluated at Dmax - 30-inch Gap
C2-3 GEN 2.19 1.57 1.18 2.58 0.400 1.67
C2-3U GEN-UB 2.08 1.83 1.21 2.53 0.400 1.67
C2-3L GEN-LB 2.20 1.35 1.15 2.52 0.375 1.62
FEMA P695 10: Supporting Studies 10-33
Table 10-6c Collapse Results for Code-Conforming Archetypes: Minimum
Seismic Criteria (SDC Dmin)
Arch.
No.
Isolator
Props.
Computed Collapse Margin Ratio Acceptable ACMR
CMR µT. SSF ACMR .TOT ACMR10%
Reinforced Concrete Special Moment Frame Space Frame Systems
Evaluated at Dmax - 30-inch Gap
C3-1 RB 4.01 1.94 1.15 4.60 0.425 1.72
C3-2 FP 4.86 2.54 1.18 5.75 0.475 1.84
Comparison of computed and acceptable values of the ACMR reveals that the
code-conforming base isolation systems all easily meet the acceptance
criteria in this Methodology. The perimeter frame system (C1) has
consistently smaller ACMRs than the space frame system (C2) due to the
smaller lateral overstrength inherent in perimeter frame systems. This study
indicates that base isolated systems have comparable levels of safety to codeconforming,
conventional fixed-base structures.
Results summarized in Table 10-6a and plotted in Figure 10-14 illustrate the
effect of the moat wall clearance distance or gap size. The smallest gap sizes
(18 in. and 24 in.) would generally not be allowed by Chapter 17 of
ASCE/SEI 7-05, and are not code-compliant. These results are included for
comparison purposes only. It is noted, however, that even systems with moat
clearances less than the code minimum are acceptable according to the
collapse criteria of the Methodology. There may be some benefit to
increasing the gap size even beyond 30 inches in SDC Dmax. This benefit is
especially apparent for the space frame systems, which have sufficient
overstrength to avoid significant nonlinear behavior even when the forces in
the isolator are large. These results indicate that the code criteria for base
isolated systems are adequate, and may be conservative in the case of moat
wall clearance criteria for structures that have sufficient overstrength.
As shown in Table 10-6b, variation in isolator properties does not have a
significant effect on the computed collapse margin ratio. Variation in
isolator properties is based on ASCE/SEI 7-05 Section 17.8.4.3, which
specifies that there not be more than a 20% change in initial stiffness during
testing of a prescribed range of prototypes. This variation is not intended to
account for differences between modeled and actual behavior, such as
softening and damping loss due to heating effects. When upper-bound
(GEN-UB) and lower-bound (GEN-LB) isolator properties are used, ACMRs
are very close to the results for the nominal properties (GEN).
10-34 10: Supporting Studies FEMA P695
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
0 10 20 30 40 50
Moat Wall Gap Size (inches)
Ratio of ACMR to Acceptable ACMR10%
RC SMF System (Perimeter Frame) - Vmax/W = 0.15
RC SMF System (Space Frame) - Vmax/W = 0.30
Acceptable ACMR
Figure 10-14 Ratios of computed ACMR to acceptable ACMR10% for reinforced
concrete special moment frame Code-Conforming isolated
archetypes with various moat wall gap sizes, evaluated for SDC
Dmax seismic criteria.
Whereas Tables 10-6a and 10-6b report collapse margins evaluated for SDC
Dmax seismic criteria, Table 10-6c reports collapse margin ratios computed
for RB and FP systems in SDC Dmin. As observed in Chapter 9, the collapse
margins tend to increase as the seismic criteria decreases such that ACMRs in
Table 10-6c are large, and high seismic criteria (SDC Dmax) govern the
acceptability of this system.
Collapse Results for Non-Code-Conforming Archetypes
Collapse results for the Non-Code-Conforming isolation archetypes are
summarized in Tables 10-7a and 10-7b. Computed CMRs include both
simulated and non-simulated failure modes for the reinforced concrete
ordinary moment frame structures.
In this study, Non-Code-Conforming archetypes are systems that violate or
exceed certain code provisions in order to examine the effects of
superstructure strength (i.e., superstructures designed for a higher or lower
base shear than required according to ASCE/SEI 7-05) and superstructure
ductility (i.e., superstructures designed as ordinary moment frames without
the ductile detailing requirements for special moment frame systems). All
results in this section are for SDC Dmax.
FEMA P695 10: Supporting Studies 10-35
Table 10-7a Collapse Results for Isolated Archetypes with Ductile
Superstructures and Normalized Design Shear Values (Vs/W)
Not Equal to the Code-Required Value (Vs/W = 0.092)
Arch.
No.
Gap
Size
(in.)
Computed Collapse Margin Ratio Acceptable ACMR
CMR µT SSF ACMR .TOT ACMR10%
reinforced concrete special moment frame Space Frame Systems Evaluated at Dmax - Vs/W
= 0.164, Vmax/W = 0.46
NC1-1 18 2.08 1.47 1.17 2.42 0.400 1.67
NC1-2 24 2.28 1.47 1.16 2.66 0.400 1.67
NC1-3 30 2.32 1.47 1.16 2.70 0.400 1.67
NC1-4 42 2.64 1.47 1.16 3.08 0.400 1.67
reinforced concrete special moment frame Space Frame Systems Evaluated at Dmax - Vs/W
= 0.046, Vmax/W = 0.24
NC2-1 18 1.62 1.76 1.21 1.95 0.400 1.67
NC2-2 24 1.93 1.76 1.21 2.33 0.400 1.67
NC2-3 30 2.03 1.76 1.21 2.45 0.400 1.67
NC2-4 42 2.4 1.76 1.21 2.89 0.400 1.67
Table 10-7b Collapse Results for Isolated Archetypes with Non-Conforming
(Non-Ductile) Superstructures of Various Normalized Design
Shear Values (Vs/W)
Arch.
No.
Gap
Size
(in.)
Computed Collapse Margin Ratio Acceptable ACMR
CMR µT SSF ACMR .TOT ACMR10%
reinforced concrete ordinary moment frame Space Frame Systems Evaluated at Dmax -
Vs/W = 0.246, Vmax/W = 0.45
NC3-1 30 1.36 1.74 1.20 1.64 0.500 1.90
NC3-2 42 1.73 1.74 1.20 2.08 0.500 1.90
reinforced concrete ordinary moment frame Space Frame Systems Evaluated at Dmax -
Vs/W = 0.164, Vmax/W = 0.30
NC4-1 30 1.08 1.93 1.23 1.32 0.500 1.90
NC4-2 42 1.62 1.93 1.23 1.99 0.500 1.90
reinforced concrete ordinary moment frame Space Frame Systems Evaluated at Dmax -
Vs/W = 0.092, Vmax/W = 0.17
NC5-1 30 1.02 1.89 1.22 1.25 0.500 1.90
NC5-2 42 1.27 1.89 1.22 1.55 0.500 1.90
The effect of superstructure strength is demonstrated by the ACMR results
shown in Table 10-7a. Figure 10-15 shows trends in the ratio of calculated
ACMR to acceptable ACMR10%. For a reinforced concrete special moment
frame space frame (NC1-3) with a design base shear of Vs/W = 0.164 (above
Code minimum) and gap size of 30 inches, the computed ACMR is 2.57, a
factor of approximately 1.7 times the limit on acceptable ACMR. When the
design base shear is Vs/W = 0.092 (C1-3), the minimum value required by
10-36 10: Supporting Studies FEMA P695
ASCE/SEI 7-05, the computed ACMR is slightly smaller, approximately 1.6
times the limiting ACMR. For a smaller design base shear, Vs/W = 0.046
(NC2-3), below Code minimum, the computed ACMR is 1.5 times the
limiting ACMR. These results suggest that a large difference in the design
base shear (varying between 0.046 and 0.164 times the weight of the
structure) has only a modest impact on the collapse results.
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
0 10 20 30 40 50
Moat Wall Gap Size (inches)
Ratio of ACMR to Acceptable ACMR10%
RC SMF Vmax/W = 0.46 (Vs/W = 0.164) - Above Code
RC SMF Vmax/W = 0.24 (Vs/W = 0.046) - Below Code
RC SMF Vmax/W = 0.30 (Vs/W = 0.092) - Code Minimum
Acceptable ACMR
Figure 10-15 Ratios of computed ACMR to acceptable ACMR10% for reinforced
concrete special moment frame isolated archetypes with various
superstructure strengths and moat wall gap sizes, evaluated for
SDC Dmax seismic criteria
The reason the design base shear does not have a greater impact on the
ACMR is that the maximum strength of these structures is not linearly related
to the design base shear (as shown in Figure 10-12). The design rules for
special moment frame systems require capacity design and strong-columnweak-
beam requirements such that the actual strength is much larger than the
design strength, particularly for space frame systems which have significant
gravity loads providing additional overstrength. As a result, the structure
designed with Vs/W = 0.046 has a true strength of Vmax/W = 0.24, accounting
for its high collapse capacity. These observations indicate that the maximum
strength of the structure (Vmax/W) is a much better indicator of collapse
performance of isolated structures than the base shear used for design of the
superstructure (Vs/W).
Figure 10-16 shows the assessed collapse performance of reinforced concrete
special moment frame and reinforced concrete ordinary moment frame
isolated archetypes as a function of the strength of the superstructure
FEMA P695 10: Supporting Studies 10-37
(Vmax/W). These isolated systems have a peak MCE response of
approximately 0.25 g. Reinforced concrete special moment frame archetypes
that have a superstructure strength greater than 0.25 g have approximately
the same ACMR (as indicated by flattening of the curves with increasing
strength). For lower strength values, systems with less strength have
progressively lower values of ACMR. Due to their ductility, reinforced
concrete special moment frame systems with superstructure strength that is
significantly less than the MCE demand meet the acceptance criteria of the
Methodology.
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
0 0.1 0.2 0.3 0.4 0.5
Superstructure Strength (V max /W)
Ratio of ACMR to Acceptable ACMR10%
RC SMF Systems - 30-in. Moat Wall Gap
RC OMF Systems - 42-in. Moat Wall Gap
RC OMF Systems - 30-in. Moat Wall Gap
Acceptable ACMR
Figure 10-16 Ratios of computed ACMR to acceptable ACMR10% for reinforced
concrete special moment frame and reinforced concrete
ordinary moment frame isolated archetypes with different
superstructure strengths, evaluated for SDC Dmax seismic criteria.
The isolated archetypes in Table 10-7b are all reinforced concrete ordinary
moment frame systems, which are not permitted for use in regions of high
seismicity (SDC C and D) by Table 12.2-1 of ASCE/SEI 7-05. The collapse
results for reinforced concrete ordinary moment frame systems are shown in
Figure 10-16 as a function of superstructure strength, and Figure 10-17 as a
function of moat wall gap size.
Only two of the reinforced concrete ordinary moment frame systems in Table
10-7b are found to have acceptable collapse performance. This result
indicates that certain less-ductile superstructures may not provide adequate
collapse safety for isolated buildings, even with large moat wall clearance.
However, isolated buildings with strong superstructures (e.g., significant
overstrength), and large moat wall clearances, do have acceptable collapse
10-38 10: Supporting Studies FEMA P695
performance (as indicated by archetypes NC3-2 and NC4-2), even when the
superstructures are less ductile.
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
0 10 20 30 40 50
Moat Wall Gap Size (inches)
Ratio of ACMR to Acceptable ACMR10%
RC OMF Vmax/W = 0.47
RC OMF Vmax/W = 0.30
RC OMF Vmax/W = 0.17
Acceptable ACMR
Figure 10-17 Ratios of computed ACMR to acceptable ACMR10% for reinforced
concrete ordinary moment frame isolated archetypes with
various moat wall gap sizes, evaluated for SDC Dmax
Reinforced concrete ordinary moment frame systems that fail to meet the
acceptance criteria (NC3-1, NC4-1, NC5-1 and NC5-2) have ACMRs
between 15% and 35% lower than the minimum acceptable value. Increasing
the moat wall clearance, increasing superstructure strength, or improving the
ductility of ordinary moment frame superstructures would likely improve
their performance sufficiently to meet the acceptance criteria. Where large
moat wall clearance is provided (e.g., 42 inches), analyses suggest that
reinforced concrete ordinary moment frame systems with superstructure
strength Vmax/W greater than approximately 0.25 g would be acceptable. For
smaller moat wall clearances, a significantly stronger or more ductile
structure would be required (see Figure 10-16). This study also did not
investigate the use of less ductile systems for lower levels of seismicity (e.g.,
SDC Dmin), where superstructure strength and ductility demands are lower.
The effect of reinforced concrete ordinary moment frame superstructure
ductility on collapse performance can be investigated by examining how the
seismic performance of isolated reinforced concrete ordinary moment frame
systems change if design rules are modified such that brittle shear failure of
columns is prevented. Assuming all other design requirements remain
unchanged, ACMRs for these structures are recalculated assuming brittle
shear failure of columns is prevented. Results are plotted in Figure 10-18.
FEMA P695 10: Supporting Studies 10-39
On average, preventing column shear failure in these example archetypes
increased ACMRs by 25%. Figure 10-18 shows that this design change is
sufficient to improve the collapse performance of reinforced concrete
ordinary moment frame systems with a 30-inch gap such that the calculated
ACMRs meet (or nearly meet) the acceptability criteria of the Methdology.
These results suggest the desirability of a comprehensive study of limited
ductility systems on isolators, as collapse performance depends on
superstructure design rules. A more detailed examination of base isolated
reinforced concrete ordinary moment frame systems may also be warranted
with shear failure and post-shear failure degradation incorporated directly
into the models (rather than using non-simulated failure modes).
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
0 0.1 0.2 0.3 0.4 0.5
Superstructure Strength (V max /W)
Ratio of ACMR to Acceptable ACMR10%
RC OMF - 42-in. Gap(w /Shear Failure)
RC OMF - 30-in. Gap (w /Shear Failure)
RC OMF - 30 in. Gap (No Shear Failure)
Acceptable ACMR
Figure 10-18 Ratios of computed ACMR to acceptable ACMR10% for reinforced
concrete ordinary moment frame isolated archetypes with
various moat wall gap sizes, evaluated for SDC Dmax, illustrating
the effect of column shear failure on assessed collapse
performance.
10.3.7 Summary and Conclusion
This study illustrates application of the Methodology to isolated structures,
which have fundamentally different dynamic response characteristics,
performance properties and collapse failure modes than those of
conventional, fixed-base structures. It demonstrates that, when evaluated in
accordance with the Methodology, base isolated systems provide levels of
safety against collapse that are comparable to conventional, fixed-base
structures. When compared to conventional structures, the major benefit of
10-40 10: Supporting Studies FEMA P695
isolating structures is in reduced earthquake damage and loss, but these
effects have not been quantified in this study.
This study also demonstrates the potential use of the Methodology as a tool
for assessing validating and developing improvements to current design
requirements, in this case requirements for isolated structures. In particular,
the Methodology can be used to investigate and possibly improve on
apparent conservatisms or limitations for isolated structures related to the
following:
Superstructure Design Strength. Performance of isolated structures is
much more closely related to the “true” maximum strength, Vmax, than the
design strength, Vs, of the superstructure due to inherent differences in the
overstrength of different systems (and designs). For systems with high
inherent overstrength, such as the special space frame designs, low design
base shears can be used. Other systems with less overstrength may behave
differently when isolated. Design base shear requirements of ASCE/SEI
7-05 may be conservative for highly ductile structures which have large
amounts of inherent overstrength. Rather than relying on generic values of
the RI factor (and unknown amounts overstrength), to define design strength,
the Methodology could be used to establish appropriate criteria for design
based on maximum strength, verified by pushover analysis.
Less Ductile (Non-Complying) Superstructures. In a similar process, the
Methodology could also be used to establish appropriate levels of true
strength for less ductile (non-complying) systems not currently permitted for
use with isolated buildings in high seismic regions. While the process would
be the same, true strength required for design of less ductile systems would
likely be significantly higher than that required for more ductile systems to
achieve the same acceptable level of collapse performance.
Moat Wall Gap Clearance (Displacement Restraint System). In the case
of moat wall clearance, superstructures with sufficient strength and ductility
were observed to be relatively insensitive to the size of the clearance. This
suggests that moat wall clearance requirements of ASCE/SEI 7-05 might be
conservative for highly ductile superstructures. The Methodology could be
used to design isolation systems with limited displacement capacity for less
intense, more common, earthquake ground motions. If the moat wall or other
displacement restraint mechanism is used to limit excessive displacement of
isolators for rare (MCE) ground motions, isolators could then be smaller and
more economical, and provide a more practical solution in regions of very
high seismic demands.
FEMA P695 11: Conclusions and Recommendations 11-1
Chapter 11
Conclusions and
Recommendations
This recommended Methodology provides a rational basis for establishing
global seismic performance factors (SPFs), including the response
modification coefficient, R factor, the system overstrength factor, .., and the
deflection amplification factor Cd of new seismic-force-resisting systems
proposed for inclusion in model building codes. The Methodology also
provides a more reliable basis for re-evaluation of seismic performance
factors of seismic-force-resisting systems currently available in model
building codes and reference standards.
This chapter describes assumptions and limitations of the Methodology,
summarizes observations and conclusions resulting from its development,
discusses an adaptation of the Methodology to collapse performance
evaluation for an individual building, and provides recommendations for
further study.
11.1 Assumptions and Limitations
The Methodology is intended to apply broadly to all buildings, recognizing
that this objective may not be fully achieved for certain seismic environments
and building configurations. Likewise, the Methodology has incorporated
certain simplifying assumptions deemed appropriate for reliable evaluation of
seismic performance. Key assumptions and potential limitations of the
Methodology are summarized in the following sections.
11.1.1 Far-Field Record Set Ground Motions
The Methodology specifies the same set of ground motions (i.e., Far-Field
record set) for collapse performance evaluation of all systems. Records of
the Far-Field record set are unambiguously defined (including scaling) to
avoid any subjectivity in the ground motions used for nonlinear dynamic
analyses. The Far-Field record set is a robust sample of strong motion
records from large magnitude events. Even so, these records have inherent
limitations for certain buildings.
11-2 11: Conclusions and Recommendations FEMA P695
Buildings at Sites near Active Faults
Two sets of ground motion data, the Far-Field record set (records at sites at
least 10 km from fault rupture) and the Near-Field record set (records at sites
within 10 km of fault rupture) were developed. An internal study,
documented in Appendix A, found that the collapse margin ratio, CMR, was
somewhat smaller for a system designed for "near-fault" (SDC E) seismic
criteria, and evaluated using the Near-Field record set, than for same system
designed for SDC D seismic criteria, and evaluated using the Far-Field
record set. This implies that somewhat smaller values of the response
modification coefficient, R, would be appropriate for design of buildings near
active faults.
Various alternatives were considered, including the development of a
separate set of seismic performance factors for design of buildings near
active faults. For simplicity, as well as consistency with ASCE/SEI 7-05, a
single set of seismic performance factors based on collapse assessments
using the Far-Field record set was chosen. In so doing, the Methodology
implicitly accepts somewhat greater life safety risk for buildings located
close to active faults. This is consistent with the approach in ASCE/SEI
7-05, which implicitly accepts somewhat greater life safety risk for buildings
near active faults by limiting MCE ground motions to deterministic values of
seismic hazard.
Although the Methodology uses the Far-Field record set to establish seismic
performance factors for a new seismic-force-resisting systems, the Near-Field
record set would be more appropriate for verifying life safety performance of
an individual building located near an active fault.
Buildings at Sites in the Central and Eastern United States or
Subject to Deep Subduction Earthquakes
The Far-Field record set is a robust sample of all strong motion records from
large magnitude events recorded at sites greater than 10 km from fault a
rupture. No attempt was made to limit records based on tectonic setting or
fault mechanism during the record selection process. The Far-Field record
set is dominated by shallow crustal earthquakes, representative of areas in the
Western United States. It does not include strong motions records from deep
subduction earthquakes, or from Central and Eastern United States
earthquakes, since such records do not exist.
Duration of strong shaking is an important parameter in collapse analyses of
degrading systems, and strong motion records were purposely selected from
large magnitude events to adequately capture shaking duration. Very large
FEMA P695 11: Conclusions and Recommendations 11-3
magnitude earthquakes associated with deep subduction zone events would
be expected to have longer durations of strong shaking, on average. Central
and Eastern United States events could have different shaking characteristics
also affecting the collapse margin ratio and related seismic performance
factors. In spite of limitations in available deep subduction zone, or Central
and Eastern United States earthquake records, actual earthquake records,
rather than artificial or theoretical ground motions, were selected as the basis
for collapse assessment.
Buildings with Very Long Periods
The usefulness of earthquake records at very long periods is limited by the
ability of strong motion instruments to accurately record long-period
vibration. Records from older instruments may only be accurate to one or
two seconds, while records from the newer instruments are typically accurate
to at least 10 seconds. The Far-Field and Near-Field record sets include only
those records deemed accurate to a period of at least 4 seconds by the agency
responsible for processing the record. Most records in these sets are accurate
to a period of at least 10 seconds.
The Methodology conservatively limits the elastic period of index archetype
configurations to a period not greater than 4 seconds to ensure valid
evaluations of collapse performance. Records from the Far-Field and Near-
Field record sets should be used with caution to evaluate collapse
performance of buildings with elastic periods greater than 4 seconds (i.e.,
very tall buildings), since the spectral content of some records may not be
valid, and the associated value of the spectral shape factor, SSF, may be
overstated by the Methodology.
11.1.2 Influence of Secondary Systems on Collapse
Performance
The Methodology evaluates collapse performance of the seismic-forceresisting
system, ignoring the influence of secondary systems, such as gravity
systems and nonstructural components, which are not included in the
designation of the seismic-force-resisting system. Such systems can either
improve or diminish the collapse performance of a system of interest.
Potential for Improved Performance
For many buildings, elements of the gravity or nonstructural systems can
significantly improve collapse performance by providing additional
resistance to lateral forces. This is particularly true when the secondary
system has a larger lateral capacity than that of that of the primary system
(FEMA, 2009).
11-4 11: Conclusions and Recommendations FEMA P695
Incorporation of elements of the gravity or nonstructural systems in index
archetype configurations was considered. It was ultimately decided to limit
participating elements to those defined as part of the seismic-force-resisting
system. It was considered inappropriate to take the beneficial effects of
elements and components that would not be subject to regulation under the
earthquake design provisions. As an alternative, it was considered
permissible to include elements of the gravity or nonstructural systems in the
assessment, if they were also included in the definition of the seismic-forceresisting
system, and subject to the criteria contained within the system
design requirements.
Potential for Reduced Performance
Partial collapse of buildings has occurred when elements of the gravity or
nonstructural systems are not able to sustain lateral deformations of the
seismic-force-resisting system. Options for explicitly modeling and
evaluating gravity system performance were considered. Ultimately it was
considered to be problematic, and not completely relevant to the evaluation
of seismic performance factors for a seismic-force-resisting system, which
must be qualified for use with many different secondary systems. Rather,
displacement compatibility requirements of ASCE/SEI 7-05 are relied upon
to protect secondary systems from collapse failure.
It should be noted that current displacement compatibility requirements of
Section 12.12.4 of ASCE/SEI 7-05 may not be adequate to protect against
premature failure of a gravity system. They only apply to SDC D (and SDC
E and F) structures, and are based on design story drift, which is substantially
less than the peak inelastic story drift of ductile seismic-force-resisting
systems at the point of incipient collapse.
11.1.3 Buildings with Significant Irregularities
Significant irregularities, including torsion (horizontal structural irregularity
Types 1a and 1b, Table 12.3-1, ASCE/SEI 7-05) and soft/weak story
(vertical structural irregularity Types 1a and 1b, Table 12.3-2, ASCE/SEI
7-05), are known contributors to building collapse. Options for explicit
modeling of irregularity were considered, but internal studies showed this to
be unnecessary for evaluating collapse margins and related seismic
performance factors for generic seismic-force-resisting systems. Limits on
the use of equivalent lateral force (ELF) analysis given in Table 12.6-1 of
ASCE/SEI 7-05, and related design conservatisms of the ELF procedure
(e.g., accidental torsion, P-delta effects), were considered adequate to either
limit the effects of significant irregularities, or require more detailed,
dynamic analysis.
FEMA P695 11: Conclusions and Recommendations 11-5
It should be noted that limits on the use of the ELF procedure are more
restrictive for SDC D (and SDC E and F) structures, and certain design
requirements (e.g., amplification of accidental torsion) do not apply to SDC
B structures, so that there may be some additional life safety risk due to
irregularity inherent to systems designed for either SDC B or SDC C criteria.
11.1.4 Redundancy of the Seismic-Force-Resisting System
Section 12.3.4.2 of ASCE/SEI 7-05 requires the seismic-force-resisting
system of structures assigned to Seismic Design Categories D, E, and F to be
designed for seismic loads increased by the redundancy factor, ., where ..=
1.3, unless the configuration meets certain requirements for redundancy.
Options for explicit modeling of non-redundant systems were considered, but
internal studies showed this to be unnecessary for evaluating collapse
margins and related seismic performance factors for generic seismic-forceresisting
systems. The Methodology assumes . = 1.0 for design of structural
system archetypes, since larger values of . would be unconservative for
collapse evaluation of archetypes that generally meet redundancy
requirements of Table 12.3-3 of ASCE/SEI 7-05.
11.2 Observations and Conclusions
In the development of this Methodology, selected seismic-force-resisting
systems were evaluated to illustrate the application of the Methodology and
verify its methods. Results of these studies provide insight into the collapse
performance of buildings and appropriate values of seismic performance
factors. Observations and conclusions in terms of generic findings applicable
to all systems, and specific findings for certain types of seismic-forceresisting
systems are described below. These findings should be considered
generally representative, but not necessarily indicative of all possible trends,
given limitations in the number and types of systems evaluated.
11.2.1 Generic Findings
The following generic findings and conclusions apply to seismic-forceresisting
systems in general.
Systems Approach
Collapse performance (and associated seismic performance factors) must be
evaluated in terms of the behavior of the overall seismic-force-resisting
system, and not the behavior of individual components or elements of the
system. Collapse failure modes are highly dependent on the configuration
and interaction of elements within a seismic-force-resisting system. Seismic
11-6 11: Conclusions and Recommendations FEMA P695
performance factors should be considered as applying to an entire seismicforce-
resisting system, and not elements comprising it.
Precision of Seismic Performance Factors
In general, there is no practical difference in the collapse performance of
systems designed with fractional differences in the response modification
coefficient, R. For example, collapse performance of structures designed for
R = 6 and R = 6.5 is essentially the same, all else being equal. There is a
discernible, but modest difference in collapse performance for systems
designed for moderately different values of R, for example R = 6 and R = 8.
There is, however, a significant difference in collapse performance for
systems designed using different multiples of R, as in R = 3 versus R = 6.
Current values of R provided in Table 12.2-1 (e.g., 3, 3-1/4 and 3-1/2) reflect
a degree of precision that is not supported by results of example collapse
evaluations.
Spectral Content of Ground Motions
Consideration of spectral content (spectral shape) of ground motions can be
very important to the evaluation of collapse performance of ductile
structures. Epsilon-neutral earthquake records that are scaled to represent
very rare ground motions (ground motions corresponding to large positive
values of epsilon) can significantly overestimate demand on ductile
structures. The Methodology incorporates a spectral shape factor, SSF, that
adjusts calculated response to account for the spectral content of rare
earthquake ground motions and avoid overestimation of nonlinear response.
Short-Period Buildings
Consistent with prior research, values of collapse margin ratio are
consistently smaller for short-period buildings, regardless of the type of
seismic-force-resisting system. Unless they have adequate design strength,
short-period buildings generally do not meet collapse performance objectives
of the Methodology.
These findings suggest a possible need for period-dependent seismic
performance factors, such as a short-period value and a 1-second value of the
response modification coefficient, R, for each system. At present, the
Methodology determines a single value of each seismic performance factor,
independent of period, and consistent with the design requirements of
ASCE/SEI 7-05. It could, however, be modified to determine perioddependent
values of each factor.
FEMA P695 11: Conclusions and Recommendations 11-7
Governing Seismic Design Category
Values of collapse margin ratio for a seismic-force-resisting system designed
and evaluated for SDC D are generally smaller than corresponding values of
collapse margin ratio for the same seismic-force-resisting system designed
and evaluated for SDC C, all else being equal.
This trend is attributed to the increasing role of gravity loads in the strength
of seismic-force-resisting components as the level of seismic design
decreases. The size and strength of seismic-force-resisting components
supporting both seismic and gravity loads is not necessarily proportional to a
decrease in seismic design loads, since gravity loads do not decrease. The
role of gravity loads in the strength of the seismic-force-resisting system is
directly related to system overstrength, ., which also tends to increase as the
level of seismic design decreases.
These findings suggest that the response modification coefficient, R, will
generally be governed by the Seismic Design Category with the strongest
ground motions for which the system is proposed for use (e.g., SDC D for
seismic-force-resisting systems that are permitted for use in all SDCs). In
general, R factors based on SDC D evaluations are conservative for design of
the same seismic-force-resisting system designed for SDC C or SDC B
seismic criteria.
These findings also suggest a possible need for seismic load-dependent
seismic performance factors. For example, a system designed for SDC C
could be assigned a larger value of R than would be required for the same
system designed for SDC D. At present, the Methodology determines a
single value of each seismic performance factor, independent of seismic
design criteria, and consistent with the design requirements of ASCE/SEI
7-05. It could, however, be modified to determine SDC-specific values of
each factor.
Overstrength
Consistent with prior research, values of collapse margin ratio are strongly
related to calculated values of overstrength, regardless of the type of seismicforce-
resisting system. For larger values of overstrength, larger values of
collapse margin ratio are observed.
Calculated values of overstrength for different index archetype models vary
widely, depending on configuration and Seismic Design Category. Results
suggest that current values of system overstrength, .O, given in Table 12.2-1
of ASCE/SEI 7-05 are not representative of the actual overstrength present in
seismic-force-resisting systems. Values in the table generally vary between 2
11-8 11: Conclusions and Recommendations FEMA P695
and 3 for all systems (except cantilevered structures), while calculated values
of . varied from as low as 1.5 to over 6 for structural system archetypes
evaluated in the development of the Methodology.
Distribution of Inelastic Response
Consistent with prior research, values of collapse margin ratio are strongly
related to the distribution of inelastic response over the height of a structure,
regardless of the type of seismic-force-resisting system. Values of collapse
margin ratio are significantly larger for systems in which the inelastic
response is more evenly distributed throughout the system.
11.2.2 Specific Findings
The following key findings and conclusions apply specifically to reinforcedconcrete
moment frame systems, wood light-frame systems, and baseisolated
systems that were evaluated as part of the development of the
Methodology.
Example Application – Reinforced Concrete Special Moment Frame
Systems
Evaluation of reinforced concrete special moment frame systems found that,
in general, designs based on current ASCE/SEI 7-05 requirements (R = 8)
meet the collapse performance objectives of this Methodology. However,
trends in collapse margin ratio results suggested a potential collapse
deficiency for taller buildings that did not meet minimum base shear
requirements consistent with requirements in the predecessor document,
ASCE 7-02. This information was made available to the ASCE 7 Seismic
Committee, and a special code change proposal was passed in 2007
(Supplement No. 2), amending the minimum base shear requirements of
ASCE/SEI 7-05 to correct for this potential deficiency.
The root cause of the potential deficiency in collapse performance of
reinforced concrete special moment frame systems is the localization of large
lateral deformations in the lower stories, and the associated detrimental
P-delta effects on post-yield behavior. While imposing a minimum value of
design base shear eliminated the potential deficiency for these systems, other
approaches could have been used, including the introduction of height limits
(which would not be practical), enhancement of P-delta and strong-columnweak-
beam design criteria (which would require additional study), or the
development of period-dependent response modification coefficients, R
(which would effectively increase the design base shear for taller buildings).
FEMA P695 11: Conclusions and Recommendations 11-9
Example Application - Reinforced-Concrete Ordinary Moment
Frame Systems
Example evaluation of reinforced concrete ordinary moment frame systems
found that, in general, designs based on current ASCE/SEI 7-05 requirements
(R = 3) meet the collapse performance objectives of this Methodology,
considering that these systems are only permitted for use in SDC B
structures.
Example Application - Wood Light-Frame Systems
Example evaluations of wood light-frame systems found that, in general,
designs based on current ASCE/SEI 7-05 requirements (R = 6) meet the
collapse performance objectives of this Methodology.
Supporting Study - Seismically-Isolated Structures
A special study of seismically-isolated buildings focused on the performance
of isolated special and ordinary reinforced concrete frame systems. This
study found that code-conforming isolated structures with ductile
superstructures (e.g., special moment frame systems) generally meet collapse
performance objectives, but that isolated structures with non-ductile
superstructures (e.g., ordinary moment frame systems) may not. However, it
also found that with higher strengths, superstructures of isolated systems did
not need to have the full ductility capacity required of conventional buildings
to achieve acceptable collapse performance. This is an important finding for
the introduction of more economically detailed superstructures of isolated
buildings, although more comprehensive studies would be required to
develop appropriate code requirements.
Pushover evaluations of isolated structures found poor correlation between
the true (maximum) strength of the superstructure and the design strength.
This suggests that nonlinear static (pushover) analyses would be more
appropriate for verifying adequate overstrength of the superstructure than
selecting superstructure strength based on the approximate values of RI, and
would also provide a more reliable and economical basis for design.
11.3 Collapse Evaluation of Individual Buildings
Although developed as a tool to establish seismic performance factors for
generic seismic-force-resisting systems, the Methodology could be readily
adapted for collapse assessment of an individual building system. As such, it
could be used to demonstrate adequate collapse performance of the structural
system of a building designed using performance-based design methods,
permitted by Section 11.1.4 of ASCE/SEI 7-05. Specific methods for
11-10 11: Conclusions and Recommendations FEMA P695
collapse evaluation of an individual building system are described in
Appendix F.
11.3.1 Feasibility
It is anticipated that buildings designed using performance-based methods
will likely be large or important structures. Projects using performancebased
design methods typically utilize detailed models for linear and
nonlinear analyses of the building, and peer review is commonly required.
Such projects are already set up to utilize many components of the
Methodology.
11.3.2 Approach
The Methodology is based on the concept of collapse level ground motions,
defined as the level of ground motions that cause median collapse (i.e., onehalf
of the records in the set cause collapse). For a building to meet the
collapse performance objectives of this Methodology, the median collapse
capacity must be an acceptable amount above the MCE ground motion
demand level (i.e., the adjusted collapse margin ratio, ACMR, must exceed
acceptable values).
By starting with an acceptable collapse probability (for MCE ground
motions) and working backwards through the Methodology, values of the
spectral shape factor, SSF, and collapse margin ratio, CMR, can be evaluated
to determine the level of ground motions corresponding to median collapse.
The Methodology can be “reverse engineered” to determine the level of
ground motions for which not more than one-half of the records should cause
collapse. By scaling the record set to this level, trial designs for a subject
building can be evaluated. If the analytical model of the trial design survives
one-half or more of the records without collapse, then the building has a
collapse probability that is equal to (or less than) the acceptable collapse
probability (for MCE ground motions), and meets collapse performance
objective of the Methodology.
11.4 Recommendations for Further Study
The following recommendations are provided for possible future studies that
would help to: (1) further improve or refine the Methodology; or (2) utilize
the Methodology to investigate and develop potential improvements to the
seismic provisions of ASCE/SEI 7-05.
FEMA P695 11: Conclusions and Recommendations 11-11
11.4.1 Studies Related to Improving and Refining the
Methodology
Comprehensive Evaluation of Existing Systems
The primary objective of this study would be to “beta test” the Methodology
to further verify that this procedure will reliably and reasonably quantify
building seismic performance for the various building systems that are, or
will eventually become adopted by current building codes and standards
organizations.
The Methodology could also be used to set minimum acceptable design
criteria for standard code-approved systems and to provide guidance in the
selection of appropriate design criteria for other systems when linear design
methods are applied. It is possible that the Methodology could then be used
to modify or eliminate those systems or requirements that cannot reliably
meet, or do not relate, to these objectives.
Component Qualification
The Methodology thus far has been developed to comprehensively evaluate
entire building systems, including variations in system configuration. This
study would be intended to modify and simplify the procedure so that it can
be used on individual building components.
The need for a simplified component methodology has been identified by
construction materials industries and codes and standards organizations for
the purpose of reliably and accurately quantifying seismic performance of
various building components that currently are, or will eventually become
available for use as seismic-force-resisting components within building
systems.
Simplified Methods
Collapse simulation is a detailed, data-intensive process, which has a high
degree of uncertainty. In order to comprehensively evaluate entire building
systems, and all permissible variations in system configuration, the
Methodology is necessarily complex. This study would investigate and test
possible short cuts in the process to simplify the overall application of the
Methodology for use when more comprehensive evaluations are not
necessarily warranted.
Modeling of Short-Period Structures
The Methodology does not currently provide any specific guidance for
modeling foundation flexibility and considering possible beneficial effects of
soil-structure-interaction on the collapse performance of short-period
11-12 11: Conclusions and Recommendations FEMA P695
structures. This study would investigate these effects and, if justified,
develop guidance for explicitly modeling of foundation flexibility and
adjustment factors (similar to SSF) that would adjust collapse margin ratios
as function of period, for evaluation of structural system archetypes assumed
to be fixed on a rigid base.
11.4.2 Studies Related to Advancing Seismic Design Practice
and Building Code Requirements (ASCE/SEI 7-05)
Period-Dependent and Seismic Load-Dependent R factors
The equivalent lateral force (ELF) procedure of ASCE/SEI 7-05 defines base
shear in terms of a single value of R for a given seismic-force-resisting
system, independent of the period of the structure or the level of seismic
design criteria. Studies show that designs based on a single value of R do not
necessarily have consistent collapse performance, and that such
inconsistencies could be reduced or eliminated by specifying perioddependent
or seismic load-dependent values of R.
Use of period-dependent and load-dependent values of R would require
substantial revision to the equivalent lateral force method and related
requirements in seismic design codes and standards, but could provide a
basis for more efficient and economical design. The Methodology could be
used to investigate period and seismic load dependency of R, and the
feasibility of incorporating such factors. Potential benefits would be greater
in regions of lower seismicity, where seismic load-dependent values of R are
likely to be larger than those now specified for the same system in regions of
high seismicity.
Consideration of Gravity and Nonstructural Systems and
Components
The Methodology could be used to investigate the importance of the gravity
system and certain nonstructural components to collapse performance, and
investigate the feasibility of enhancing current seismic design requirements
to more appropriately incorporate these systems in the seismic design
process. This would include accounting for both the possible beneficial and
detrimental effects of these systems on collapse performance. Comparisons
of collapse results for archetypical models both with and without selected
gravity system components could be made to quantify the results.
Structural Irregularities and Redundancy
The Methodology could be used to investigate the importance of structural
system regularity and redundancy to collapse performance, and investigate
possible changes to current seismic design requirements with regard to these
FEMA P695 11: Conclusions and Recommendations 11-13
characteristics. Comparisons of collapse results for archetypical models both
with and without features of regularity and redundancy could be made to
quantify the effects.
Structures with Isolation and Damping Systems
Due to limited U.S. experience in strong earthquakes, code development
committees have necessarily and appropriately imposed certain conservative
restrictions on design requirements for seismically isolated and damped
structures. The Methodology could be used to evaluate the importance of
current design requirements for isolated systems, and investigate the
feasibility of removing unnecessary conservatism. Comparisons of collapse
results of archetypical models with non-ductile superstructures of varying
strength levels could be made to quantify required strength for meeting
collapse performance objectives. Removal of unnecessary conservatism
would reduce system cost and support greater use of these types of protective
systems.
FEMA P695 A: Ground Motion Record Sets A-1
Appendix A
Ground Motion Record Sets
This appendix describes the selection of ground motion record sets for
collapse assessment of building structures using nonlinear dynamic analysis
(NDA) methods. It summarizes the characteristics of the Far-Field and Near-
Field record sets and defines the scaling methods appropriate for collapse
evaluation of building archetypes based on incremental dynamic analysis.
The Methodology utilizes the Far-Field record set for nonlinear dynamic
analysis and related collapse assessment of archetype models.
Both the Far-Field and Near-Field record sets have average epsilon values
that are lower than expected for Maximum Considered Earthquake (MCE)
ground motions, and therefore can substantially underestimate calculated
collapse margin ratios without appropriate adjustment for spectral shape
effects. Adjustment of results from nonlinear dynamic analysis using these
record sets is described in Section A.4 (and Appendix B).
Three ground-motion-related studies are included at the end of this appendix.
The first study (Section A.11) compares collapse performance for archetypes
evaluated using Far-Field and Near-Field record sets, respectively, and shows
that collapse margins are somewhat smaller for the Near-Field set record.
The second study (Section A.12) addresses the robustness of the Far-Field
record set and shows that the selection criteria of Section A.7 yield a
sufficiently large number of representative records (for sites not close to fault
rupture). The third study (Section A.13) investigates record-to-record
variability, .RTR, of collapse results and develops a simple relationship for
estimating .RTR as a function of period-based ductility, .T, for systems that
have limited period elongation.
A.1 Introduction
Ground motion record sets include a set of ground motions recorded at sites
located greater than or equal to 10 km from fault rupture, referred to as the
“Far-Field” record set, and a set of ground motions recorded at sites less than
10 km from fault rupture, referred to as the “Near-Field” record set. The
Near-Field record set includes two subsets: (1) ground motions with strong
pulses, referred to as the “NF-Pulse” record subset, and (2) ground motions
without such pulses, referred to as the “NF-No Pulse” record subset.
A-2 A: Ground Motion Record Sets FEMA P695
A.2 Objectives
The Methodology requires a set of records that can be used for nonlinear
dynamic analysis of buildings and evaluation of the probability of collapse
for Maximum Considered Earthquake (MCE) ground motions. These
records meet a number of conflicting objectives, described below.
. Code (ASCE/SEI 7-05) Consistent – The records should be consistent
(to the extent possible) with the ground motion requirements of Section
16.1.3.2 of ASCE/SEI 7-05 Minimum Design Loads for Buildings and
Other Structures (ASCE, 2006a) for three-dimensional analysis of
structures. In particular, “ground motions shall consist of pairs of
appropriate horizontal ground motion acceleration components that shall
be selected and scaled from individual recorded events.”
. Very Strong Ground Motions – The records should represent very
strong ground motions corresponding to the MCE motion. In high
seismic regions where buildings are at greatest risk, few recorded ground
motions are intense enough, and significant upward scaling of the
records is often required.
. Large Number of Records – The number of records in the set should be
“statistically” sufficient such that the results of collapse evaluations
adequately describe both the median value and record-to-record (RTR)
variability of collapse capacity.
. Structure Type Independent – Records should be broadly applicable to
collapse evaluation of a variety of structural systems, such as systems
that have different dynamic response properties or performance
characteristics. Accordingly, records should not depend on period, or
other building-specific properties of the structure.
. Site Hazard Independent – The records should be broadly applicable to
collapse evaluation of structures located at different sites, such as sites
with different ground motion hazard functions, site and source
conditions. Accordingly, records should not depend on hazard deaggregation,
or other site- or hazard-dependent properties.
No single set of records can fully meet all of the above objectives due, in
part, to inherent limitations in available data. Large magnitude events are
rare, and few existing earthquake ground motion records are strong enough
to collapse large fractions of modern, code-compliant buildings. In the
United States, strong-motion records date to the 1933 Long Beach
earthquake, with only a few records obtained from each event until the 1971
San Fernando earthquake.
FEMA P695 A: Ground Motion Record Sets A-3
Even with many instruments, strong motion instrumentation networks (e.g.,
Taiwan and California) provide coverage for only a small fraction of all
regions of high seismicity. Considering the size of the earth and period of
geologic time, the available sample of strong motion records from largemagnitude
earthquakes is still quite limited (and potentially biased by records
from more recent, relatively well-recorded events).
A.3 Approach
The Methodology requires a set of ground motion records for collapse
assessment of archetypical models that are appropriate for incremental
dynamic analysis (Vamvatsikos and Cornell, 2002), as adapted herein.
Incremental dynamic analysis (IDA) makes use of multiple response history
analyses for a given ground motion record of increasing intensity until
collapse occurs or the model otherwise reaches a collapse limit state. This
process is repeated for a set of ground motion records of sufficient number to
determine median collapse and record-to-record variability.
The Methodology follows the IDA concept of increasing ground motion
intensity to collapse, but applies each sequential increase in intensity,
collectively, to the entire set of ground motion records. Similar to IDA, the
Methodology characterizes intensity in terms of response spectral
acceleration at the fundamental period, T, of the system of interest, except
that intensity is defined collectively by the median spectral acceleration of
the record set, ST, rather than by different intensities for each record. Record
set intensity, ST, is the parameter used to define record set intensity when the
record set is scaled to a particular level of ground motions (e.g., MCE
spectral acceleration). The median value of collapse spectral acceleration,
CT Sˆ
, is the value of ST when the record set is scaled such that one-half of the
records affect collapse.
To ensure broad representation of different recorded earthquakes, sets of
ground motions contain records selected from all large-magnitude events in
the PEER NGA database (PEER, 2006a). The PEER NGA database is
described in Section A.6. A sufficient number of the strongest ground
motion records are selected from each event to permit statistical evaluation of
record-to-record variability.
Record selection does not distinguish ground motion records based on either
site condition or source mechanism. However, distance to fault rupture is
used to develop separate Far-Field and Near-Field record sets (for
examination of potential differences in collapse fragility due to near-source
directivity and pulse effects).
A-4 A: Ground Motion Record Sets FEMA P695
A.4 Spectral Shape Consideration
Spectral shape, i.e., frequency content of ground motions, can significantly
influence the calculation of collapse fragility and the collapse margin ratio of
“ductile” structures. Baker and Cornell (2006) have shown that rare ground
motions in the Western United States, such as those corresponding to the
MCE, have a distinctive spectral shape that, for a given fundamental-period
spectral acceleration, causes the record to be less damaging than other
records of less intensity.
In essence, the shape of the spectrum of rare ground motions drops off more
rapidly at periods both greater and less than the fundamental period of
interest (i.e., has less energy), as compared to spectra of other (less rare)
records. The amount by which spectral shape can influence the collapse ratio
is a function of the “rareness” of the ground motions. For ductile structures
located in coastal California, accounting for this spectral shape effect can
cause a 40% to 60% increase in the collapse margin ratio (i.e., median
collapse capacity).
The ground motion record sets do not directly incorporate the effect of
spectral shape. Direct incorporation of spectral shape would necessarily
require records to be selected based on the fundamental period of the
structure, resulting in a different set of records for each structure of differing
period. Rather, collapse margin ratios calculated using the ground motion
record sets are adjusted for spectral shape effects based on structure
deformation capacity and seismic design category, described in Chapter 7,
using factors developed in Appendix B.
A.5 Maximum Considered Earthquake and Design
Earthquake Demand (ASCE/SEI 7-05)
The seismic provisions of ASCE/SEI 7-05 specify ground motions and
design requirements in terms of the structure’s Seismic Design Category
(SDC), which is a function of the level of design earthquake (DE) ground
motions and the Occupancy Category of the structure. The Methodology is
based on life safety and assumes all structures to be either Occupancy
Category I or II (i.e., structures that do not have special functionality
requirements). Seismic Design Categories for Occupancy I and II structures
vary from SDC A in regions of very low seismicity (which is of least
interest) to SDC E in regions of highest seismicity near active faults.
The seismic provisions of ASCE/SEI 7-05 define MCE demand in terms of
mapped values of short-period spectral acceleration, SS, and 1-second
spectral acceleration, S1, site coefficients, Fa and Fv, and a standard response
FEMA P695 A: Ground Motion Record Sets A-5
spectrum shape. For seismic design of the structural system, ASCE/SEI 7-05
defines the DE demand as two-thirds of the MCE demand. Archetypical
systems are designed for DE ground motions and evaluated for collapse
using the corresponding set of MCE ground motions.
Mapped values of spectral acceleration vary greatly by seismic region. The
Methodology defines MCE and DE ground motions by the range of spectral
accelerations associated with Seismic Design Categories B, C, and D,
respectively. For each SDC, maximum and minimum ground motions are
based on the respective upper-bound and lower-bound values of DE spectral
acceleration, as given in Table 11.6-1 of ASCE/SEI 7-05, for short-period
response, and in Table 11.6-2 of ASCE/SEI 7-05, for 1-second response.
MCE spectral accelerations are derived from DE spectral accelerations for
site coefficients corresponding to Site Class D (stiff soil) following the
requirements of Section 11.4 of ASCE/SEI 7-05.
Tables A-1A and A-1B list values of spectral acceleration, site coefficients
and design parameters for maximum and minimum ground motions of
Seismic Design Categories B, C, and D. Figure A-1 shows MCE response
spectra for these ground motions.
Table A-1A Summary of Mapped Values of Short-Period Spectral
Accelerations, Site Coefficients, and Design Parameters Used
for Collapse Evaluation of Seismic Design Categories D, C,
and B Structure Archetypes
Seismic Design Category Maximum Considered Earthquake Design
Maximum Minimum SS (g) Fa SMS (g) SDS (g)
D 1.5 1.0 1.5 1.0
C D 0.55 1.36 0.75 0.50
B C 0.33 1.53 0.50 0.33
B 0.156 1.6 0.25 0.167
Table A-1B Summary of Mapped Values of 1-Second Spectral
Accelerations, Site Coefficients, and Design Parameters Used
for Collapse Evaluation of Seismic Design Categories D, C,
and B Structure Archetypes
Seismic Design Category Maximum Considered Earthquake Design
Maximum Minimum S1 (g) Fv SM1 (g) SD1 (g)
D 0.60 1.50 0.90 0.60
C D 0.132 2.28 0.30 0.20
B C 0.083 2.4 0.20 0.133
B 0.042 2.4 0.10 0.067
A-6 A: Ground Motion Record Sets FEMA P695
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4
Period (seconds)
Spectral Acceleration (g)
MCE SDC D (maximum)
MCE SDC D (min) or SDC C (max)
MCE SDC C (min) or SDC B (max)
MCE SDC B (minimum)
Figure A-1 Plots of MCE response spectral accelerations used for collapse
evaluation of Seismic Design Categories D, C, and B structure
archetypes.
The 1-second value of spectral acceleration shown in Table A-1B, S1 = 0.60,
is rounded upward slightly for convenience, and should be taken as S1 <
0.60g, for the purpose of evaluating minimum base shear design
requirements, i.e., Equation 12.8-6 of ASCE/SEI 7-05 should not be used for
design of index archetypes.
Maximum values of seismic criteria for SDC D (Ss = 1.5 g and S1 = 0.60 g)
are based on the effective boundary between deterministic (near-source) and
probabilistic regions of MCE ground motions, as defined in Section 21.2 of
ASCE/SEI 7-05. The Methodology purposely excludes SDC D structures at
deterministic (near-source) sites defined by 1-second spectral acceleration
equal to or greater than 0.60 g, i.e., Section 11.6 of ASCE/SEI 7-05 defines
SDC D structures as having 1-second spectral acceleration values as large as
0.75 g.
The seismic response coefficient, Cs, is used for design of the structure
archetypes. This coefficient is based on the fundamental period, T, which is
computed using Equation A-1. T is the code-defined period, and not the
period computed using eigenvalue analysis of the structural model.
x 0.25 seconds
u a u t n T . C T . C C h . (A-1)
FEMA P695 A: Ground Motion Record Sets A-7
where hn is the height, in feet, of the building above the base to the highest
level of the structure, and values of the coefficient, Cu, are given in Table
12.8-1 and values of period parameters, Ct and x, are given in Table 12.8-2 of
ASCE/SEI 7-05.
Period parameters, Ct and x are a function of structure type, distinguishing
between steel moment frames, reinforced concrete moment frames,
eccentrically braced steel frames and all other stiff structural systems.
Example values of the fundamental period, T, and corresponding MCE
spectral acceleration, SMT, are given in Table A-2 for concrete moment
resisting frame structures of various heights, ranging from 1-story to 20-
stories. Table A-2 lists example values of the fundamental period and MCE
spectral acceleration corresponding to maximum and minimum seismic
criteria of Seismic Design Categories B, C, and D.
Table A-2 Example values of the Fundamental Period, T, and
Corresponding MCE Spectral Acceleration, SMT, for
Reinforced Concrete Moment Frame Structures of Various
Heights
System
Properties
Seismic Design Category (SDC) Max and Min Seismic Criteria
SDC Dmax
SDC Cmax SDC Bmax
SDC Bmin
No. of
Stories
Height
(ft.)
SDC Dmin SDC Cmin
T (sec.) SMT
(g)
T (sec.) SMT
(g)
T (sec.) SMT
(g)
T (sec.) SMT
(g)
1 15 0.26 1.50 0.27 0.75 0.30 0.50 0.31 0.25
2 28 0.45 1.50 0.48 0.62 0.52 0.38 0.55 0.183
4 54 0.81 1.11 0.87 0.34 0.95 0.21 0.99 0.101
8 106 1.49 0.60 1.60 0.188 1.74 0.115 1.81 0.055
12 158 2.13 0.42 2.29 0.131 2.49 0.080 2.59 0.039
20 262 3.36 0.27 3.60 0.083 3.92 0.051 4.08 0.024
Section A.8 discusses scaling of record sets and provides factors for scaling
the Far Field record set to match MCE spectral acceleration, SMT, for Seismic
Design Categories B, C and D, respectively.
A.6 PEER NGA Database
The PEER NGA database is an extension of the earlier PEER Strong Motion
Database that was first made publicly available in 1999. The PEER NGA
database is composed of over 3,550 ground motion recordings that represent
over 160 seismic events (including aftershock events) ranging in magnitude
from M4.2 to M7.9.
A-8 A: Ground Motion Record Sets FEMA P695
Each ground motion recording includes two horizontal components of
acceleration; the vertical acceleration component is also available for many
records. In addition, the rotated fault-normal and fault-parallel motion
components are also available. This database contains records from a wide
range of domestic and international sources including the United States
Geological Survey, the California Division of Mines and Geology, the
Central Weather Bureau, the Earthquake Research Department of Turkey,
and many others.
The PEER NGA database was chosen for use in this study for several
reasons, including the large number of ground motion records, and the
availability of fault-normal and fault-parallel components of ground motion.
This report utilizes version 6.0 of the PEER NGA database. After this work
was completed, version 7.3 of the database was released, and some ground
motion parameters differ slightly in the new version, such as PGV. Even so,
the variations are small and have no tangible impact on the FEMA P695
Methodology.
A.7 Record Selection Criteria
This section describes record selection criteria developed to meet
Methodology objectives. Each criterion is listed, followed by a brief
discussion of the intent of the rule.
. Source Magnitude. Large-magnitude events pose the greatest risk of
building collapse due to inherently longer durations of strong shaking
and larger amounts of energy released. Ground motions of smaller
magnitude (M < 6.5) events can cause building damage (typically of a
nonstructural nature), but are not likely to collapse new structures. Even
when small magnitude events generate strong ground motions, the
duration of strong shaking is relatively short and the affected area is
relatively small. In contrast, large-magnitude events can generate strong,
long duration, ground motions over a large region, affecting a much
larger population of buildings.
. Source Type. Record sets include ground motions from earthquakes
with either strike-slip or reverse (thrust) sources. These sources are
typical of shallow crustal earthquakes in California and other Western
United States locations. Few strong-motion records are available from
other source mechanisms.
. Site Conditions. Record sets include ground motions recorded on either
soft rock (Site Class C) or stiff soil (Site Class D) sites. Records on soft
soil (Site Class E) or sites susceptible to ground failure (Site Class F) are
FEMA P695 A: Ground Motion Record Sets A-9
not used. Relatively few strong-motion records are available for Site
Class B (rock) sites.
. Site-Source. The 10 km source-to-site distance boundary between Near-
Field and Far-Field records is arbitrary, but generally consistent with the
“near fault” region of MCE design values maps in ASCE/SEI 7-05.
Several different measures of this distance are available. For this project,
the source-to-site distance was taken as the average of Campbell and
Joyner-Boore fault distances provided in the PEER NGA database.
. Number of Records per Event. Strong-motion instruments are not
evenly distributed across seismically active regions. Due to the number
of instruments in place at the time of the earthquake, some largemagnitude
events have generated many records, while others have
produced only a few. To avoid potential event-based bias in record sets,
not more than two records are taken from any one earthquake for a
record set. The two-record limit was applied separately to the Near-Field
“Pulse” and “No Pulse” record sets, respectively. When more than two
records of an event pass the other selection criteria, the two records with
highest peak ground velocity are selected.
. Strongest Ground Motion Records. The limits of greater than 0.2 g on
PGA and greater than 15 cm/sec on PGV are arbitrary, but generally
represent the threshold of structural damage (for new buildings) and
capture a large enough sample of the strongest ground motions (recorded
to date) to permit calculation of record-to-record variability.
. Strong-Motion Instrument Capability. Some strong-motion
instruments, particularly older models, have inherent limitations on their
ability to record long-period vibration accurately. Most records have a
valid frequency content of at least 8 seconds, but some records do not,
and records not valid to at least 4 seconds are excluded from the record
sets. The record sets are considered valid for collapse evaluation of tall
buildings with elastic fundamental periods up to about 4 seconds.
. Strong-Motion Instrument Location. Strong-motion instruments are
sometimes located inside buildings (e.g., ground floor or basement) that,
if large, can influence recorded motion due to soil-structure-foundation
interaction. Instead, instruments located in free-field location or on
ground floor of a small building should be used.
A.8 Scaling Method
Scaling of ground motion records is a necessary element of nonlinear
dynamic analysis, since few, if any, available unscaled records are strong
A-10 A: Ground Motion Record Sets FEMA P695
enough to collapse modern buildings. The scaling process of the
Methodology involves two elements: normalization and scaling.
Normalization of Records. Individual records of a given set are normalized
by their respective peak ground velocities. In essence, some records are
factored upwards (and some factored downwards), while maintaining the
same overall ground motion strength of the record set. Normalization by
peak ground velocity is a simple way to remove unwarranted variability
between records due to inherent differences in event magnitude, distance to
source, source type and site conditions, while still maintaining the inherent
aleatory (i.e., record-to-record) variability necessary for accurately predicting
collapse fragility.
Normalization is done with respect to the value of peak ground velocity
computed in the PEER NGA database (PGVPEER), which is the geometric
mean of PGV of the two horizontal components considering different record
orientations. The geometric mean (or geomean) is the square root of the
product of the two horizontal components and is a common parameter used
to characterize ground motions.
The following formulas define the normalization factor, NMi, and calculation
of normalized horizontal components of the ith record, respectively:
PEER, NMi . Median(PGV i )/PGVPEER,i (A-2)
,i i ,i
,i i ,i
NTH NM TH
NTH NM TH
2 2
1 1
. .
. .
(A-3)
where:
NMi = Normalization factor of both horizontal components of the ith
record (of the set of interest),
PGVPEER,i = Peak ground velocity of the ith record (PEER NGA
database),
Median (PGVPEER,i) = Median of PGVPEER,i values of records in the set,
NTH1,i = Normalized ith record, horizontal component 1,
NTH2,i = Normalized ith record, horizontal component 2,
TH1,i = Record i, horizontal component 1 (PEER NGA database),
and
TH2,i = Record i, horizontal component 2 (PEER NGA database).
FEMA P695 A: Ground Motion Record Sets A-11
Records and their corresponding values of PGVPEER are taken directly from
the PEER NGA database. For Near-Field records, the components rotated to
the fault normal (FN) and fault parallel (FP) directions are utilized.
Horizontal components of Far-Field records are not rotated based on the
orientation with respect to the fault and their as-recorded orientation is used.
Horizontal components of each record are normalized by the same
normalization factor, NMi, to maintain the relative, as-recorded strength of
the two components (which can be important for analysis of threedimensional
archetype models.
Table A-3 provides median values of 5%-damped spectral acceleration,
NRT Sˆ
, of normalized Far-Field and Near-Field record sets, respectively. The
Far-Field record set is described in Section A.9 and the Near-Field record set
is described in Section A.10, and Tables A-4D and A-6D summarize
normalization factors for each record of the Far-Field and Near-Field ground
motion sets, respectively.
Scaling of Record Sets. For collapse evaluation, the Methodology requires
the set of normalized ground motion records to be collectively scaled upward
(or downward) to the point that causes 50 percent of the ground motions to
collapse the archetype analysis model being evaluated. This approach is
used to determine the median collapse capacity, CT Sˆ
, of the model. Once
this has been established, the Methodology requires that the collapse margin
ratio, CMR, be computed between median collapse capacity, CT Sˆ
, and MCE
demand, SMT.
In the same manner, record sets can be scaled to a specific level of ground
motions e.g., MCE spectral acceleration. Record normalization and record
set scaling to match a particular level of ground motions parallels the ground
motion scaling requirements of Section 16.1.3.2 of ASCE/SEI 7-05, with the
notable exception that the median value of the scaled record set need only
match the MCE demand at the fundamental period, T, rather than over the
range of periods required by ASCE/SEI 7-05.
Table A-3 provides scaling factors for anchoring the normalized Far-Field
record set to the MCE demand level of interest. Scaling factors (for
anchoring the normalized Far-Field record set to MCE demand) depend on
the fundamental period of the building, T, which is shown in the left column
of the table. Figure A-2 illustrates the median spectrum of the normalized
Far-Field record set anchored to various levels of MCE demand at a period of
1 second.
A-12 A: Ground Motion Record Sets FEMA P695
Table A-3 Median 5%-Damped Spectral Accelerations of Normalized
Far-Field and Near-Field Record Sets and Scaling Factors for
Anchoring the Normalized Far-Field Record Set to MCE
Spectral Demand1.
Period
T = CuTa
(sec.)
Median Value of
Normalized Record Set
ˆ
SNRT (g)
Scaling Factors for Anchoring Far-
Field Record Set to MCE Spectral
Demand
Near-Field
Set
Far-Field
Set SDC Dmax
SDC Cmax SDC Bmax
SDC Bmin
SDC Dmin SDC Cmin
0.25 0.936 0.779 1.93 0.96 0.64 0.32
0.3 1.020 0.775 1.94 0.97 0.65 0.32
0.35 0.939 0.761 1.97 0.99 0.66 0.33
0.4 0.901 0.748 2.00 1.00 0.67 0.33
0.45 0.886 0.749 2.00 0.89 0.59 0.30
0.5 0.855 0.736 2.04 0.82 0.54 0.27
0.6 0.833 0.602 2.49 0.83 0.55 0.28
0.7 0.805 0.537 2.40 0.80 0.53 0.27
0.8 0.739 0.449 2.50 0.83 0.56 0.28
0.9 0.633 0.399 2.50 0.83 0.56 0.28
1.0 0.571 0.348 2.59 0.86 0.58 0.29
1.2 0.476 0.301 2.49 0.83 0.55 0.28
1.4 0.404 0.256 2.51 0.84 0.56 0.28
1.6 0.356 0.208 2.70 0.90 0.60 0.30
1.8 0.319 0.168 2.98 0.99 0.66 0.33
2.0 0.284 0.148 3.05 1.02 0.68 0.34
2.2 0.258 0.133 3.08 1.03 0.68 0.34
2.4 0.230 0.118 3.18 1.06 0.71 0.35
2.6 0.210 0.106 3.28 1.09 0.73 0.36
2.8 0.190 0.091 3.53 1.18 0.79 0.39
3.0 0.172 0.080 3.75 1.25 0.83 0.42
3.5 0.132 0.063 4.10 1.37 0.91 0.46
4.0 0.104 0.052 4.29 1.43 0.95 0.48
4.5 0.086 0.046 4.34 1.45 0.96 0.48
5.0 0.072 0.041 4.43 1.48 0.98 0.49
1. Spectral acceleration values and scaling factors may be based on linear interpolation for
periods not listed in the table.
In Figure A-2, the median spectrum of the normalized Far-Field record set is
factored by 2.59 to match the maximum MCE spectral acceleration of SDC
Dmax at a period of 1 second, and factored by 0.86 to match the minimum
MCE spectral acceleration of SDC Dmin at a period of 1 second. The figure
shows that the median spectrum of the Far-Field record set tends to match
FEMA P695 A: Ground Motion Record Sets A-13
MCE spectral acceleration at periods near the “anchor” period, but can
substantially deviate from MCE spectra at other periods. Such deviations are
unavoidable due to the approximate shape of code-defined spectra.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4
Period (seconds)
Spectral Acceleration (g)
FF Record Set Scaled to MCE SDC Dmax
FF Set Scaled to MCE SDC Dmin/Cmax
FF Set Scaled to MCE SDC Cmin/Bmax
FF Record Set Scaled to MCE SDC Bmin
Figure A-2 Example anchoring of median spectrum of the Far-Field record
set to MCE spectral acceleration at 1 second for maximum and
minimum seismic criteria of Seismic Design Categories B, C
and D.
A.9 Far-Field Record Set
The Far-Field record set includes twenty-two records (44 individual
components) selected from the PEER NGA database using the criteria from
Section A.7 of this appendix.
For each record, Table A-4A summarizes the magnitude, year, and name of
the event, as well as the name and owner of the station. The twenty-two
records are taken from 14 events that occurred between 1971 and 1999. Of
the 14 events, eight were California earthquakes and six were from five
different foreign countries. Event magnitudes range from M6.5 to M7.6 with
an average magnitude of M7.0 for the Far-Field record set.
For each record, Table A-4B summarizes site and source characteristics,
epicentral distances, and various other measures of site-source distance. Site
characteristics include shear wave velocity and the corresponding NEHRP
Site Class. Sixteen sites are classified as Site Class D (stiff soil sites) and the
remaining six are classified as Site Class C (very stiff soil sites). Fifteen
records are from events of predominantly strike-slip faulting and the
A-14 A: Ground Motion Record Sets FEMA P695
remaining seven records are from events of predominantly thrust (or reverse)
faulting.
Table A-4A Summary of Earthquake Event and Recording Station Data
for the Far-Field Record Set
M Year Name Name Owner
1 6.7 1994 Northridge Beverly Hills - Mulhol USC
2 6.7 1994 Northridge Canyon Country-WLC USC
3 7.1 1999 Duzce, Turkey Bolu ERD
4 7.1 1999 Hector Mine Hector SCSN
5 6.5 1979 Imperial Valley Delta UNAMUCSD
6 6.5 1979 Imperial Valley El Centro Array #11 USGS
7 6.9 1995 Kobe, Japan Nishi-Akashi CUE
8 6.9 1995 Kobe, Japan Shin-Osaka CUE
9 7.5 1999 Kocaeli, Turkey Duzce ERD
10 7.5 1999 Kocaeli, Turkey Arcelik KOERI
11 7.3 1992 Landers Yermo Fire Station CDMG
12 7.3 1992 Landers Coolwater SCE
13 6.9 1989 Loma Prieta Capitola CDMG
14 6.9 1989 Loma Prieta Gilroy Array #3 CDMG
15 7.4 1990 Manjil, Iran Abbar BHRC
16 6.5 1987 Superstition Hills El Centro Imp. Co. CDMG
17 6.5 1987 Superstition Hills Poe Road (temp) USGS
18 7.0 1992 Cape Mendocino Rio Dell Overpass CDMG
19 7.6 1999 Chi-Chi, Taiwan CHY101 CWB
20 7.6 1999 Chi-Chi, Taiwan TCU045 CWB
21 6.6 1971 San Fernando LA - Hollywood Stor CDMG
22 6.5 1976 Friuli, Italy Tolmezzo --
ID Earthquake Recording Station
No.
Site-source distances are given for the closest distance to fault rupture,
Campbell R distance, and Joyner-Boore horizontal distance to the surface
projection of the rupture. Based on the average of Campbell and Boore-
Joyner fault distances, the minimum site-source distance is 11.1 km, the
maximum distance is 26.4 km and the average distance is 16.4 km for the
Far-Field record set.
FEMA P695 A: Ground Motion Record Sets A-15
Table A-4B Summary of Site and Source Data for the Far-Field Record
Set
NEHRP
Class
Vs_30
(m/sec) Epicentral Closest to
Plane Campbell Joyner-
Boore
1 D 356 Thrust 13.3 17.2 17.2 9.4
2 D 309 Thrust 26.5 12.4 12.4 11.4
3 D 326 Strike-slip 41.3 12 12.4 12
4 C 685 Strike-slip 26.5 11.7 12 10.4
5 D 275 Strike-slip 33.7 22 22.5 22
6 D 196 Strike-slip 29.4 12.5 13.5 12.5
7 C 609 Strike-slip 8.7 7.1 25.2 7.1
8 D 256 Strike-slip 46 19.2 28.5 19.1
9 D 276 Strike-slip 98.2 15.4 15.4 13.6
10 C 523 Strike-slip 53.7 13.5 13.5 10.6
11 D 354 Strike-slip 86 23.6 23.8 23.6
12 D 271 Strike-slip 82.1 19.7 20 19.7
13 D 289 Strike-slip 9.8 15.2 35.5 8.7
14 D 350 Strike-slip 31.4 12.8 12.8 12.2
15 C 724 Strike-slip 40.4 12.6 13 12.6
16 D 192 Strike-slip 35.8 18.2 18.5 18.2
17 D 208 Strike-slip 11.2 11.2 11.7 11.2
18 D 312 Thrust 22.7 14.3 14.3 7.9
19 D 259 Thrust 32 10 15.5 10
20 C 705 Thrust 77.5 26 26.8 26
21 D 316 Thrust 39.5 22.8 25.9 22.8
22 C 425 Thrust 20.2 15.8 15.8 15
Site-Source Distance (km)
ID
No.
Site Data Source
(Fault
Type)
For each record, Table A-4C summarizes key record information from the
PEER NGA database. This record information includes the record sequence
number and file names of the two horizontal components, as well as the
lowest frequency (longest period) for which frequency content is considered
fully reliable.
Maximum values of as-recorded peak ground acceleration, PGAmax, and peak
ground velocity, PGVmax, are also given for each record. The term
“maximum” implies that the larger value of the two components is reported.
Peak ground acceleration values vary from 0.21 g to 0.82 g with an average
PGAmax of 0.43 g. Peak ground velocity values vary from 19 cm/second to
115 cm/second with an average PGVmax of 46 cm/second.
A-16 A: Ground Motion Record Sets FEMA P695
Table A-4C Summary of PEER NGA Database Information and
Parameters of Recorded Ground Motions for the Far-Field
Record Set
Component 1 Component 2
1 953 0.25 NORTHR/MUL009 NORTHR/MUL279 0.52 63
2 960 0.13 NORTHR/LOS000 NORTHR/LOS270 0.48 45
3 1602 0.06 DUZCE/BOL000 DUZCE/BOL090 0.82 62
4 1787 0.04 HECTOR/HEC000 HECTOR/HEC090 0.34 42
5 169 0.06 IMPVALL/H-DLT262 IMPVALL/H-DLT352 0.35 33
6 174 0.25 IMPVALL/H-E11140 IMPVALL/H-E11230 0.38 42
7 1111 0.13 KOBE/NIS000 KOBE/NIS090 0.51 37
8 1116 0.13 KOBE/SHI000 KOBE/SHI090 0.24 38
9 1158 0.24 KOCAELI/DZC180 KOCAELI/DZC270 0.36 59
10 1148 0.09 KOCAELI/ARC000 KOCAELI/ARC090 0.22 40
11 900 0.07 LANDERS/YER270 LANDERS/YER360 0.24 52
12 848 0.13 LANDERS/CLW-LN LANDERS/CLW-TR 0.42 42
13 752 0.13 LOMAP/CAP000 LOMAP/CAP090 0.53 35
14 767 0.13 LOMAP/G03000 LOMAP/G03090 0.56 45
15 1633 0.13 MANJIL/ABBAR--L MANJIL/ABBAR--T 0.51 54
16 721 0.13 SUPERST/B-ICC000 SUPERST/B-ICC090 0.36 46
17 725 0.25 SUPERST/B-POE270 SUPERST/B-POE360 0.45 36
18 829 0.07 CAPEMEND/RIO270 CAPEMEND/RIO360 0.55 44
19 1244 0.05 CHICHI/CHY101-E CHICHI/CHY101-N 0.44 115
20 1485 0.05 CHICHI/TCU045-E CHICHI/TCU045-N 0.51 39
21 68 0.25 SFERN/PEL090 SFERN/PEL180 0.21 19
22 125 0.13 FRIULI/A-TMZ000 FRIULI/A-TMZ270 0.35 31
PGA max
(g)
PEER-NGA Record Information Recorded Motions
ID
No. Record File Names - Horizontal Records
Seq. No.
Lowest
Freq (Hz.)
PGV max
(cm/s.)
For each record, Table A-4D summarizes the 1-second spectral acceleration
of both horizontal components, peak ground velocity reported in the PEER
NGA database, PGVPEER, normalization factors, NMi, and values of PGAmax
and PGVmax after normalization by PGVPEER.
FEMA P695 A: Ground Motion Record Sets A-17
Table A-4D Summary of Factors Used to Normalize Recorded Ground
Motions, and Parameters of Normalized Ground Motions
for the Far-Field Record Set
Comp. 1 Comp. 2
1 1.02 0.94 57.2 0.65 0.34 41
2 0.38 0.63 44.8 0.83 0.40 38
3 0.72 1.16 59.2 0.63 0.52 39
4 0.35 0.37 34.1 1.09 0.37 46
5 0.26 0.48 28.4 1.31 0.46 43
6 0.24 0.23 36.7 1.01 0.39 43
7 0.31 0.29 36.0 1.03 0.53 39
8 0.33 0.23 33.9 1.10 0.26 42
9 0.43 0.61 54.1 0.69 0.25 41
10 0.11 0.11 27.4 1.36 0.30 54
11 0.50 0.33 37.7 0.99 0.24 51
12 0.20 0.36 32.4 1.15 0.48 49
13 0.46 0.28 34.2 1.09 0.58 38
14 0.27 0.38 42.3 0.88 0.49 39
15 0.35 0.54 47.3 0.79 0.40 43
16 0.31 0.25 42.8 0.87 0.31 40
17 0.33 0.34 31.7 1.17 0.53 42
18 0.54 0.39 45.4 0.82 0.45 36
19 0.49 0.95 90.7 0.41 0.18 47
20 0.30 0.43 38.8 0.96 0.49 38
21 0.25 0.15 17.8 2.10 0.44 40
22 0.25 0.30 25.9 1.44 0.50 44
ID
No. PGVPEER
(cm/s.)
Normalization
Factor
As-Recorded Parameters
1-Sec. Spec. Acc. (g)
Normalized Motions
PGAmax
(g)
PGVmax
(cm/s.)
Normalization factors vary from 0.41 to 2.10. After normalization, peak
ground acceleration values vary from 0.18 g to 0.58 g with an average
PGAmax of 0.40 g. Peak ground velocity values vary from 36 cm/second to
54 cm/second with an average PGVmax of 42 cm/second. Table A-5 shows
that normalization of the records (by PGVPEER) has reduced the dispersion in
PGVmax to a level consistent with that of PGAmax without appreciably
affecting average values of PGAmax or PGVmax for the record set.
A-18 A: Ground Motion Record Sets FEMA P695
Table A-5 Far-Field Record Set (as-Recorded and After
Normalization): Comparison of Maximum, Minimum and
Average Values of Peak Ground Acceleration (PGAmax) and
Peak Ground Velocity (PGVmax), Respectively
Parameter
Value
PGAmax (g) PGVmax (g)
As-Recorded Normalized As-Recorded Normalized
Maximum 0.82 0.58 115 54
Minimum 0.21 0.18 19 36
Max/Min Ratio 3.9 3.2 6.1 1.5
Average 0.43 0.43 46 42
Figure A-3 shows response spectra of individual records of the normalized
Far-Field record set, as well as the median and one and two standard
deviation spectra of the set in log format. Figure A-4 shows median and onestandard
deviation spectra, as well as a plot of the standard deviation (natural
log of spectral acceleration) of the normalized Far-Field record set in linear
form.
0.01
0.1
1
10
0.01 0.1 1 10
Period (seconds)
Spectral Acceleration (g)
Median Spectrum - Far-Field Set
+ 1 LnStdDev Spectrum - FF Set
+ 2 LnStdDev Spectrum - FF Set
Figure A-3 Response spectra of the forty-four individual components of
the normalized Far-Field record set and median, one and
two standard deviation response spectra of the total record
set.
FEMA P695 A: Ground Motion Record Sets A-19
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Period (seconds)
Spectral Acceleration (g)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Standard Deviation - Ln (Sa)
Median Spectrum - Far-Field Set
+ 1 LnStdDev Spectrum - FF Set
Standard Deviation - Ln (Sa)
Figure A-4 Median and one standard deviation response spectra of
the normalized Far-Field record set, and plot of the
standard deviation (natural log) of response spectral
acceleration.
The median spectrum of the normalized Far-Field record set has frequency
content consistent with that of a large magnitude (M7) event recorded at 15
km from fault rupture. The median spectral acceleration at short periods is
about 0.8 g and median spectral acceleration at 1-second is about 0.35 g. The
domain of “constant acceleration” transitions to the domain of “constant
velocity” at about 0.5 second, consistent with soft rock and stiff soil site
response. Record-to-record variability (standard deviation of the logarithm
of spectral acceleration) ranges from about 0.5 at short periods to about 0.6 at
long periods, consistent with the values of dispersion from common ground
motion attenuation relations (e.g., see Figure 8, Campbell and Borzorgnia,
2003).
Recent research has shown that spectral shape is an important aspect of
collapse capacity prediction (Baker and Cornell, 2006) and that this is related
to a parameter called epsilon, .. Epsilon is defined as the number of standard
deviations between the observed spectral value and the median prediction
from an attenuation function such that it depends on both the period and the
attenuation function used. Figure A-5 is a plot of the median value of .
calculated for the forty-four components of the Far-Field record set. These
values are based on the attenuation relationship developed by Abrahamson
and Silva (1997).
Median values of . are generally small and near zero at all periods beyond 1
second, indicating that the Far-Field record set is approximately “. -neutral.”
A-20 A: Ground Motion Record Sets FEMA P695
Accordingly, collapse margin ratios (CMRs) based on incremental dynamic
analysis using these records do not account for the effects of spectral shape,
discussed in Section A.4, and are later increased to account for these effects,
using the factors described in Section 7.4 and developed in Appendix B.
-2.0
-1.6
-1.2
-0.8
-0.4
0.0
0.4
0.8
1.2
1.6
2.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Period (seconds)
Median Epsilon(T)
Far-Field Set
Figure A-5 Plot of median value of epsilon for the Far-Field
record set.
A.10 Near-Field Record Set
The Near-Field record set includes twenty-eight records (56 individual
components) selected from the PEER NGA database using the criteria from
Section A.7 of this appendix. Fourteen records have pulses (Pulse subset)
and fourteen records do not have pulses (No-Pulse subset), as judged by
wavelet analysis classification of the records (Baker, 2007).
For each record, Table A-6A summarizes the magnitude, year and name of
earthquake events and the name and owner of the station. The twenty-eight
records are taken from 14 events that occurred between 1976 and 2002. Of
the 14 events, seven were United States earthquakes (six in California) and
seven were from five different foreign countries. Event magnitudes range
from M6.5 to M7.9 with an average magnitude of M7.0.
FEMA P695 A: Ground Motion Record Sets A-21
Table A-6A Summary of Earthquake Event and Recording Station Data
for the Near-Field Record Set
M Year Name Name Owner
1 6.5 1979 Imperial Valley-06 El Centro Array #6 CDMG
2 6.5 1979 Imperial Valley-06 El Centro Array #7 USGS
3 6.9 1980 Irpinia, Italy-01 Sturno ENEL
4 6.5 1987 Superstition Hills-02 Parachute Test Site USGS
5 6.9 1989 Loma Prieta Saratoga - Aloha CDMG
6 6.7 1992 Erzican, Turkey Erzincan --
7 7.0 1992 Cape Mendocino Petrolia CDMG
8 7.3 1992 Landers Lucerne SCE
9 6.7 1994 Northridge-01 Rinaldi Receiving Sta DWP
10 6.7 1994 Northridge-01 Sylmar - Olive View CDMG
11 7.5 1999 Kocaeli, Turkey Izmit ERD
12 7.6 1999 Chi-Chi, Taiwan TCU065 CWB
13 7.6 1999 Chi-Chi, Taiwan TCU102 CWB
14 7.1 1999 Duzce, Turkey Duzce ERD
15 6.8 6.8 Gazli, USSR Karakyr --
16 6.5 1979 Imperial Valley-06 Bonds Corner USGS
17 6.5 1979 Imperial Valley-06 Chihuahua UNAMUCSD
18 6.8 1985 Nahanni, Canada Site 1 --
19 6.8 1985 Nahanni, Canada Site 2 --
20 6.9 1989 Loma Prieta BRAN UCSC
21 6.9 1989 Loma Prieta Corralitos CDMG
22 7.0 1992 Cape Mendocino Cape Mendocino CDMG
23 6.7 1994 Northridge-01 LA - Sepulveda VA USGS/VA
24 6.7 1994 Northridge-01 Northridge - Saticoy USC
25 7.5 1999 Kocaeli, Turkey Yarimca KOERI
26 7.6 1999 Chi-Chi, Taiwan TCU067 CWB
27 7.6 1999 Chi-Chi, Taiwan TCU084 CWB
28 7.9 2002 Denali, Alaska TAPS Pump Sta. #10 CWB
ID
No.
Earthquake Recording Station
No Pulse Records Subset
Pulse Records Subset
For each record, Table A-6B summarizes site and source characteristics,
epicentral distances, and various measures of site-source distance. Site
characteristics include shear wave velocity and the corresponding NEHRP
Site Class. Eleven sites are classified as Site Class D (stiff soil sites), fifteen
are classified as Site Class C (very stiff soil sites), and the remaining two are
classified as Site Class B (rock sites). Fourteen records are from events of
predominantly strike-slip faulting and the remaining fourteen records are
from events of predominantly thrust (or reverse) faulting.
A-22 A: Ground Motion Record Sets FEMA P695
Based on the average of Campbell and Boore-Joyner fault distances, the
minimum site-source distance is 1.7 km, the maximum distance is 8.8 km,
and the average distance is 4.2 km.
Table A-6B Summary of Site and Source Data for the Near-Field
Record Set
NEHRP
Class
Vs_30
(m/sec) Epicentral Closest to
Plane Campbell Joyner-
Boore
1 D 203 Strike-slip 27.5 1.4 3.5 0.0
2 D 211 Strike-slip 27.6 0.6 3.6 0.6
3 B 1000 Normal 30.4 10.8 10.8 6.8
4 D 349 Strike-slip 16.0 1.0 3.5 1.0
5 C 371 Strike-slip 27.2 8.5 8.5 7.6
6 D 275 Strike-slip 9.0 4.4 4.4 0.0
7 C 713 Thrust 4.5 8.2 8.2 0.0
8 C 685 Strike-slip 44.0 2.2 3.7 2.2
9 D 282 Thrust 10.9 6.5 6.5 0.0
10 C 441 Thrust 16.8 5.3 5.3 1.7
11 B 811 Strike-slip 5.3 7.2 7.4 3.6
12 D 306 Thrust 26.7 0.6 6.7 0.6
13 C 714 Thrust 45.6 1.5 7.7 1.5
14 D 276 Strike-slip 1.6 6.6 6.6 0.0
15 C 660 Thrust 12.8 5.5 5.5 3.9
16 D 223 Strike-slip 6.2 2.7 4.0 0.5
17 D 275 Strike-slip 18.9 7.3 8.4 7.3
18 C 660 Thrust 6.8 9.6 9.6 2.5
19 C 660 Thrust 6.5 4.9 4.9 0.0
20 C 376 Strike-slip 9.0 10.7 10.7 3.9
21 C 462 Strike-slip 7.2 3.9 3.9 0.2
22 C 514 Thrust 10.4 7.0 7.0 0.0
23 C 380 Thrust 8.5 8.4 8.4 0.0
24 D 281 Thrust 3.4 12.1 12.1 0.0
25 D 297 Strike-slip 19.3 4.8 5.3 1.4
26 C 434 Thrust 28.7 0.6 6.5 0.6
27 C 553 Thrust 8.9 11.2 11.2 0.0
28 C 553 Strike-slip 7.0 8.9 8.9 0.0
Pulse Records Subset
No Pulse Records Subset
ID
No.
Site Data Source
(Fault
Type)
Site-Source Distance (km)
For each record, Table A-6C summarizes key record information from the
PEER NGA database. This record information includes the record sequence
number and the file names of the two horizontal components as well as the
lowest frequency (longest period) for which frequency content is considered
fully reliable.
FEMA P695 A: Ground Motion Record Sets A-23
Maximum values of as-recorded peak ground acceleration, PGAmax, and peak
ground velocity, PGVmax, are given for each record. The term “maximum”
implies the larger peak ground velocity of the two components is reported.
Peak ground acceleration values range from 0.22 g to 1.43 g with an average
PGAmax of 0.60 g. Peak ground velocity values range from 30 cm/second to
167 cm/second with an average PGVmax of 84 cm/second.
Table A-6C Summary of PEER NGA Database Information and
Parameters of Recorded Ground Motions for the Near-
Field Record Set
FN Component FP Component
1 181 0.13 IMPVALL/H-E06_233 IMPVALL/H-E06_323 0.44 111.9
2 182 0.13 IMPVALL/H-E07_233 IMPVALL/H-E07_323 0.46 108.9
3 292 0.16 ITALY/A-STU_223 ITALY/A-STU_313 0.31 45.5
4 723 0.15 SUPERST/B-PTS_037 SUPERST/B-PTS_127 0.42 106.8
5 802 0.13 LOMAP/STG_038 LOMAP/STG_128 0.38 55.6
6 821 0.13 ERZIKAN/ERZ_032 ERZIKAN/ERZ_122 0.49 95.5
7 828 0.07 CAPEMEND/PET_260 CAPEMEND/PET_350 0.63 82.1
8 879 0.10 LANDERS/LCN_239 LANDERS/LCN_329 0.79 140.3
9 1063 0.11 NORTHR/RRS_032 NORTHR/RRS_122 0.87 167.3
10 1086 0.12 NORTHR/SYL_032 NORTHR/SYL_122 0.73 122.8
11 1165 0.13 KOCAELI/IZT_180 KOCAELI/IZT_270 0.22 29.8
12 1503 0.08 CHICHI/TCU065_272 CHICHI/TCU065_002 0.82 127.7
13 1529 0.06 CHICHI/TCU102_278 CHICHI/TCU102_008 0.29 106.6
14 1605 0.10 DUZCE/DZC_172 DUZCE/DZC_262 0.52 79.3
15 126 0.06 GAZLI/GAZ_177 GAZLI/GAZ_267 0.71 71.2
16 160 0.13 IMPVALL/H-BCR_233 IMPVALL/H-BCR_323 0.76 44.3
17 165 0.06 IMPVALL/H-CHI_233 IMPVALL/H-CHI_323 0.28 30.5
18 495 0.06 NAHANNI/S1_070 NAHANNI/S1_160 1.18 43.9
19 496 0.13 NAHANNI/S2_070 NAHANNI/S2_160 0.45 34.7
20 741 0.13 LOMAP/BRN_038 LOMAP/BRN_128 0.64 55.9
21 753 0.25 LOMAP/CLS_038 LOMAP/CLS_128 0.51 45.5
22 825 0.07 CAPEMEND/CPM_260 CAPEMEND/CPM_350 1.43 119.5
23 1004 0.12 NORTHR/0637_032 NORTHR/0637_122 0.73 70.1
24 1048 0.13 NORTHR/STC_032 NORTHR/STC_122 0.42 53.2
25 1176 0.09 KOCAELI/YPT_180 KOCAELI/YPT_270 0.31 73.0
26 1504 0.04 CHICHI/TCU067_285 CHICHI/TCU067_015 0.56 91.8
27 1517 0.25 CHICHI/TCU084_271 CHICHI/TCU084_001 1.16 115.1
28 2114 0.03 DENALI/ps10_199 DENALI/ps10_289 0.33 126.4
No Pulse Records Subset
Pulse Records Subset
ID
No.
PEER-NGA Record Information Recorded Motions
Record
Seq. No.
Lowest
Freq (Hz.)
File Names - Horizontal Records PGAmax
(g)
PGVmax
(cm/s.)
A-24 A: Ground Motion Record Sets FEMA P695
For each record, Table A-6D summarizes the 1-second spectral acceleration
(both horizontal components), peak ground velocity reported in the PEER
database, PGVPEER (i.e., single value of peak ground velocity based on the
geometric mean of rotated components), normalization factors (NMi), and
values of PGAmax and PGVmax after normalization by PGVPEER.
Normalization is done separately for Pulse subset and No-Pulse subsets.
Table A-6D Summary of Factors Used to Normalize Recorded
Ground Motions, and Parameters of Normalized
Ground Motions for the Near-Field Record Set
FN Comp. FP Comp.
1 0.43 0.60 83.9 0.90 0.40 100.1
2 0.66 0.64 78.3 0.96 0.44 104.4
3 0.25 0.41 43.7 1.72 0.53 78.2
4 0.97 0.51 71.9 1.04 0.44 111.6
5 0.47 0.32 46.1 1.63 0.62 90.6
6 0.98 0.37 68.8 1.09 0.53 104.2
7 0.92 0.70 69.6 1.08 0.68 88.6
8 0.43 0.34 97.2 0.77 0.62 108.4
9 1.96 0.47 109.3 0.69 0.59 114.9
10 0.89 0.65 94.4 0.80 0.58 97.7
11 0.29 0.28 26.9 2.79 0.62 83.2
12 1.33 1.10 101.6 0.74 0.60 94.4
13 0.60 0.58 87.5 0.86 0.25 91.5
14 0.54 0.73 69.6 1.08 0.56 85.6
15 0.81 0.42 65.0 0.86 0.61 61.4
16 0.44 0.44 49.8 1.13 0.86 49.8
17 0.41 0.37 28.2 1.99 0.56 60.8
18 0.53 0.29 44.1 1.27 1.50 55.9
19 0.16 0.29 28.7 1.95 0.87 67.8
20 0.55 0.45 49.0 1.15 0.73 64.0
21 0.53 0.50 47.9 1.17 0.60 53.3
22 0.42 0.73 84.4 0.66 0.95 79.4
23 0.62 1.00 72.6 0.77 0.56 54.2
24 0.81 0.40 47.7 1.18 0.50 62.6
25 0.38 0.35 62.4 0.90 0.28 65.6
26 0.75 0.75 72.3 0.78 0.44 71.3
27 2.54 0.86 90.3 0.62 0.72 71.5
28 0.69 0.82 98.5 0.57 0.19 72.0
1-Sec.Spec. Acc. (g) PGVPEER
(cm/s.)
Pulse Records Subset
No Pulse Records Subset
ID
No.
As-Recorded Parameters Normalization
Factor
Normalized Motions
PGAmax
(g)
PGVmax
(cm/s.)
FEMA P695 A: Ground Motion Record Sets A-25
Normalization factors vary from 0.57 to 2.79. After normalization, peak
ground acceleration values range from 0.19 g to 1.50 g with an average
PGAmax of 0.60 g. Peak ground velocity values range from 50 cm/second to
115 cm/second with an average PGVmax of 80 cm/second. Table A-7 shows
that normalization of the records (by PGVPEER) has substantially reduced the
dispersion in PGVmax without greatly affecting average values of PGAmax or
PGVmax, or the dispersion in PGAmax.
Table A-7 Near-Field Record Set (As-Recorded and After
Normalization): Comparison of Maximum, Minimum
and Average Values of Peak Ground Acceleration
(PGAmax) and Peak Ground Velocity (PGVmax),
Respectively
Parameter
Value
PGAmax (g) PGVmax (g)
As-Recorded Normalized As-Recorded Normalized
Maximum 1.43 1.50 167 115
Minimum 0.22 0.19 30 50
Max/Min Ratio 6.5 7.9 5.6 2.3
Average 0.60 0.60 84 80
Figure A-6 shows response spectra of individual records of the normalized
Near-Field record set, as well as the median and one and two standard
deviation spectra of the set in log format. Figure A-7 shows median and onestandard
deviation spectra, as well as a plot of the standard deviation (natural
log of spectral acceleration) of the normalized Near-Field record set in linear
format.
The median spectrum of the normalized Near-Field record set has frequency
content consistent with a large magnitude (M7) event recorded relatively
close (5 km) to the fault rupture. The median spectral acceleration at short
periods is about 1.0 g and median spectral acceleration at 1-second is about
0.6 g. Record-to-record variability (standard deviation of the logarithm of
spectral acceleration) ranges from about 0.4-0.5 at short periods to about 0.6
at long periods, generally consistent (except at very short periods) with the
values of dispersion from common ground motion (attenuation) relations
(e.g., see Figure 8, Campbell and Borzorgnia, 2003).
A-26 A: Ground Motion Record Sets FEMA P695
0.01
0.1
1
10
0.01 0.1 1 10
Period (seconds)
Spectral Acceleration (g)
Median Spectrum - Near-Field Set
+1 Logarithmic St. Dev. Spectrum
+2 Logarithmic St. Dev. Spectrum
Figure A-6 Response spectra of the fifty-six individual components of
the normalized Near-Field record set, and median, one
and two standard deviation response spectra of the total
record.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Period (seconds)
Spectral Acceleration (g)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Standard Deviation of Ln(Sa)
Median Spectrum - Near-Field Set
+1 Logarithmic St. Dev. Spectrum
Logarithmic St. Dev.
Figure A-7 Median and one standard deviation response spectra of
the normalized Near-Field record set, and plot of the
standard deviation (natural log) of response spectral
acceleration.
Figure A-8 is a plot of the median value of epsilon calculated for the fifty-six
components of the Near-Field record set. For comparison, Figure A-8 also
shows the median value of . calculated for the Far-Field record (from Figure
A-5). These values are based on the attenuation relationship developed by
FEMA P695 A: Ground Motion Record Sets A-27
Abrahamson and Silva (1997). Median values of . are generally small and
near zero at all periods, with mildly positive . values occurring at periods of
2.5 to 4.5 seconds, indicating that the Near-Field record set is approximately
“epsilon neutral.” Accordingly, collapse margin ratios (CMRs) based on
incremental dynamic analysis using these records do not account for the
effects of spectral shape, discussed in Section A.4, and are later increased to
account for these effects, using the factors of Section 7.4 and Appendix B.
-2.0
-1.6
-1.2
-0.8
-0.4
0.0
0.4
0.8
1.2
1.6
2.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Period (seconds)
Median Epsilon(T)
Far-Field Set
Near-Field Set
Figure A-8 Median value of epsilon for the Near-Field record set (and plot
of median value of epsilon for the Far-Field record set for
comparison).
A.11 Comparison of Far-Field and Near-Field Record
Sets
This section compares median response spectra of the Far-Field and Near-
Field record sets and collapse margin ratios (CMRs) calculated using these
record sets. Comparisons of collapse margin ratios are made to determine if
margins are, in general, substantially less for SDC E structures subjected to
near-fault seismic demands than for SDC D structures subjected to SDC Dmax
seismic demands. These comparisons necessarily consider that higher
seismic loads are required for design of SDC E structures (than for SDC D
structures).
The three record sets used in this section to evaluate collapse margin are: (1)
Far-Field record set (full set of 44 records); (2) Near-Field record set (full set
of 56 records); and (3) Near-Field subset of pulse records in the fault normal
(FN) direction (14 Near-Field records that are oriented in the fault normal
direction and which have pulses). All records are normalized and scaled as
A-28 A: Ground Motion Record Sets FEMA P695
described in Section A.8. These sets permit assessment of the effect of
differences in frequency content and response characteristics of records on
collapse margin. For example, are collapse margins similar for Far-Field and
Near-Field record sets (when both sets are anchored to the same level of SDC
E seismic criteria)? Similarly, are collapse margins substantially lower for
Near-Field FN-Pulse records than for the full Near-Field record set?
Figure A-9 shows unscaled median response spectra for Near-Field and Far-
Field record sets, respectively, and the ratio of these spectra. The ratio of
median spectra varies from about 1.2 at short periods to about 1.6 at a period
of 1-second (and over 2.0 at periods beyond 2 seconds) consistent with
mapped values of ground motion required by ASCE/SEI 7-05, and other
codes, for structural design near active sources. For example, the near-source
coefficients of the 1997 UBC (ICBO, 1997), summarized in Table A-8,
increase seismic design loads by 1.2 in acceleration domain and 1.6 in
velocity domain for structures located 5 km of an active fault capable of
generating large magnitude earthquakes. Corresponding increases in MCE
seismic criteria are used for evaluation of collapse margin for SDC E seismic
demands.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Period (seconds)
Spectral Acceleration (g) 0.0
0.4
0.8
1.2
1.6
2.0
2.4
2.8
3.2
3.6
4.0
Spectra Ratio
Median Spectrum - Near-Field Record Set
Median Spectrum - Far-Field Record Set
Spectra Ratio - Near-Field/Far-Field
Figure A-9 Median response spectra of normalized Near-Field and
Far-Field record sets, and the ratio of these spectra.
FEMA P695 A: Ground Motion Record Sets A-29
Table A-8 Near-Source Coefficients of the 1997 UBC (from Tables 16-
S and 16-T, ICBO, 1997)
Spectral
Domain
Closest Distance to Fault
=2 km 5 km 10 km .15 km
Acceleration 1.5 1.2 1.0 1.0
Velocity 2.0 1.6 1.2 1.0
Figure A-10 compares unscaled median response spectra of the Far-Field
record set, the full Near-Field record set, and the fault normal records of the
Near-Field Pulse subset. At short periods, the three response spectra are
similar, but they diverge significantly at long periods. As expected, at long
periods, the median response spectrum of fault normal records of the Near-
Field Pulse subset shows substantially greater seismic demand than both the
Near-Field and Far-Field record sets.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Period (seconds)
Spectral Acceleration (g)
Median - Near-Field Pulse (FN) Subset
Median - Near-Field Record Set
Median - Far-Field Record Set
Figure A-10 Median response spectra of the normalized Near-
Field record set, the Far-Field record set, and fault
normal records of the Near-Field Pulse subset.
Potential differences in collapse margin ratio due to differences in the
frequency content of the three record sets are investigated using archetypes
of reinforced-concrete special moment-frame structures (i.e., archetypes of
Chapter 9, perimeter frame and 30-foot bay configuration). Three archetype
heights are considered: 1-story, 4-story and 20-story. In the case of the 20-
story archetype, two designs are prepared. One design, 20-Story–NL, is
without lower-bound limits on design base shear (i.e., design ignores
Equation 12.8-6 of ASCE/SEI 7-05). The other design, 20-Story-02, has
A-30 A: Ground Motion Record Sets FEMA P695
lower-bound base shear limits, including also the base shear limit of
Equation 9.5.5.2.1-3 of ASCE 7-02 (ASCE, 2003).
Each reinforced concrete special moment frame archetype is designed in
accordance with ASCE/SEI 7-05 criteria for both SDC Dmax and SDC E
design requirements (i.e. one design is prepared for SDC Dmax and another
for SDC E), respectively, with certain exceptions for 20-story archetypes.
Note that the SDC E demands were not previously defined in Table A-1,
since SDC E is not used in the basic assessment Methodology of this
document; SDC E is only used for this comparison, so the SDC E demands
are given in Table A-9.
Table A-9 summarizes fundamental periods and seismic design coefficients
for each of the eight archetypes. As shown, the seismic coefficients of
archetypes designed for SDC E requirements are 20% to 60% greater than
those of archetypes designed for SDC Dmax requirements.
Table A-9 Summary of Key Reinforced-Concrete Special Moment Frame
Archetype Properties and Seismic Coefficients Used to Evaluate
the Collapse Margin Ratio (CMR)
Building Archetype Seismic Coefficient, Cs
Height - ID T (sec.) Cs Limits SDC Dmax SDC E Ratio
1-Story 0.26 NA 0.125 0.15 1.2
4-Story 0.81 NA 0.092 0.149 1.6
20-Story - 02 3.36 ASCE/SEI 7-021 0.0441 0.0532 1.2
20-Story - NL 3.36 None3 0.022 0.036 1.6
1. Archetype design includes lower-bound limit of ASCE 7-02, Equation 9.5.5.2.1-3.
2. Archetype design slightly less than Cs = 0.06 limit of ASCE/SEI 7-05, Equation 12.8-6.
3. Archetype design ignores lower-bound limit of ASCE/SEI 7-05, Equation 12.8-6.
Tables A-10A and A-10B summarize collapse margin ratios (CMR’s) for the
eight archetypes of Table A-9. In both tables, CMRs of the four archetypes
designed and evaluated for SDC Dmax (far-field) sites are compared with
CMR’s of the same four archetypes designed and evaluated for SDC E (nearfield)
sites.
Table A-10A compares two CMR values computed for SDC E sites, against
the baseline CMR values (for SDC Dmax design assessed using the far-field
ground motion set). The SDC E CMR values are computed both (a) using the
Far-Field ground motion set, and (b) using the Near-Field ground motion set.
Both sets are anchored to SDC E demands (from Table A-9) according to
section A.8. Note that the near-field record set should be used when
FEMA P695 A: Ground Motion Record Sets A-31
assessing performance at a SDC E (near-field) site; the far-field set is only
used for comparison and to help explain the observed differences between
the computed CMR values for SDC E and SDC Dmax sites.
Table A-10A Summary of Selected Collapse Margin Ratios (CMRs) for
Reinforced-Concrete Special Moment Frame Archetypes –
Comparison of CMRs for Far-Field and Near-Field Record Sets
Building
Archetype
Far-Field
Baseline (Sdc
Dmax)
Near-Field Designs Evaluated For Sdc E Seismic
Demand (Mce)
Near-Field Record Set Far-Field Record Set
Height - ID CMR CMR CMR/
Baseline
CMR CMR/
Baseline
1-Story 1.26 0.86 68% 1.03 82%
4-Story 1.98 1.32 67% 1.55 78%
20-Story - 02 1.62 1.34 83% 1.39 86%
20-Story - NL 0.82 1.01 123% 0.91 111%
Table A-10A shows the CMR values for SDC E sites, computed using both
the Far-Field and Near-Field record sets. When the Near-Field record set is
used for the SDC E evaluation, the CMR is an average of 30% lower as
compared to SDC Dmax (with the exception of the 20-Story-NL archetype,
which will be discussed later).
This 30% difference is caused by two aspects: (1) the higher seismic demand
of SDC E (it has been shown in Chapter 9 that CMR values are typically
lower for buildings designed and assessed for higher seismic demands) and
(2) the use of the Near-Field record set instead of the Far-Field set. To
clearly separate these two effects, columns two and five show the CMR
values for SDC Dmax and SDC E, respectively, both evaluated using the Far-
Field record set. This shows that the CMR values are an average of 20%
lower for SDC E, simply due to SDC E having higher seismic demand. This
shows that the use of Near-Field records, rather than Far-Field records, only
leads to an average 10% reduction in the CMR.
Table A-10B shows how low the CMR could become if one used only the
subset of Near-Field motions which are fault-normal and have pulses.
However, this is not recommended for performance evaluation. For
comparison, this table also replicates the results from the full Near-Field set.
This table shows that the resulting CMR values are an average of 45% lower
than the baseline SDC Dmax case, and 15% lower than the values computed
when the full Near-Field set is utilized.
A-32 A: Ground Motion Record Sets FEMA P695
Table A-10B Summary of Selected Collapse Margin Ratios (CMRs) for
Reinforced-Concrete Special Moment Frame Archetypes -
Comparison of CMRs for the Near-Field Record Set and the
Near-Field Pulse FN Record Subset
Building
Archetype
Far-Field
Baseline
(SDC Dmax)
Near-Field Designs Evaluated for SDC E Seismic
Demand (MCE)
Near-Field Record Set NF - Pulse FN Subset
Height - ID CMR CMR CMR/
Baseline
CMR CMR/
Baseline
1-Story 1.26 0.86 68% 0.67 53%
4-Story 1.98 1.32 67% 0.99 50%
20-Story - 02 1.62 1.34 83% 0.92 57%
20-Story - NL 0.82 1.01 123% 0.70 85%
The 20-Story-NL archetype is the only building that does not follow the
trends described in the preceding discussion. One possible reason for this is
that the minimum base shear limit was not imposed in this design, which
caused the design strengths to become very low (Cs = 0.022 g to 0.036 g),
and the collapse capacity of extremely weak structures becomes more
sensitive to changes in design strength. Another possible contributing factor
is that the fundamental period, T1, is longer for this design, and the spectral
shapes change slightly for periods above 3.5 seconds (see Figure A-9).
In summary, as compared to Far-Field motions, when a structure is subjected
to the full set of Near-Field records, the CMR is typically 10% lower.
However, when a structure is subjected to the subset of FN pulse-type
records, the CMR is decreased by about 25%. This shows that the increase in
the seismic response coefficient, Cs, required for design of structures located
near faults does not appear sufficient to result in performance comparable to
that of the same system (i.e., same R factor) located further away from fault
rupture.
Tables A-10A and A-10B also show that without lower-bound limits on
design base shear, collapse margins for the 20-Story-NL archetype are very
low; but with lower-bound limits, collapse margins for the 20-Story-02
archetype are approximately the same as the 4-Story archetype. While the
margins for the 20-Story-NL archetype are very low, it should be noted that a
60% increase in SDC E design strength affects comparable or better
performance than the baseline (SDC D) archetype. In fact, a 60% increase in
design base shear is able to cause about the same collapse margin for the 20-
Story-NL (SDC E) archetype, when evaluated with fault normal components
of the Near-Field pulse records, as that of the 20-Story-NL archetype (SDC
D) evaluated with Far-Field records (Table A-10B).
FEMA P695 A: Ground Motion Record Sets A-33
A.12 Robustness of Far-Field Record Set
The purpose of this Methodology is to assess the collapse safety of newly
proposed structural systems. To this end, the record selection criteria of
section A.7 result in an appropriate set of strong ground motions that are
representative of those that may cause collapse of a new building (to the
extent possible, given the limited number of recordings available).
This section evaluates the “robustness” of the ground motion selection
criteria to demonstrate that the final collapse capacity predictions are not
highly sensitive to small changes to the selection criteria. Specifically, the
peak ground acceleration (PGA) > 0.2g and peak ground velocity (PGV) >
15cm/s limits are evaluated, as well as the requirement that the two highest
PGV records be selected when there are many candidate records for an event.
The other selection criteria are not investigated here because they are less
subjective, or in the case of magnitudes larger than M6.5, small changes to
this requirement are expected to have minimal impact on the collapse
capacity predictions.
A.12.1 Approach to Evaluating Robustness
To investigate the effects of the PGA and PGV selection criteria, a set of 192
records is selected based on liberal criteria (PGA > 0.05g and PGV > 3cm/s).
To reduce the number of ground motions, and still maintain the selection
criteria, 20 pairs of motions are randomly selected from each event. This
affected the records from Chi-Chi (261 candidate records), Loma Prieta (53
records), Northridge (70 records), and Landers (24 candidate records).
To evaluate the quantitative differences between ground motion sets, each set
is used to predict the median collapse capacity (the CMR and ACMR) of three
reinforced concrete special moment frame building archetypes. To cover the
range of structural periods, the following building archetypes are utilized in
these comparisons: 1-story reinforced concrete special moment frame (T =
0.26 s), 4-story reinforced concrete special moment frame (T = 0.81 s), and a
12-story reinforced concrete special moment frame (T = 2.13 s). These are
three archetypes are taken from the reinforced concrete special moment
frame example of Section 9.2. Each archetype is a perimeter-frame systems
designed for SDC Dmax seismic criteria. Table A.11 summarizes design
properties and collapse margin results for these archetypes.
A-34 A: Ground Motion Record Sets FEMA P695
Table A-11 Summary of Design Properties and Collapse Margins of the
Three Reinforced Concrete Special Moment Frame Building
Archetypes Used to Evaluate Far-Field Record Set Robustness
Archetype
Design
ID No.
Design Configuration and Properties Collapse Margin
No. of
Stories
Framing
Type
Seismic
SDC
Period T
(sec.)
CMR ACMR
20611 1 P Dmax 0.26 1.96 2.61
10031 4 P Dmax 0.81 1.61 2.27
10132 12 P Dmax 2.13 1.45 2.33
1. Design data and collapse margins taken from Tables 9-2, 9-3 and 9-8.
2. Design data and collapse margins taken from Tables 9-9 and 9-10.
A.12.2 Effects of PGA Selection Criteria Alone
Table A-12 shows the effects that the minimum PGA requirement has on the
computed values of CMR. This table includes results for PGA > 0.05g (192
records), PGA > 0.10g (105 records), PGA > 0.15g (62 records), and PGA >
0.20g (36 records). These results clearly show that the CMR values increase
as the minimum PGA limit is increased. Comparing the PGA limit of 0.05g
and 0.20g, the CMR increases 31% for the 1-story building, 19% for the 4-
story building, and 10% for the 12-story building, with an overall average
value of 20%. The PGA limit, unsurprisingly, has a more significant effect
on shorter-period structures.
Table A-12 Effects of the PGA Selection Criterion on the Computed CMR
Values for Three Reinforced Concrete Special Moment Frame
Buildings
Selection Criteria and Number of Selected Records
Min PGA (g) 0.05 0.10 0.15 0.20
Min PGV (cm/s) 3.0 3.0 3.0 3.0
No. of Records 192 105 62 36
Arch. ID No. CMR
2061 1.76 1.97 2.24 2.31
1003 1.26 1.36 1.39 1.51
1013 1.41 1.46 1.49 1.56
Mean Value 1.48 1.60 1.71 1.79
However, the CMR is not the appropriate parameter for use in a complete
comparison of the ground motion sets. The Adjusted Collapse Margin Ratio
(ACMR) value should be used because it is the complete collapse capacity
result after accounting for the effects of spectral shape. In order to compute
the ACMR values, we must compute the Spectral Shape Factor values, as
FEMA P695 A: Ground Motion Record Sets A-35
discussed in Appendix B. The following discussion provides the rationale
for approximate SSF values that can be used to compute approximate values
of ACMR for this comparison.
Appendix B will show that the SSF value is, in part, based on the mean e
values of the ground motion set. Figure A-11 shows these mean e values for
the Far-Field ground motion set, and it can be reasoned that the mean e
values of the PGA > 0.20 g set (36 records) should be similar due to the
similarities in the selection criteria of the two sets. In contrast, the PGA >
0.05 g (192 records) set is more representative of all possible records, so the
mean e values is expected to be approximately zero at all periods. This
observation suggests that the high PGA limit used for the Far-Field set is
causing the observed mean e = 0.6 at short periods.
0 0.5 1 1.5 2 2.5 3 3.5 4
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Period (T) [sec]
Mean of .(T),records
Basic Far-Field Ground Motion Set (22x2 records)
Simplified Trend
Figure A-11 Mean e values, . (T),records , for the Far-Field ground
motion set (duplicated from Figure B-3 of App. B).
Using the above rationale, Table A-13 presents the approximate SSF and
ACMR values for the sets selected with PGA > 0.05 g (192 records) and PGA
> 0.20 g (36 records). This shows that the differences in the mean e values
counteract much of the trend seen with the CMR values. For the three
buildings considered here, the ACMR value increases by an average of 8%
when the ground motions are selected based on PGA > 0.20 g rather than a
more liberal selection using PGA > 0.05 g.
A-36 A: Ground Motion Record Sets FEMA P695
Table A-13 Effects of the PGA Selection Criterion on the Approximate
ACMR Values for Three Reinforced Concrete Special
Moment Frame Buildings
Selection Criteria and Number of Selected Records
Min PGA (g) 0.05 0.20 0.05 0.20
Min PGV (cm/s) 3.0 3.0 3.0 3.0
No. of Records 192 36 192 36
Arch. ID No. approx. SSF approx. ACMR
2061 1.61 1.33 2.83 3.07
1003 1.61 1.41 2.04 2.12
1013 1.61 1.61 2.27 2.51
Mean Value 1.61 1.45 2.38 2.57
A.12.3 Effects of PGV Selection Criteria Alone
Similar to previous PGA comparisons, Table A-14 shows the effect that the
minimum PGV requirement has on the CMR. This shows that the impacts of
the PGV requirement vary for the three buildings considered. Comparing the
PGV limit of 3 cm/s and 15 cm/s, the CMR decreases 14% for the 1-story
building, decreases 2% for the 4-story building, and increases 5% for the 12-
story building, with an overall average decrease of 5%.
Table A-14 Effects of the PGV Selection Criterion on the Computed
CMR Values for Three RC SMF Buildings
Selection Criteria and Number of Selected Records
Min PGA (g) 0.05 0.05 0.05 0.05
Min PGV (cm/s) 3.0 5.0 10.0 15.0
No. of Records 192 186 130 85
Arch. ID No. approx. SSF approx. ACMR
2061 1.76 1.73 1.59 1.50
1003 1.26 1.26 1.20 1.24
1013 1.41 1.42 1.47 1.48
Mean Value 1.48 1.47 1.42 1.41
For evaluating the PGV selection criterion, comparing the CMR value is
sufficient (comparison of ACMR is not needed) because the PGV value is
expected to affect the mean e values at moderate or long periods. Since
Figure A-11 showed that even the Far-Field set (selected for high PGV > 15
cm/s) has e = 0 at moderate/long periods, it is expected that the other larger
ground motion sets should also have e = 0 at those periods. Therefore, the
SSF values should approximately equal for each of the record sets shown in
Table A-14.
FEMA P695 A: Ground Motion Record Sets A-37
A.12.4 Effects of both PGA and PGV Selection Criteria
Simultaneously, as well as Selection of Two Records
from Each Event
The previous sections investigated the minimum PGA and PGV selection
criteria individually, and showed that they have somewhat counteracting
effects on the resulting collapse capacity predictions (CMR and ACMR). To
create the Far-Field record set, these two criteria are imposed simultaneously,
and then two records are selected from each event. To select the two records
from each event, the records with highest PGV are used.
Table A-15 shows results of this progression to obtain the final Far-Field
ground motion set (final Far-Field set results shown in bold-italic). The three
ground motion sets used to produces the collapse results shown in this table
were selected as follows:
Set 1. Set selected using liberal selection criteria, with PGA > 0.05 g and
PGV > 3 cm/s, and 20 records selected randomly from each event (192
records).
Set 2. Set selected from the above set of 192, but with strict imposed
selection criteria of PGA > 0.20g and PGV > 15 cm/s (32 records).
Set 3. Set selected using the strict selection criteria, of PGA > 0.20 g and
PGV > 15 cm/s, and two records with the highest PGV selected from each
event (22 records). This is the final Far-Field record set used in the
Methodology. Note that due to slight differences in the scaling method, the
CMR values do not match exactly with those reported elsewhere in this
report, so relative comparisons should be made of values within this table.
The results in Table A-15 follow directly from the trends shown previously
for PGA and PGV criteria alone. Comparing Set 1 and Set 2, the CMR value
increased by an average of 18% (CMR = 1.48 versus 1.74), which is driven
mostly by the effects of the PGA selection criterion and the resulting
differences in spectral shape. The selection of two high-PGV records from
each event make the difference reduce to 12% when comparing Set 1 to Set 3
(CMR = 1.48 versus 1.65).
A-38 A: Ground Motion Record Sets FEMA P695
Table A-15 Effects of the PGA and PGV Selection Criteria on the
Computed CMR and ACMR Values, for Three Reinforced
Concrete Special Moment Frame Buildings, as well as The
Effects of Selecting Two Records from Each Event
Selection Criteria and Number of Selected Records
Record Set No. 1 2 3 1 2 3 1 2 3
Min PGA (g) 0.05 0.20 0.20 0.05 0.20 0.20 0.05 0.20 0.20
Min PGV (cm/s) 3.0 15.0 15.0 3.0 15.0 15.0 3.0 15.0 15.0
Record Selection1 Rand Rand High Rand Rand High Rand Rand High
No. of Records 192 32 22 192 32 22 192 32 22
Arch. ID No. CMR approx. SSF approx. ACMR
2061 1.76 2.06 1.94 1.61 1.33 1.33 2.83 2.73 2.58
1003 1.26 1.53 1.43 1.61 1.41 1.41 2.04 2.16 2.01
1013 1.41 1.62 1.59 1.61 1.61 1.61 2.27 2.61 2.56
Mean Value 1.48 1.74 1.65 1.61 1.45 1.45 2.38 2.50 2.38
1. "Rand" indicates random selection from candidate records and "High" indicates biased
selection from candidate records based on highest PGV values (Sec. A.7).
As with the previous comparisons of the PGA selection criterion, the
differences in the SSF value account for most of the above differences in
CMR, and cause the differences in ACMR to be much smaller. Comparing
Set 1 and Set 2, ACMR increased by only an average of 5% (CMR = 2.38
versus 2.50). The selection of the two high-PGV records from each event
eliminate the difference when comparing the Set 1 to Set 3 (CMR = 2.38
versus 2.38). Note that this average difference of 0% is rather coincidental,
and the individual differences are non-zero, being -9%, -1%, and +13%, for
the three buildings considered. Even the individual 1% to 13% differences in
median collapse capacity are very small when comparing two sets of ground
motions that were selected based on such different selection criteria.
A.12.5 Summary of the Robustness of the Far-Field Set
ACMR collapse capacity results are not highly sensitive to the PGA and PGV
selection criteria or to the approach used to select the two motions from each
event. More specifically, when comparing the Far-Field set (Set 3) to a set of
records selected with much looser PGA and PGV criteria, the ACMR values
vary between -9% to +13% for the three buildings considered, with an
average difference of 0% which is partially coincidence when using such a
small sample size. Relatively speaking, these differences of 0% to 13% are
exceptionally small when comparing two ground motion sets selected using
such differing criteria. This shows that the Far-Field ground motion set can
be considered robust with respect to these selection criteria.
FEMA P695 A: Ground Motion Record Sets A-39
A.13 Assessment of Record-to-Record Variability in
Collapse Fragility
According to the Methodology, the ground motion record set is scaled as
described in Section A.8 in order to evaluate the median collapse capacity,
ˆ . CT S To assess the collapse probability and accordingly the ACMR
acceptance criteria, an estimate of the record-to-record variability in collapse
capacity (ßRTR) is also needed. ßRTR is the lognormal standard deviation
associated with uncertainty in response of an archetype structure due to
differences in frequency content and other characteristics of ground motion
records. Through the course of this project, it has become evident that ßRTR
values are relatively consistent across different types of structures, so a fixed
value of ßRTR = 0.40 is used in the Methodology for most structures. This
section validates the choice of the 0.40 value, but shows that for structures
that have limited period elongation before collapse, ßRTR < 0.40 is appropriate
and provides guidance on how to determine ßRTR for these structures.
Record-to-record variability in the collapse fragility for a particular structure
can be computed from the results of incremental dynamic analysis of the
ground motion record set. The value of ßRTR obtained will depend on the
scaling and normalization of the records. Two normalization and scaling
methods are considered:
1. The PGV normalization method developed in this Methodology and
described in Section A.8 normalizes the records by PGV and then scales
the normalized record set as a group, in order to determine the median
collapse capacity. An illustration of the scaled Far-Field record set is
shown in Figure A-3.
2. An alternative method, termed Sa-component scaling, scales records
according to the component spectral acceleration of each record
individually evaluated at the period of the building. When this method is
utilized, all records have exactly the same spectral acceleration at the
period of interest. This is in contrast to the PGV normalization method,
which maintains variability at all periods. This method is commonly
used in research applications and is conceptually appealing because each
individual record individually matches the target Sa level. In addition,
this scaling method avoids double counting of uncertainties in prediction
of the mean annual frequency of collapse, a frequently used metric of
seismic performance.
The advantage of the PGV normalization method is that the entire record set
can be scaled by the same scale factor, shown in Table A-3. This
significantly reduces the complexity of the calculations for the user of this
A-40 A: Ground Motion Record Sets FEMA P695
Methodology. Sa-component scaling requires computing different scale
factors for each record at each different period of interest.
Differences between the PGV normalization method and the Sa-component
scaling method are illustrated in Figures A-12 and A-13. Figure A-12
compares the median of the record sets according to the two scaling methods.
The median spectral acceleration of the records at all periods is virtually the
same under the two scaling methods. Also, the median of the set under the
Sa-component method does not depend on the period at which the scaling
occurs. For this reason, it can be shown that the computed median collapse
capacity will be the same regardless of which scaling procedure is used (see
Appendix 6B of Haselton and Deierlein, 2007).
0 1 2 3 4 5
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Period [seconds]
Sa [g]
Normalization by PGV
Scaling to Sa(0.25s)
Scaling to Sa(0.50s)
Scaling to Sa(1.00s)
Scaling to Sa(2.00s)
Figure A-12 Comparison of median spectra under the FEMA P695
normalization/scaling method and the Sa-component
scaling method (scaled to four different periods: 0.25 s,
0.50 s, 1.0 s, and 1.2 s). The five curves on this plot
are virtually indistinguishable.
Figure A-13 compares the logarithmic standard deviation of the spectral
ordinates, sLN,Sa, for different scaling methods. The PGV normalization
method gives relatively constant values of the sLN,Sa, regardless of the period
of interest. For the Sa-component method, sLN,Sa is zero at the period of
interest, since all records have the same spectral ordinate. Away from the
scaling period, sln,Sa values obtained from Sa-component scaling are similar to
those obtained in the FEMA P695 approach.1 Values of sLN,Sa obtained from
1 The sLN,Sa values obtained by Sa-component scaling are notably larger at long
periods away from the scaling period. The PGV normalization avoids this high
uncertainty. Regardless, sLN,Sa at long periods are not expected to significantly
FEMA P695 A: Ground Motion Record Sets A-41
Sa-component scaling indicate that the .RTR for brittle structures, which
collapse at periods close to the period of the building, T, is expected to be
smaller than the .RTR for structures that undergo significant period
elongations before collapse. For brittle systems, Sa-component scaling
avoids double-counting of record-to-record variability that occurs in the PGV
normalization method due to the enforced variation in the record set at all
periods.
0 1 2 3 4 5
0
0.2
0.4
0.6
0.8
1
Period [seconds]
.
LN(Sa)
Normalization by PGV
Scaling to Sa(0.25s)
Scaling to Sa(0.50s)
Scaling to Sa(1.00s)
Scaling to Sa(2.00s)
Figure A-13 Comparison of logarithmic standard deviation of
spectral ordinates, s LN,Sa, for FEMA P695 scaling
method and Sa-component scaling method at
different periods.
To investigate the relationship between ßRTR and the building period
elongation prior to collapse, ßRTR values were computed using the Sacomponent
scaling method for the reinforced concrete special moment
frames and ordinary moment frames discussed in Chapter 9. Note that some
ordinary moment frame data include hypothetical brittle non-simulated
failure modes in order to get data points for very brittle systems. Period
elongation is quantified by the period-based ductility parameter, .T, defined
as the ratio of ultimate roof drift to the yield roof drift: .T = .u
/.y,eff (see
Appendix B and Chapter 7 for more details; period elongation prior to
collapse would approximately be the square-root of this ratio).
The relationship between ßRTR and period-based ductility, .T, is shown in
Figure A-14. The reinforced concrete special moment frames have an
average period-based ductility, .T = 9, and average ßRTR = 0.4 in the collapse
fragility. Because these structures undergo significant period elongation
impact predicted structural response because the periods with high sln,Sa are far
away from the scaling period (the elastic period of the building).
A-42 A: Ground Motion Record Sets FEMA P695
before collapse, they respond away from the pinched region of the spectrum.
For reinforced concrete ordinary moment frames, which undergo less period
elongation prior to collapse, Figure A-14 reveals smaller average ßRTR values,
and shows that the ßRTR value clearly reduces as the building becomes less
ductile (for .T < 3).
Figure A-14 Relationship between period-based ductility ( .T
= ( .u
/.y,eff)
and record-to-record variability (.RTR) from collapse data for
reinforced concrete frames. The FEMA P695 simplified
relationship for predicting .RTR is superimposed.
Figure A-14 also superimposes the FEMA P695 simplified relationship for
predicting ßRTR. Most structures are relatively ductile and have .T > 3. For
these structures ßRTR should be taken as 0.40 and the totally system collapse
uncertainty is given by Table 7-2. For structures with .T = 3, ßRTR can be
reduced according to Equation A-4:
. RTR . 0.1. 0.1.T . 0.4 (A-4)
where ßRTR must be greater than or equal to 0.2. If Equation A-4 is used, the
total system collapse uncertainty can be computed according to Equation 7-4,
repeated below:
TOT RTR DR TD MDL
. . . 2 . . 2 . . 2 . . 2 (A-5)
Values of .TOT obtained should be rounded to nearest 0.05 for use in Table 7-
3.
FEMA P695 A: Ground Motion Record Sets A-43
This reduction in .RTR is expected to apply only to a limited number of
structural systems that are either very brittle or unusual structures that have
limited period elongation prior to collapse, such as base-isolated structures.
A.14 Summary and Conclusion
This appendix describes the Far-Field and Near-Field record sets and the
scaling methods appropriate for collapse evaluation of building archetypes
using IDA.
Both the Far-Field and Near-Field record sets have average . values that are
lower than expected for MCE motions, and therefore can substantially
underestimate the CMR without appropriate adjustment for spectral shape
effects. The adjustment method is discussed in Section 7.4 and Appendix B.
The Far-Field record set is generally appropriate for collapse evaluation of
buildings, but can slightly overestimate the CMR of buildings at sites close to
fault rupture (e.g., distances less than 10 km).
The Near-Field record set is generally appropriate for collapse evaluation of
buildings at sites close to fault rupture (i.e., distances less than 10 km). Note
that the Near-Field record set is not specifically required as part of the basic
assessment Methodology. Even so, the Near-Field record set was developed
for comparative purposes and is documented here to both substantiate the
earlier comparisons and for use in other studies.
FEMA P695 B: Adjustment of Collapse Capacity Considering B-1
Effects of Spectral Shape
Appendix B
Adjustment of Collapse Capacity
Considering Effects of Spectral
Shape
This appendix describes the background and development of simplified
spectral shape factors that depend on the fundamental period of the building,
as well as the expected elongation of structural period as the structure
collapses. These factors are used to adjust the collapse capacity to account
for the frequency content (spectral shape) of the ground motion record set.
B.1 Introduction
A challenge associated with analytical prediction of structural collapse is the
selection of ground motions for use in dynamic analysis. A characteristic of
ground motions that can affect collapse capacity is the spectral shape. For
rare ground motions in California, such as Maximum Considered Earthquake
(MCE) ground motions, the spectral shape is much different than the shape
of a structural design spectrum contained in ASCE/SEI 7-05 (ASCE, 2006a)
or a uniform hazard spectrum (Baker, 2005; Baker and Cornell, 2006).
Figure B-1 shows the acceleration spectrum of a Loma Prieta ground motion1
(PEER, 2006a). This motion has a MCE intensity at a period of 1.0 second,
which is Sa(1.0 sec) = 0.9 g for this example. This spectrum is labeled as
“2% in 50 year Sa” which is the same as the MCE for this site. This figure
also shows the intensity predicted by the Boore et al. (1997) attenuation
prediction, consistent with the event and site associated with this ground
motion. These predicted spectra include the median spectrum and the
plus/minus one and two standard deviation spectra, assuming that Sa values
are lognormally distributed.
1 This motion is from the Saratoga station and is owned by the California
Department of Mines and Geology. For this illustration, this spectrum was scaled by
a factor of +1.4, in order to make the Sa(1 s) demand the same as the MCE demand.
For the purposes of this example, please consider this spectrum to be unscaled, since
later values (e.g., e) are computed using unscaled spectra.
B-2 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
In Figure B-1, this MCE motion has an unusual spectral shape with a “peak”
from 0.6 to 1.8 seconds that is much different from the shape of a uniform
hazard spectrum. This peak occurs around the period for which the motion is
said to have an MCE intensity, and at this period the observed Sa(1 s) is
much higher (0.9 g) than the mean expected Sa(1 s) from the attenuation
function (0.3 g). This peaked shape makes intuitive sense because it seems
unlikely that a ground motion with a much larger than expected spectral
acceleration (meaning much higher than the mean expected) at one period
would have similarly large spectral accelerations at all other periods.
0 0.5 1 1.5 2
0
0.5
1
1.5
2
Sa
component
[g]
Period [seconds]
Figure B-1 Comparison of an observed spectrum with spectra predicted by
Boore, Joyner, and Fumal (1997); after Haselton and Baker
(2006).
Epsilon, e, is defined as the number of logarithmic standard deviations
between the observed spectral value and the median prediction from an
attenuation function. At a period of 1.0 second, the spectral value is 1.9
logarithmic standard deviations above the predicted mean spectral value, so
this record is said to have “e = 1.9 at 1.0 second.” Similarly, this record has e
= 1.1 at 1.8 seconds. Thus, the e value is a function of the ground motion
record, the period of interest, and the attenuation function used for ground
motion prediction.
Trends shown in Figure B-1 are general to sites in coastal California where e
values ranging from 1.0 to 2.0 are typically expected for the MCE (or 2% in
50 year) ground motion level. These positive e values come from the fact
that the return period of the ground motion (i.e., 2,475 years for a 2% in 50
year motion) is much longer than the return period of the event that causes
e = +1.9 at 1.0s
e = +1.1 at 1.8s
Mean BJF
Mean + 2s
Mean - 2s
Observed Loma Prieta
Spectrum with 2% in 50
year Sa(1s)
FEMA P695 B: Adjustment of Collapse Capacity Considering B-3
Effects of Spectral Shape
the ground motion (i.e., 150-500 years for typical events in California).
Record selection for structural analyses at such sites should reflect the
expected e for the site and the ground motion hazard level of interest.
It should be noted that the expected e value is both hazard-level and site
dependent. For example, for 50% in 5 year ground motions in coastal
California, e values ranging from 0.5 to -2.0 are expected (Haselton et al.,
2007, chapter 4). In the eastern United States, e values ranging from 0.25 to
1.0 are expected for a 2% in 50 year motion. Negative e values for a 50% in
5 year motion stems from the fact that the return period of the ground motion
(i.e., 10 years) is much shorter than the typical return period of the event that
causes the ground motion (e.g., 150 to 500 years). The Eastern United States
has low positive e values because seismic events are less frequent than in
California, but the return periods are still shorter than the return period of a
2% in 50 year motion (i.e., 2,475 years).
Collapse capacity is defined as the Sa(T1) value that causes dynamic sidesway
collapse (termed SCT1). Research has shown that collapse capacity is higher
for motions with a peaked spectral shape relative to motions without a
peaked spectral shape. This is especially true when the peak of the spectrum
is near the fundamental period of the building (T1), and ground motions are
scaled based on Sa(T1) (Haselton and Baker, 2006; Baker, 2006; Baker, 2005;
Goulet et al., 2006; Zareian, 2006). Spectral accelerations at periods other
than T1 are often important to the collapse response of a building. For
example, the period elongation as the building responds inelastically makes
spectral values for period greater than T1 to become important to collapse
response. In addition, higher mode effects make periods less than T1 to also
become important to collapse response. Positive e peaked spectra typically
have lower spectral demands at periods away from T1.
Past studies have shown that if e(T1) = 0 ground motions are used when e(T1)
= 1.5 to 2.0 ground motions are appropriate, the median collapse capacity is
under-predicted by a factor of 1.3 to 1.8 for relatively ductile structures. In
cases where it is expected that the collapse-level ground motions will have
high positive e(T1) values, such as with modern buildings in high seismic
areas of California, properly accounting for these values is critical.
The most direct approach to account for spectral shape is to select ground
motions that have the appropriate e(T1) expected for the site and hazard level
of interest. This approach is difficult when assessing the collapse capacities
of many buildings with differing T1, because it would require a unique
ground motion set for each building. To address this issue, this Appendix
develops a simplified method which involves the use of a general set of
B-4 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
ground motion records (selected independent of e values), and a correction to
the median collapse capacity estimates to account for spectral shape. In this
process, the spectral shape is quantified by the e(T1) value expected for the
site and hazard level of interest.
B.2 Previous Research on Simplified Methods to Account
for Spectral Shape (Epsilon)
Several recent studies have focused on how spectral shape (e) affects collapse
capacity and pre-collapse structural responses (Baker, 2006; Baker, 2005;
Goulet et al., 2006; Haselton and Baker, 2006; Zareian, 2006). This
Appendix does not attempt to present a full literature review of this past
work.
The purpose of this Appendix is to develop a simplified method to account
for spectral shape, e. One recent study by Haselton and Deierlein (2007,
Chapter 3) developed such a method, and the following is an overview of
their work.
To develop the simplified method, Haselton and Deierlein first predicted the
collapse capacities (in terms of SCT1) of 65 modern reinforced concrete
special moment frame buildings. For the collapse assessment of each
building, 80 ground motions were utilized, with the goal of finding a
relationship between the collapse capacity, SCT1, and e(T1). For illustration
purposes, Figure B-2 shows an example of representative findings for a
single building2. This figure shows the results of linear regression analysis
that is used to define the relationship: LN[SCT1] = ß0 + ß1e. The value of ß0
indicates the average collapse capacity when e = 0, and the value of ß1
indicates how sensitive the collapse capacity, SCT1, is to changes in the e
value. To achieve the goal of a simplified correction method, the ß1 value is
a required ingredient. For this specific building, ß1 = 0.315.
It is observed from Figure B-2 that there is a great deal of scatter in the data.
Even so, the p-value for the regression is 1.3x10-6, which shows that the
trend between collapse capacity and e(T1) is statistically significant (note that
the p-value must only be less than 0.01-0.05 for the trend to be statistically
defensible, and this value is orders of magnitude smaller). Additionally,
similar statistically significant trends are observed for the 118 buildings
considered in this study (discussed later in section B.3.3), so this adds
2 This figure is representative of findings by Haselton and Deierlein, but this
specific figure is taken from collapse analyses of a 5-story wood light-frame
building, which was investigated as part of this current study (model No. 15; R =
6, wood-only), and is used in a later section of this Appendix.
FEMA P695 B: Adjustment of Collapse Capacity Considering B-5
Effects of Spectral Shape
confidence for utilizing these observed trends in an adjustment method to
account for proper spectral shape, e.
As a side note, this observed scatter in the data (Figure B-2) also suggests
that using a spectral shape correction method will increase the overall
uncertainty in the median collapse capacity estimate. This is true, but
accounting for the trend with e also decreases the record-to-record variability.
To approximately account for both of these effects in a simple manner, we
will (a) use a record-to-record variability of 0.4, which neglects the reduction
in record-to-record variability associated with e-selected records, and (b)
neglect the additional uncertainty associated with the variability in the Figure
B-2 regression line.
-3 -2 -1 0 1 2 3
0
1
2
3
4
5
SCT1(T1=0.64s) [g]
.(T1=0.64s)
Observation
Outlier
Regression
5/95% CIs on Mean
Best-Fit: LN(SCT1) = 0.542 + 0.315.
Figure B-2 Relationship between collapse spectral acceleration, SCT1, and e(T1)
for a single 5-story wood light-frame building (No. 15; R = 6 design,
wood-only model). This includes linear regression analysis results
which relate LN[SCT1(T1)] to e(T1), along with confidence intervals
(CIs) for the best-fit line. For this example, . 0 = 0.524 and . 1 =
0.315.
The study by Haselton and Deierlein (2007) found an average value of ß1 =
0.29 to be exceptionally consistent for modern reinforced concrete special
moment frame buildings with various heights (1-story to 20-stories) and
various designs, such as perimeter frame, space frame, and various bay
widths.
The relationship between ß1 and inelastic building deformation capacity was
also investigated. For buildings with larger inelastic deformation capacity,
the effective period elongates more prior to structural collapse. This causes
spectral values at period above T1 to have larger impact on collapse response,
and subsequently causes the spectral shape of the ground motion to become
B-6 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
more important (thus increasing ß1). To investigate the impact of inelastic
building deformation capacity, a set of 26 1967-era RC frame buildings was
investigated in the same manner as the previous set of reinforced concrete
special moment frame buildings. This revealed an average value of ß1 =
0.18, which is 35% lower than that of the reinforced concrete special moment
frame buildings; this confirms that the ground motion spectral shape, e, is
less important for buildings with lower deformation capacity. To further add
to the comparison, a set of 20 reinforced concrete ordinary moment frame
buildings was also considered, which showed an average value of ß1 = 0.19,
consistent with the finding for the 1967-era reinforced concrete frame
buildings.
Using a subset of the above data, Haselton and Deierlein created a simplified
equation that predicts ß1 based on the deformation capacity of the building
(judged from static pushover) and the building height.
This Appendix expands on the work by Haselton and Deierlein, and proceeds
to create an adapted version of their simplified methodology.
B.3 Development of a Simplified Method to Adjust
Collapse Capacity for Effects of Spectral Shape
(Epsilon)
The simplified method developed in this Appendix allows one to correct the
collapse capacity distribution without needing to compute the e(T1) values of
the ground motion records, and without needing to perform a regression
analysis. It involves using the Far-Field record set (Appendix A) for
structural collapse analyses, and then applying an adjustment factor to the
median collapse capacity ( CT Sˆ ).
The simplified correction factor depends on the following:
1. Differences between:
a. the e(T) values of the ground motions used in the structural
analyses (i.e., the Far-Field record set), and
b. the e(T) value expected for the site and ground motion hazard
level of interest.
2. How drastically the ground motion e values affect the building collapse
capacity; quantified by ß1, as described in section B.2. The ground
motion e values will have a greater effect on buildings with larger
inelastic deformation capacity, which are those that have more extensive
period elongation prior to collapse.
FEMA P695 B: Adjustment of Collapse Capacity Considering B-7
Effects of Spectral Shape
B.3.1 Epsilon Values for the Ground Motions in the Far-Field
Set
To adjust collapse capacity predictions for spectral shape, the epsilon, e(T),
values for the ground motion set used for collapse simulation are needed.
Figure B-3 shows the mean e(T) values for the Far-Field record set,
computed using the Abrahamson and Silva attenuation function (1997). The
mean e(T) values computed using the Boore et al. (1997) attenuation function
are similar but are not shown. For the Far-Field record set, mean e(T) values
are approximately 0.6 for periods less than 0.5 seconds, and are nearly 0.0
for periods greater than 1.5 seconds.
0 0.5 1 1.5 2 2.5 3 3.5 4
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Period (T) [sec]
Mean of .(T),records
Basic Far-Field Ground Motion Set (22x2 records)
Simplified Trend
Figure B-3 Mean e values for the Basic Far-Field ground motion set
. (T),records .
The simplified trend from Figure B-3 can be express using Equation B-1:
. .. . , ( ) 0.6 1.5 records . T . .T (B-1)
where , 0.0 ( ) 0.6 records .. T . .
Note that for consistency with the Methodology, the period used for
, ( )records . T is the code-defined fundamental period (T = CuTa), and not the
fundamental period computed from eigenvalue analysis (T1).
B.3.2 Target Epsilon Values
The expected epsilon, 0 . , value (used in this Appendix as the “target e”)
depends on both the site and hazard level of interest. A target e is needed in
the process of adjusting the median collapse capacity to account for spectral
shape.
To quantify the target e for various seismic design categories around the
United States, data from the United States Geological Survey (USGS) is
B-8 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
utilized. The USGS conducted the seismic hazard analysis for the United
States and used dissagregation to determine the mean expected e values ( 0 . )
for various periods and hazard levels of interest (Harmsen et al., 2003;
Harmsen 2001). Luco, Harmsen, and Frankel of the USGS provided
electronic 0 . data for use in this study.
Figure B-4 shows a map of expected 0 . (1s) for a 2% in 50 year motion in
Western United States. Values of 0 . (1s) = 0.50 to 1.25 are typical in areas
other than the seismic regions of California. The values are higher in most of
California, since the earthquake events have shorter return periods, with
typical values being 0 . (1s) = 1.25 to 1.75, and some values ranging upward
to 3.0.
Figure B-5 shows a map of expected 0 . (1s) for a 2% in 50 year motion in
Eastern United States. It shows typical values of 0 . (1s) = 0.75 to 1.0, with
some values reaching up to 1.25. Expected 0 . (1s) values fall below 0.75 for
the New Madrid Fault Zone, portions of the eastern coast, most of Florida,
southern Texas, and areas in the north-west portion of the map.
To illustrate the effects of period, Figure B-5b shows 0 . (0.2s) instead of
0 . (1s). This shows that typical 0 . (0.2s) are slightly lower and more
variable, having a typical range of 0.25 to 1.0. This is in contrast to the
typical range of 0.75 to 1.0 for 0 . (1s).
Figure B-4 Predicted 0 . values from dissagregation of ground motion
hazard for Western United States. The values are for a 1.0
second period and the 2% exceedance in 50 year motion
(Harmsen et al., 2003).
FEMA P695 B: Adjustment of Collapse Capacity Considering B-9
Effects of Spectral Shape
Figure B-5 Mean predicted 0 . values from dissagregation of ground
motion hazard for Eastern United States. The values are for (a)
1.0 second and (b) 0.2 second periods and the 2% exceedance
in 50 year motion (Harmsen et al., 2003).
To better quantify the 0 . values presented in the previous figures, Table B-1
and Table B-2 show the average 0 . and average spectral acceleration values
for Seismic Design Categories B, C, and D. These data are given for four
levels of ground motion: the motion with 10% exceedance in 50 years, 2% in
50 years, 1% in 50 years, and 0.5% in 50 years. These tables also show the
number of zip code data points included in each SDC.
Table B-1 Tabulated 0 . Values for Various Seismic Design Categories
Seismic Design
Category
Average e Values Number
of Zip
Code
Data
Points
e0(0.2s) e0(1.0s)
e10/50 e2/50 e1/50 e0.5/50 e10/50 e2/50 e1/50 e0.5/50
SDC B 0.14 0.42 0.49 0.55 0.31 0.80 0.94 1.04 20,142
SDC C 0.11 0.51 0.63 0.75 0.23 0.74 0.88 1.00 7,456
SDC D 0.25 0.88 1.09 1.27 0.33 0.99 1.21 1.39 6,461
SDC D,
0.35 < S1 < 0.599g
0.32 0.97 1.21 1.41 0.39 1.01 1.24 1.45 1,305
(a) (b)
B-10 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
Table B-2 Tabulated Spectral Demands for Various Seismic Design
Categories
Seismic Design
Category
Average Sa Values Number
of Zip
Code
Data
Points
Sa(0.2s) [g] Sa(1.0s) [g]
Sa10/50 Sa2/50 Sa1/50 Sa0.5/50 Sa10/50 Sa2/50 Sa1/50 Sa0.5/50
SDC B 0.06 0.18 0.26 0.39 0.02 0.06 0.08 0.11 20,142
SDC C 0.11 0.31 0.46 0.66 0.04 0.10 0.14 0.19 7,456
SDC D 0.50 1.05 1.35 1.68 0.18 0.38 0.49 0.62 6,461
SDC D,
0.35 < S1 < 0.599g
0.61 1.31 1.69 2.09 0.21 0.46 0.61 0.77 1,305
Seismic Design Category D is treated differently from the other categories,
since it is the category with highest spectral demand and often controls the
collapse performance assessment, as was shown in the Chapter 9 examples.
Since the higher spectral demands often control the performance, sites in
SDC D with higher values of S1 (i.e., 0.35 < S1 < 0.599) are used to define
the target e.
It should be noted that the 0 . values in Table B-1 are a bit lower than some
may expect for SDC D sites in seismic zones of California. This comes from
the values in Table B-1 being averages for all SDC D sites in the United
States. The SDC D sites are located in seismic regions of California, as well
as in the Eastern United States. Seismic sources in these two regions have
widely differing return periods, which causes 0 . values to vary, with the
values in California being larger. Average 2% in 50 year and 0.5% in 50
year 0 . (1.0s) values are listed below for selected California cities. These
values are averages over all SDC D zip codes in a given city, and are
comparable to 0.99 and 1.39 values from Table B-1 which are for the entire
United States. Since these values are averages over the city, the values at
each specific site in the city may be higher or lower than these values.
. 1.5 and 1.9 in San Francisco (average over 16 zip codes)
. 1.7 and 2.1 in Oakland (average over 10 zip codes)
. 1.6 and 2.0 in Berkeley (average over 3 zip codes)
. 1.6 and 2.1 in San Jose (average over 29 zip codes)
. 1.3 and 1.7 in Los Angeles (average over 58 zip codes)
. 2.0 and 2.2 in Riverside (average over 8 zip codes)
FEMA P695 B: Adjustment of Collapse Capacity Considering B-11
Effects of Spectral Shape
Table B-1 presented both 0 . (0.2 s) and 0 . (1.0 s) values. The 0 . (1.0 s)
values are used to develop the target e values, since most building structures
have periods closer to 1.0 second, or greater than 1.0 second. Buildings with
periods near 0.2 seconds are relatively rare.
To complete the determination of target e values for each Seismic Design
Category, the proper ground motion hazard level must be established. Since
the spectral shape, e, adjustment will be used to modify the median collapse
capacity the appropriate hazard level should be near this median. Table 7-3
shows that for typical uncertainty levels, the median collapse capacity must
be roughly twice the MCE, for a structure to pass a 10% conditional collapse
probability acceptance criterion. Therefore, the ground motion hazard level
that should be used in establishing the target e should have spectral
acceleration demand that is twice (or more) of the 2% in 50 year demand
(since the 2% in 50 year motion and MCE are considered to be identical for
Seismic Design Categories B, C, and D). Table B-1 shows that the 0.5% in
50 year demand approximately meets this criterion, and is still conservative
for SDC D, so the 0.5% in 50 year 0 . (1.0s) is used as the target e. Based on
this, the target e value for Seismic Design Categories B and C is 1.0, and the
target e value for SDC D is 1.5.
SDC E must be treated differently, because the 2% in 50 year motion and the
MCE can differ widely in near-fault regions. At sites close to a fault, the
MCE motion has a shorter return period than the 2% in 50 year motion (i.e.,
2,475 years). This is the result of the methodology used to construct the
1997 NEHRP maps, as explained in Appendix B of FEMA 369 (FEMA,
2001). In the near-field, the MCE is set to be 50% larger than the median
predicted motion; this is approximately one logarithmic standard deviation
greater than the median, so by definition e = 1.0. To approximately account
for the fact that the median collapse capacity is larger than the MCE motion,
a target e = 1.2 is used for SDC E.
B.3.3 Impact of Spectral Shape (.) on Median Collapse
Capacity
For buildings with larger inelastic deformation capacity, the effective period
elongates more significantly before structural collapse, causing the spectral
values at periods greater than T1 to have more drastic impacts on collapse
response. This subsequently causes the spectral shape, e, of the ground
motions to become more important. The ß1 value (defined in section B.2) is
used to quantify how drastically the spectral shape, e, affects the collapse
capacity, so the ß1 is larger for buildings with larger deformation capacity.
The purpose of this section is to create a predictive equation to estimate the
B-12 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
proper ß1 value for any building. This section presents data for 118
buildings, and then shows how the data were used to create the predictive
equation for ß1.
Quantification of Building Period Elongation Prior to Collapse
Fundamentally, the spectral shape, e, is important to the building collapse
capacity because the building period elongates prior to collapse; this makes
the building response become affected also by the spectral values at periods
greater than the initial period T1. In creating a predictive equation for ß1, it
would be ideal to base the prediction directly on the expected amount of
period elongation, such as a ratio of the near-collapse period to the initial
undamaged period. However, the period near collapse is an ill-defined
parameter, so this method instead uses a surrogate for period elongation.
To begin creating this surrogate for period elongation, the building
deformation capacity is first quantified using a pushover analysis, in
accordance with the guidelines of Section 6.3. Figure B-6 shows an
idealized pushover curve and shows that, according to Section 6.3, the
ultimate roof displacement (.u) is defined as the roof displacement associated
with 20% loss of base shear strength. For illustration, Figure B-6 also shows
the effective yield roof displacement (.y,eff), but this term is actually
computed from the building period, as will be described later in this section.
Vmax
.y,eff .u
0.8Vmax
V
Base
Shear
Roof Displacement
Figure B-6 Idealized nonlinear static pushover curve (from Section 6.3).
The ultimate building deformation capacity is divided by the yield
deformation to compute the period-based ductility parameter, µT = .u
/ .y,eff,
FEMA P695 B: Adjustment of Collapse Capacity Considering B-13
Effects of Spectral Shape
that is used as the surrogate for period elongation. The square root of µT can
be approximately thought of as the ratio of the near-collapse period to the
initial undamaged period. However, this is not precisely the case because the
deformed shape of the building changes as the building is damaged. Even so,
µT is an acceptable surrogate to approximately quantify the period elongation
prior to collapse.
To compute the effective yield roof displacement (.y,eff), the initial
undamaged fundamental period of the building, T1, is first computed using
eigenvalue analysis for the undamaged structural model. This period is then
converted into a roof displacement using the guidelines provided in Section
3.3.3.3.2 of ASCE/SEI 41-06 (ASCE, 2006b). To provide more stability to
the Methodology, the code defined period of T = CuTa must instead be used
for cases when T > T1. This requirement is imposed to ensure that analysts
do not over-predict the value of ß1 by using a structural model that has an
initial stiffness that is unreasonably large. The independent peer review
panel should be careful to scrutinize the initial stiffness assumptions of the
structural model.
Equation B-2 shows the ASCE/SEI 41-06 equation for effective yield roof
displacement (.y,eff), with the proper modifications for computing yield
displacement (e.g., coefficients based on elastic response, and the use of
yield base shear for Sa), and with the previously described period
requirement. This equation is presented in Section 6.3 as Equation 6-7.
max 2
, 0 2 1 (max( , ))
y eff 4
C V g T T
W
.
.
. . . .. ..
(B-2)
where C0 relates the SDOF displacement to the roof displacement, computed
according to ASCE/SEI 41-06 Section 3.3.3.3; Vmax/W is the maximum base
shear normalized by building weight, g is the gravity constant; T is the codedefined
fundamental period (i.e., CuTa); and T1 is the undamaged
fundamental period of the structural model computed using eigenvalue
analysis.
In summary, when creating the predictive equation for ß1, the period-based
ductility parameter, µT, will be used as a surrogate for period elongation prior
to collapse.
Predictive Equation for ß1: Database of Buildings Used as the Basis
for Creating the Equation
In order to create a predictive equation for ß1, the first step is to understand
what the ß1 values should be for various structural models that have various
deformation capacities. To this end, the project team assembled a large
B-14 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
database of structural models and then, for each model, completed the
collapse analyses and performed the regression to compute the ß1 value (as
previously outlined in Section B.2). This section summarizes the set of
structural models utilized, provides the important properties for each
building, and report the ß1 values computed using regression for each model.
The results from these structural models are utilized to develop the predictive
equation for ß1, as discussed in the next section.
These five building sets include a total of 118 buildings, which are described
in the following list. Partial documentation of the results for these models is
provided in tables that follow this list.
. 30 code-conforming reinforced concrete special moment frame buildings
from Haselton and Deierlein (2007, Chapter 6). These buildings range
from 1-20 stories, and are representative of currently designed buildings
(ASCE/SEI 7-02 and ACI 318-02) in high seismic regions of California.
Eighteen of these buildings are included in an example in section 9.2 of
this report, and the full results for these 18 buildings are included in
Table B-3 below. More complete documentation is provided in Haselton
and Deierlein (2007). The average ß1 for this set of buildings is 0.29 and
the average µT is 11.5.
. 30 reinforced concrete special moment frame buildings that were
designed and analyzed as part of a design sensitivity study completed by
Haselton and Deierlein (2007, Chapter 7). This set of building designs
includes variations in design strength (R value), design strong-column
weak beam ratio, and design drift limits. These buildings are useful for
this study because they include subsets of buildings that are identical,
except for a single design change (such as the strong-column weak beam
ratio, which affects the building deformation capacity); this allows trends
to be seen more clearly. These data are not fully documented in this
Appendix, though some of these data are used later in Figure B-7; the
reader is referred to Chapter 7 of Haselton and Deierlein (2007) for full
documentation.
- 16 code-conforming reinforced concrete ordinary moment frame
buildings, from Section 9.3 of this report. These buildings range
from 2-12 stories, and are representative of currently designed
buildings in the eastern United States. The full results for these 20
buildings are included in Table B-4 below. The average ß1 for this
set of buildings is 0.19, and the average µT is 3.4. As expected, this
set of buildings shows lower ß1 values being associated with
buildings having lower deformation capacity.
FEMA P695 B: Adjustment of Collapse Capacity Considering B-15
Effects of Spectral Shape
. 26 non-ductile reinforced concrete frame buildings from Liel (2008).
These buildings range from 2-12 stories, and are representative of
existing 1967-era buildings in high seismic regions of California. The
full results for these 26 buildings are included in Table B-5 below. The
average ß1 for this set of buildings is 0.18, and the average µT is 2.9,
consistent with the observation made from the set of reinforced concrete
ordinary moment frame buildings, that lower ß1 values are to be expected
for buildings with lower deformation capacity.
. 16 wood light-frame buildings, from Section 9.4 of this report. These
buildings range from 1-5 stories, and are representative of currently
designed buildings in high-seismic regions of the United States (SDC D).
The full results for these 16 buildings are included in Table B-6 below.
The average ß1 for this set of buildings is 0.33 and the average µT is 7.8.
It is also observed that there is a modest difference between the ß1 values
for the buildings designed for SDC Dmax versus SDC Dmin.
The following tables (Table B-3 through Table B-6) provide documentation
of the results obtained using the above models. These tables include the ß1
values for each model, obtained from regression analysis, as well as the
model properties that are important for prediction of ß1.
B-16 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
Table B-3 Documentation of Building Information and .1
Regression
Results for the Set of Reinforced Concrete Special Moment
Frame Buildings
Arch.
ID
Design
Configuration Building Information ß1 from
Regr.
No. of
Stories
Cs [g] O
T1
[s]
T [s] .y,eff / hr .u
/ hr µT ß1
Maximum Seismic (Dmax) and Low Gravity (Perimeter Frame) Designs, 20' Bay Width
2069 1 0.125 1.6 0.71 0.26 0.0055 0.077 14.0 0.27
2064 2 0.125 1.8 0.66 0.45 0.0034 0.067 19.6 0.26
1003 4 0.092 1.6 1.12 0.81 0.0035 0.038 10.9 0.27
1011 8 0.050 1.6 1.71 1.49 0.0023 0.023 9.8 0.31
1013 12 0.044 1.7 2.01 2.13 0.0023 0.026 11.4 0.29
1020 20 0.044 1.6 2.63 3.36 0.0032 0.018 5.6 0.26
Maximum Seismic (Dmax) and High Gravity (Space Frame) Designs, 20' Bay Width
2061 1 0.125 4.0 0.42 0.26 0.0048 0.077 16.1 0.39
1001 2 0.125 3.5 0.63 0.45 0.0061 0.085 14.0 0.26
1008 4 0.092 2.7 0.94 0.81 0.0041 0.047 11.3 0.26
1012 8 0.050 2.3 1.80 1.49 0.0037 0.028 7.5 0.32
1014 12 0.044 2.1 2.14 2.13 0.0028 0.022 7.7 0.25
1021 20 0.044 2.0 2.36 3.36 0.0040 0.023 5.7 0.30
Comparison of Results - SDC Dmax & Dmin Seismic Design Conditions, 20' Bay Width
4011 8 0.017 1.8 3.00 1.60 0.0028 0.010 3.6 0.25
4013 12 0.017 1.8 3.35 2.28 0.0023 0.010 4.3 0.20
4020 20 0.017 1.8 4.08 3.60 0.0021 0.008 3.9 0.15
4021 20 0.017 2.8 4.03 3.60 0.0031 0.012 3.8 0.20
Comparison of Results - 20-Foot and 30-Foot Bay Width Designs (SDC Dmax)
1009 4 0.092 1.6 1.16 0.81 0.0037 0.050 13.4 0.32
1010 4 0.092 3.3 0.86 0.81 0.0042 0.056 13.2 0.27
FEMA P695 B: Adjustment of Collapse Capacity Considering B-17
Effects of Spectral Shape
Table B-4 Documentation of Building Information and . 1 Regression
Results for the Set of Reinforced Concrete Ordinary Moment
Frame Buildings
Arch.
ID
Design
Configuration Building Information ß1 from
Regr.
No. of
Stories
Cs [g] O
T1
[s]
T [s] .y,eff / hr .u
/ hr µT ß1
Minimum Seismic, SDC Bmin, Low Gravity (Perimeter Frame) Designs
9101 2 0.041 2.0 1.56 0.45 0.0059 0.022 3.7 0.25
9103 4 0.023 1.8 2.81 0.81 0.0048 0.014 3.0 0.17
9105 8 0.012 2.6 4.58 1.49 0.0026 0.008 3.0 0.07
9107 12 0.010 2.3 5.8 2.13 0.0027 0.007 2.5 0.10
Minimum Seismic, SDC Bmin, High Gravity (Space Frame) Designs
9102 2 0.041 6.6 0.85 0.45 0.0067 0.020 3.0 0.15
9104 4 0.023 5.3 1.49 0.81 0.0050 0.010 2.1 0.27
9106 8 0.012 6.0 2.53 1.49 0.0045 0.014 3.0 0.22
9108 12 0.010 6.0 2.85 2.13 0.0034 0.031 9.1 0.21
Maximum Seismic, SDC Bmax, Low Gravity (Perimeter Frame) Designs
9201 2 0.087 1.6 1.23 0.45 0.0069 0.024 3.5 0.28
9203 4 0.048 1.6 1.93 0.81 0.0047 0.018 3.8 0.24
9205 8 0.026 1.5 3.39 1.49 0.0033 0.009 2.8 0.12
9207 12 0.018 1.7 4.43 2.13 0.0028 0.009 3.0 0.17
Maximum Seismic, SDC Bmax, High Gravity (Space Frame) Designs
9202 2 0.087 2.9 0.81 0.45 0.0058 0.019 3.3 0.09
9204 4 0.048 3.0 1.36 0.81 0.0050 0.011 2.2 0.27
9206 8 0.026 3.1 2.35 1.49 0.0045 0.014 3.0 0.19
9208 12 0.018 3.8 2.85 2.13 -- -- -- 0.16
B-18 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
Table B-5 Documentation of Building Information and . 1 Regression
Results for the Set of 1967-era Reinforced Concrete Buildings
Arch.
ID
Design Configuration Building Information
ß1
from
Regr.
No.
Stor.
Perim. /
Space
Cs [g] O
T1
[s]
T [s] .y,eff / hr .u
/ hr µT ß1
3001 2 Space 0.086 1.9 1.08 0.45 0.0067 0.019 2.9 0.16
3002 2 Perim. 0.086 1.6 1.04 0.45 0.0052 0.035 6.7 0.22
3003 4 Perim. 0.068 1.2 1.96 0.81 0.0059 0.013 2.2 0.18
3004 4
Space
0.068 1.3 1.98 0.81 0.0065 0.016 2.4 0.20
3009 4 0.068 1.5 1.98 0.81 0.0075 0.016 2.1 0.15
3010 4 0.068 1.4 1.98 0.81 0.0070 0.015 2.1 0.15
3012 4 0.068 1.4 1.98 0.81 0.0070 0.016 2.3 0.19
3032 4 0.068 1.6 1.98 0.81 0.0081 0.018 2.2 0.19
3015 8
Perim.
0.054 1.1 2.36 1.49 0.0033 0.007 2.1 0.16
3034 8 0.054 1.3 1.84 1.49 0.0024 0.009 3.8 0.22
3016 8
Space
0.054 1.6 2.20 1.49 0.0042 0.011 2.6 0.18
3017 8 0.054 1.6 2.17 1.49 0.0041 0.011 2.7 0.19
3018 8 0.054 1.6 2.20 1.49 0.0042 0.012 2.9 0.19
3019 8 0.054 1.6 2.20 1.49 0.0042 0.011 2.6 0.16
3020 8 0.054 1.6 2.20 1.49 0.0042 0.011 2.6 0.16
3021 8 0.054 1.6 2.20 1.49 0.0042 0.011 2.6 0.19
3022 12
Perim.
0.047 1.1 2.75 2.13 0.0026 0.005 1.9 0.10
3035 12 0.047 1.3 2.23 2.13 0.0020 0.006 2.9 0.19
3023 12
Space
0.047 1.9 2.26 2.13 0.0031 0.010 3.3 0.16
3024 12 0.047 2.0 2.19 2.13 0.0030 0.010 3.3 0.19
3026 12 0.047 2.0 2.26 2.13 0.0032 0.010 3.1 0.18
3027 12 0.047 1.9 2.26 2.13 0.0031 0.010 3.3 0.19
3028 12 0.047 2.0 2.26 2.13 0.0032 0.007 2.2 0.16
3029 12 0.047 1.9 2.26 2.13 0.0031 0.012 3.9 0.18
3031 12 0.047 1.9 2.26 2.13 0.0031 0.012 3.9 0.15
3033 12 0.047 2.2 2.26 2.13 0.0035 0.012 3.4 0.22
FEMA P695 B: Adjustment of Collapse Capacity Considering B-19
Effects of Spectral Shape
Table B-6 Documentation of Building Information and . 1 Regression
Results for the Set of Wood Light-frame Buildings
Arch.
ID
Design
Configuration Building Information ß1 from
Regression
No. of
Stories
Cs [g] O
T1
[s]
T [s] .y,eff / hr .u / hr µT ß1
High Seismic (SDC Dmax) - Low Aspect Ratios - R = 6
1 1 0.167 1.6 0.51 0.25 0.0055 0.045 8.3 0.29
5 2 0.167 2.0 0.52 0.26 0.0045 0.045 10.1 0.31
9 3 0.167 1.6 0.61 0.36 0.0032 0.033 10.3 0.29
High Seismic (SDC Dmax) - High Aspect Ratios - R = 6
2 1 0.167 3.1 0.38 0.25 0.0060 0.045 7.4 0.27
6 2 0.167 2.8 0.46 0.26 0.0048 0.040 8.3 0.35
10 3 0.167 2.8 0.47 0.36 0.0034 0.041 12.1 0.32
13 4 0.167 2.7 0.52 0.45 0.0031 0.023 7.2 0.30
15 5 0.167 2.4 0.64 0.53 0.0034 0.022 6.4 0.32
Low Seismic (SDC Dmin) - Low Aspect Ratios - R = 6
11 3 0.063 1.6 1.10 0.41 0.0039 0.025 6.4 0.41
Low Seismic (SDC Dmin) - High Aspect Ratios - R = 6
3 1 0.063 2.7 0.65 0.25 0.0059 0.047 7.9 0.42
4 1 0.063 4.1 0.53 0.25 0.0059 0.045 7.7 0.29
7 2 0.063 3.1 0.74 0.30 0.0052 0.038 7.3 0.35
8 2 0.063 2.5 0.80 0.30 0.0050 0.033 6.6 0.34
12 3 0.063 3.1 0.83 0.41 0.0043 0.030 7.0 0.36
14 4 0.063 2.6 0.99 0.51 0.0041 0.025 6.1 0.31
16 5 0.063 2.4 1.12 0.60 0.0041 0.021 5.2 0.34
Predictive Equation for ß1: Development and Evaluation of Equation
Using the database of results for the 118 buildings, the goal is to create an
equation that can be used to predict the value of ß1, without needing to
perform a regression analysis. In an earlier section, it was already decided
that ß1 will be predicted using the period-based ductility parameter, µT =
.u/.y,eff, as a surrogate for the period elongation that the building undergoes
prior to collapse. Therefore, the goal of this section is to determine the
proper form of this equation.
To most clearly see the relationship between ß1 and µT, Figure B-7 shows the
trends using three sets of reinforced concrete special moment frame
B-20 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
buildings, each with constant height. These three sets of buildings are
described as follows:
. Four-story perimeter frames, with the design R value varied from 4 to 12.
. Four-story space frames, with the design strong-column weak-beam ratio
varied from 0.4 to 3.0.
. Twelve-story space frames, with the design R value varied from 4 to 12.
These data show a clear trend between period-based ductility (µT) and ß1 for
observed µT values up to 8-9, and suggest that the effects are saturated for
larger values of µT. For these buildings, the ß1 value saturates at 0.32 for two
of the sets, and 0.26 for the third set. These observations suggest that the
predictive equation for ß1 should saturate at a µT value somewhere between 8
and 11, and saturate at a value of ß1 that is between 0.26 and 0.32.
Figure B-7 Relationship between . 1 and period-based ductility, . T, for
three sets of reinforced concrete special moment frame
buildings of constant height within each set.
Considering both the reinforced concrete special moment frame data (Figure
B-7 and Table B-3) and the wood light-frame data (Table B-6), it was
decided that the predictive equation for ß1 should saturate at a value of ß1 =
0.32 for µT . 8. This will accurately reflect the average values from the wood
data, accurately reflect the saturation point for the reinforced concrete special
moment frame data, and slightly over-predict the average ß1 values for the
reinforced concrete special moment frame data.
Now that the saturation point has been established for the ß1 predictive
equation, the equation can be created by choosing an appropriate functional
0 5 10 15
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
.T
.1
4-story perimeter frame - design R varied
4-story space frame - SCWB varied
12-story space frame - design R varied
FEMA P695 B: Adjustment of Collapse Capacity Considering B-21
Effects of Spectral Shape
form, and then using the full set of data from the 118 buildings, with standard
linear regression analysis to predict log(ß1) (Chatterjee et al., 2000). Several
function forms were evaluated, and a power form was chosen, in order to
best match the trends in the data, and enforce that ß1 = 0.0 when µT = 1.0.
Small corrections were then applied to the equation to enforce the desired
saturation point of ß1 = 0.32 at µT = 8, while still accurately fitting the data
for lower values of µT.
The final relationship between ß1 and µT is shown in Equation B-3:
. .. .0.42
1
.ˆ . 0.14 .T .1 (B-3)
where the limit is enforced of µT . 8.0.
Predictive Equation for ß1: Comparisons of Predictions and
Observations
In order to evaluate the prediction accuracy of Equation B-3, Figure B-8
compares the predicted and observed values of ß1. Figure B-8a shows the
data points for all of the 118 buildings used to create the equation, and Figure
B-8b shows the average values for each of the building subsets. Overall, this
shows that the equation accurately predicts the expected values of ß1. More
specifically, this shows that the equation accurately predicts the ß1 values for
non-ductile buildings, accurately predicts the values for wood buildings in
SDC Dmax, slightly under-predicts the values for wood buildings in SDC
Dmin, but still accurately predicts the values for wood buildings on average
(with overall average µT = 7.4 and ß1 = 0.32), and slightly over-predicts the
values for ductile reinforced concrete special moment frame buildings.
B.4 Final Simplified Factors to Adjust Median Collapse
Capacity for the Effects of Spectral Shape
Following the rationale above, the spectral shape factor, SSF, can be
computed using Equation B-4.
. . 1 , exp ( ) ( ) o records SSF . ... . T .. T .. (B-4)
where ß1 depends on building inelastic deformation capacity (Equation B-3);
0 . depends on SDC and is equal to 1.0 for SDC B/C, 1.5 for SDC D, and 1.2
for SDC E (Section B.3.2); and , ( )records . T is for the Far-Field record set
(Section B.3.1).
B-22 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
Figure B-8 Comparison of predicted and observed values of . 1, for all 118
buildings used to create the predictive equation. Figure (a)
shows all the data, and Figure (b) shows the average values for
the individual subsets of building types.
Table B-7 through Table B-9 present values of spectral shape factor, SSF,
for various levels of building period-based ductility, µT, and various building
periods, using Equation B-4. Table B-7 presents values for Seismic Design
Categories B and C, Table B-8 presents values for SDC D, and Table B-9
presents values for SDC E.
To compute the adjusted collapse margin ratio, multiply the SSF value by the
collapse margin ratio that was predicted using the Far-Field record set, as
shown in Equation B-5:
ACMR . SSF *CMR (B-5)
0 5 10 15 20
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
.
T
.
1
Prediction
RC SMF - Chp. 9
RC SMF - Haselton
RC OMF - Chp. 9
RC 1967-era - Liel
Wood - Chp. 9 - SDC Dmax
Wood - Chp. 9 - SDC Dmin
0 5 10 15 20
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0.5
.
T
.
1
Prediction
RC SMF - Chp. 9
RC SMF - Haselton
RC OMF - Chp. 9
RC 1967-era - Liel
Wood - Chp. 9 - SDC Dmax
Wood - Chp. 9 - SDC Dmin
(b)
(a)
FEMA P695 B: Adjustment of Collapse Capacity Considering B-23
Effects of Spectral Shape
Table B-7 Spectral Shape Factors for Seismic Design Categories B, C, and
Dmin
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.02 1.04 1.06 1.08 1.09 1.12 1.14
0.6 1.00 1.02 1.05 1.07 1.09 1.11 1.13 1.16
0.7 1.00 1.03 1.06 1.08 1.10 1.12 1.15 1.18
0.8 1.00 1.03 1.06 1.08 1.11 1.14 1.17 1.20
0.9 1.00 1.03 1.07 1.09 1.13 1.15 1.19 1.22
1.0 1.00 1.04 1.08 1.10 1.14 1.17 1.21 1.25
1.1 1.00 1.04 1.08 1.11 1.15 1.18 1.23 1.27
1.2 1.00 1.04 1.09 1.12 1.17 1.20 1.25 1.30
1.3 1.00 1.05 1.10 1.13 1.18 1.22 1.27 1.32
1.4 1.00 1.05 1.10 1.14 1.19 1.23 1.30 1.35
. 1.5 1.00 1.05 1.11 1.15 1.21 1.25 1.32 1.37
Table B-8 Spectral Shape Factors for Seismic Design Category Dmax
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.05 1.10 1.13 1.18 1.22 1.28 1.33
0.6 1.00 1.05 1.11 1.14 1.20 1.24 1.30 1.36
0.7 1.00 1.06 1.11 1.15 1.21 1.25 1.32 1.38
0.8 1.00 1.06 1.12 1.16 1.22 1.27 1.35 1.41
0.9 1.00 1.06 1.13 1.17 1.24 1.29 1.37 1.44
1.0 1.00 1.07 1.13 1.18 1.25 1.31 1.39 1.46
1.1 1.00 1.07 1.14 1.19 1.27 1.32 1.41 1.49
1.2 1.00 1.07 1.15 1.20 1.28 1.34 1.44 1.52
1.3 1.00 1.08 1.16 1.21 1.29 1.36 1.46 1.55
1.4 1.00 1.08 1.16 1.22 1.31 1.38 1.49 1.58
. 1.5 1.00 1.08 1.17 1.23 1.32 1.40 1.51 1.61
B-24 B: Adjustment of Collapse Capacity Considering FEMA P695
Effects of Spectral Shape
Table B-9 Spectral Shape Factors for Seismic Design Category E
T
(sec.)
Period-Based Ductility,..T
1.0 1.1 1.5 2 3 4 6 . 8
= 0.5 1.00 1.03 1.06 1.09 1.12 1.14 1.18 1.21
0.6 1.00 1.04 1.07 1.10 1.13 1.16 1.20 1.23
0.7 1.00 1.04 1.08 1.11 1.14 1.17 1.22 1.26
0.8 1.00 1.04 1.09 1.12 1.16 1.19 1.24 1.28
0.9 1.00 1.05 1.09 1.12 1.17 1.21 1.26 1.31
1.0 1.00 1.05 1.10 1.13 1.18 1.22 1.28 1.33
1.1 1.00 1.05 1.11 1.14 1.20 1.24 1.30 1.36
1.2 1.00 1.06 1.11 1.15 1.21 1.25 1.32 1.38
1.3 1.00 1.06 1.12 1.16 1.22 1.27 1.35 1.41
1.4 1.00 1.06 1.13 1.17 1.24 1.29 1.37 1.44
. 1.5 1.00 1.07 1.13 1.18 1.25 1.31 1.39 1.46
B.5 Application to Site Specific Performance Assessment
Simplified spectral shape adjustment factors can be modified for site-specific
and building-specific collapse performance assessment. To compute the SSF
for a site-specific collapse performance assessment, use Equation B-4 with
the following values:
. ß1 should be computed using Equation B-3.
. 0 . (T) should not be based on Section B.3.2. Instead, 0 . (T) should be
based directly on the dissagregation of the probabilistic seismic hazard
analysis for the site of interest and for the ground motion hazard level of
the median collapse capacity.
. . (T),records will differ depending on the ground motion set utilized in the
performance assessment. If the site is within 10 km of an active fault
capable of producing an event larger than magnitude M6.5, then the
Near-Field ground motion record set should be used (Appendix A);
otherwise, the Far-Field record set should be used. If the Far-Field
record set is used, then , ( )records . T should be taken from Figure B-3 and
Equation B-1. If the Near-Field record set is used, then , ( )records . T
should be taken to equal , , ( )records NF . T from Figure B-9 and Equation B-
6.
FEMA P695 B: Adjustment of Collapse Capacity Considering B-25
Effects of Spectral Shape
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Period (T) [sec]
Mean of .(T),records,NF
Near-Field Ground Motion Set (28x2 records)
Simplified Trend
Figure B-9 Mean e values for the Near-Field ground motion set,
. (T),records,NF .
The Near-Field set is nearly e-neutral at all periods, but a slight simplified
trend from Figure B-9 can be approximated using Equation B-6:
. .. . ,records,NF . (T) . 0.2 T .1.5 (B-6)
where ,records,NF 0.0 .. (T) . 0.2 .
FEMA P695 C: Development of Index Archetype Configurations C-1
Appendix C
Development of Index Archetype
Configurations
This appendix illustrates how index archetype configurations are developed
for reinforced concrete moment frame systems and wood light-frame shear
wall systems.
Development of structural system archetypes considers both structural
configuration issues and seismic behavioral effects described in Chapter 4.
Consideration of structural configuration issues is discussed in this appendix.
Examination of seismic behavioral effects is described in Appendix D. Index
archetype configurations are subsequently used to develop index archetype
designs, which are then used to develop index archetype models.
Development and calibration of nonlinear index archetype models is
described in Appendix E.
C.1 Development of Index Archetype Configurations
for a Reinforced Concrete Moment Frame System
In this section, index archetype configurations are developed for a reinforced
concrete special moment frame system conforming to design requirements
contained in ASCE/SEI 7-05 Minimum Design Loads for Buildings and
Other Structures (ASCE, 2006a), and ACI 318-05 Building Code
Requirements for Structural Concrete (ACI, 2005). The index archetype
configurations described here are used in the reinforced concrete moment
frame system example of Section 9.2. When the Methodology is applied to a
proposed new seismic-force-resisting system, index archetype configurations
will be based on existing code requirements, as appropriate, and new design
requirements developed specifically for the proposed system.
C.1.1 Establishing the Archetype Design Space
To establish the archetype design space, design parameters that significantly
affect seismic performance are first identified, and then the bounds on each
design parameter are established. The overall range of permissible
configurations, structural design parameters, and other features that define
the application limits for a seismic-force-resisting system are specified in the
system design requirements and in existing code requirements, as applicable.
C-2 C: Development of Index Archetype Configurations FEMA P695
Parameters that were identified as having a critical impact on the collapse
performance of reinforced concrete moment frame systems, along with
related physical properties and associated design variables, are listed in Table
C-1. This organizational approach is useful for deciding which of the many
design variables should be the focus of further investigation in index
archetype configurations.
Table C-1 Important Parameters, Related Physical Properties, and
Design Variables for Reinforced Concrete Moment Frame
Systems
Important Parameter Related Physical
Properties
Design Variables
Column and beam
plastic rotation capacity
Axial load ratio
Building height, bay width, ratio of
tributary areas for gravity and
lateral loads
Column aspect ratio
Building height, bay width, story
heights, allowable reinforcement
ratio
Confinement ratio Confinement ratio used in design
Stirrup spacing Stirrup spacing used in design
Longitudinal bar
diameter
Longitudinal bar diameter used in
design
Reinforcement ratios Reinforcement ratio allowed in
design
Concrete strength Concrete strength used in design
Element Strengths
All element strengths Conservatism of engineer, dead
and live loads used in design
Beam strengths Slab width (steel) assumed
effective
Column strengths
Ratio of factored to expected axial
loads, level of conservatism in
applying strong-column weakbeam
provision
Number of stories in
collapse mechanism
Strength/stiffness
irregularities
Presence of strength or stiffness
irregularity, ratio of first to upper
story heights, how column heights
are stepped down over height
Lateral stiffness of frame Member sizes in
frame
Member/joint/footing stiffness
used in design
Gravity system
strength/stiffness Gravity system Not considered in this assessment
Key design variables are those that are likely to have a significant impact on
the collapse performance of the proposed system. Key design variables
identified for reinforced concrete moment frame systems, along with their
applicable ranges, are listed in Table C-2.
FEMA P695 C: Development of Index Archetype Configurations C-3
Table C-2 Key Design Variables and Ranges Considered in the Design
Space for Reinforced Concrete Moment Frame Systems
Key Design Variable Range Considered in Archetype Design Space
Structural System
Special Reinforced Concrete
Moment Frame (as per ASCE/SEI 7-
05, ACI 318-05)
All designs meet code requirements
Seismic framing system Perimeter and space frames
Configuration
Building Height 1 to 20 stories
Bay Width 20 to 30 feet
First and upper story heights 15 and 13 feet
Element Design
Confinement ratio and stirrup
spacing
Conforming to ACI 318-05
Concrete compressive strength 5 to 7 ksi
Longitudinal bar diameter #8 and #9 are commonly used
Strength/stiffness irregularities As permitted by existing code
Loading
Ratio of tributary areas for gravity
and lateral loads
0.1 (perimeter frame) to 1.0 (space frame)
Design floor loads 175 psf
Lower and upper bounds on design
floor load 150 to 200 psf
Design floor live load Constant: 50 psf
The design variables and ranges identified in Table C-2 provide the basis for
identifying a finite number of design variations for use in developing index
archetype configurations.
Figure C-1 illustrates one example of how index archetype configurations
might change for a design variable associated with the ratio of tributary areas
for gravity and lateral loads (Agrav/Alat). In the case of moment frame
systems, this variable is primarily affected by whether the building is
designed as a space frame or a perimeter frame system. Table C-2 identifies
a range of 0.1 to 1.0 for this design variable, which is schematically shown in
the figure.
C-4 C: Development of Index Archetype Configurations FEMA P695
Figure C-1 Different index archetype configurations for varying ratios of
tributary areas for gravity Agrav and lateral loads Alat.
C.1.2 Identifying Index Archetype Configurations and
Populating Performance Groups
Identification of a set of index archetype configurations requires
consideration of performance group binning in accordance with Section
4.3.1. As a minimum, performance groups should consider the design
ground motion intensities for the governing Seismic Design Category
(maximum and minimum SDC), two fundamental period domains (longperiod
and short-period), variations in gravity load intensity (high and low
gravity loads), and building configuration. In the case of reinforced concrete
moment frames, "high gravity" systems are space frames and "low gravity"
systems are perimeter frames, as shown in Figure C-1. This results in the
need for index archetype configurations to populate at least eight
performance groups, as described in Table 4-3:
. High gravity loads, maximum SDC, short-period archetypes
. High gravity loads, maximum SDC, long-period archetypes
. High gravity loads, minimum SDC, short-period archetypes
. High gravity loads, minimum SDC, long-period archetypes
. Low gravity loads, maximum SDC, short-period archetypes
. Low gravity loads, maximum SDC, long-period archetypes
. Low gravity loads, minimum SDC, short-period archetypes
. Low gravity loads, minimum SDC, long-period archetypes
Additional performance groups need to be defined where other design
variables are important. For example, performance groups for reinforced
concrete moment frame also include variation in bay width (column spacing)
of 20 feet and 30 feet, respectively, which may be an important parameter for
Perimeter Frame
(Agrav/Alat = 0.16)
Space Frame
(Agrav/Alat = 1.0)
FEMA P695 C: Development of Index Archetype Configurations C-5
system performance (Table 9-1). Each performance group must be robustly
populated by buildings that have representative configurations and heights
across the entire archetype design space. Each performance group is
expected to contain at least three buildings, except in cases where that
performance group represents an unusual rather than typical configuration of
the structural system of interest.
Index archetype configurations must be conceived in such a way that index
archetype models will capture important system behavior. For moment
frames in general, three framing bays are considered to be the minimum
number feasible for capturing variations in behavior related to interior and
exterior columns and beam-column joints, strong-column weak-beam design
provisions, and induced column axial loads due to overturning effects. As a
result, the three-bay, variable story-height configuration shown in Figure C-2
was selected as the simplest model still capable of capturing important
collapse performance behaviors for reinforced concrete moment frame
systems.
Figure C-2 Index archetype model for reinforced concrete moment frame
systems.
To develop a set of index archetype configurations, the range of parameters
that will be commonly utilized in design and construction must be
understood. As identified in Table C-2, building heights between one and
twenty stories are expected, since most buildings taller than twenty stories
would include a core wall in addition to moment frames. Considering typical
office occupancies, story heights are expected to be relatively consistent,
taken as 15 feet for the first story and 13 feet for the upper stories. Plan
dimensions of 120 feet by 120 feet and 120 feet by 180 feet, along with bay
widths ranging between 20 feet and 30 feet, are also expected to be typical
for such buildings.
C-6 C: Development of Index Archetype Configurations FEMA P695
Figure C-3 illustrates two set of index archetype configurations representing
the range of expected building heights for 20-foot and 30-foot
configurations, respectively, that are used to populate reinforced-concrete
moment frame performance groups. In this case, performance groups with
short-period archetypes include 1-story and 2-story (and 3-story) buildings
and performance groups with long-period archetypes include 4-story and
taller buildings.
13' (typ.)
15'
12-story
20-story
1-story
2-story
4-story
8-story
3 bays @ 20' 3 bays @ 20' 3 bays @ 20' 3 bays @ 20' 3 bays @ 20' 3 bays @ 20'
(a) Twenty-foot bay width
13' (typ.)
15'
3 bays @ 30' 3 bays @ 30' 3 bays @ 30' 3 bays @ 30' 3 bays @ 30' 3 bays @ 30'
12-story
20-story
1-story
2-story
4-story
8-story
(b) Thirty-foot bay width
Figure C-3 Index archetype configurations for a reinforced concrete
moment frame system.
Maximum and minimum seismic loads are defined by the range of possible
design loads for Seismic Design Category D (SDC Dmax and SDC Dmin).
High and low gravity load intensities are represented by the space frame or
perimeter frame configurations.
Based on the information summarized in Table C-1 and Table C-2, and
consideration of performance group binning and nonlinear analysis
modeling, a matrix of index archetype configurations, such as the one
described in Table C-3, can be developed. This matrix of index archetype
configurations completely populates the performance groups expected to be
most typical: 20-foot bay spacing of different heights, space and perimeter
FEMA P695 C: Development of Index Archetype Configurations C-7
frames, designed for the maximum seismic load intensity (SDC Dmax). Taller
buildings (4, 8, 12, and 20 stories) respond in the long-period range and are
classified in separate performance groups from shorter short-period buildings
(1, 2, and 3 stories). In order to verify that these performance groups control
the assessment, a small number of additional archetype configurations are
considered in other performance groups, including archetypes with a 30-foot
bay width and archetypes designed for minimum seismic load intensity (SDC
Dmin).
Table C-3 Matrix of Index Archetype Configurations for a Reinforced
Concrete Moment Frame System
Perform.
Group No.
Archetype ID
(Ch. 9)
No. of
Stories
Period
Domain
Bay
Width
Gravity
Loads
SDC
PG-1
2061 1
Short 20'
High
1001 2 (Space) Dmax
-- 3
PG-2
1008 4
Long 20' High
(Space)
Dmax
1012 8
5014 12
5021 20
PG-5
2069 1
Short 20'
Low
2064 2 (Perimeter) Dmax
-- 3
PG-6
1003 4
Long 20' Low
(Perimeter)
Dmax
1011 8
5013 12
5020 20
PG-4 6021 20 Long 20' High
(Space)
Dmin
PG-8
6011 8
Long 20'
Low
6013 12 (Perimeter) Dmin
6020 20
PG-10 1009 4 Long 30' High
(Space)
Dmax
PG-14 1010 4 Long 30' Low
(Perimeter)
Dmax
C.1.3 Preparing Index Archetype Designs and Index
Archetype Models
Each of the index archetype configurations in Table C-3 are used to prepare
index archetype designs, which are represented with index archetype models.
In the examples of Chapter 9, index archetype designs for reinforced
C-8 C: Development of Index Archetype Configurations FEMA P695
concrete moment frames were prepared in accordance with the requirements
of ASCE/SEI 7-05 and ACI 318-05. Complete adherence to design
requirements is essential for adequately evaluating the performance of a class
of buildings designed using the proposed system. Designs should address
minimum design requirements, utilize nominal material properties, and not
be overly conservative. Assumptions used in the development of index
archetype designs for reinforced concrete moment frames are listed in Table
C-4.
While index archetype designs are prepared using nominal material
properties and other standard assumptions, index archetype models should be
defined based on the expected behavior of the building. For modeling and
assessment of the mean performance of reinforced concrete moment frame
systems, this included the use of expected material properties, element
stiffness assumptions based on test data, tributary slab contributions for beam
strength and stiffness, and expected gravity loads. More details of modeling
for reinforced concrete moment frame systems are described in Appendix D
and Appendix E.
Table C-4 Index Archetype Design Assumptions for a Reinforced Concrete
Moment Frame System
Design Parameter Design Assumption
Assumed Stiffness
Member stiffness assumed in
design: Beams 0.5EIg (ASCE/SEI 41)
Member stiffness assumed in
design: Columns
0.7EIg for all axial load levels (based on practitioner
recommendation)
Slab consideration Slab not included in stiffness/strength design of
beams
Footing rotational stiffness assumed
in design: 2-4 story Effective stiffness of grade beam
Footing rotational stiffness assumed
in design: 8-20 story
Basement assumed; exterior columns fixed at
basement wall, interior columns consider stiffness of
first floor beam and basement column
Joint stiffness assumed in design Elastic joint stiffness
Expected Design Conservatisms
Conservatism applied in element
flexural and shear (capacity)
strength design
1.15 times required strength
Conservatism applied in joint
strength design 1.0 times required strength
Conservatism applied in strongcolumn
weak-beam design Use expected ratio of 1.3 instead of 1.2
FEMA P695 C: Development of Index Archetype Configurations C-9
C.2 Development of Index Archetype Configurations
for a Wood Light-Frame Shear Wall System
In this section, index archetype configurations are developed for a wood
light-frame shear wall system conforming to design requirements contained
in ASCE/SEI 7-05. The system consists of wood light-frame bearing wall
structures braced with wood structural panel shear walls. These archetypes
are used in the wood light-frame bearing wall example of Section 9.4.
C.2.1 Establishing the Archetype Design Space
The following design variables were considered in the development of index
archetype configurations for this system: (1) number of stories; (2)
Maximum and minimum seismic intensity for the governing Seismic Design
Category (SDC D in this case); (3) building occupancy and use; and (4) shear
wall aspect ratio.
Wood light-frame buildings of upto three stories are common across most of
the United States. Wood light-frame multi-family residential buildings of
four to five stories represent a growing trend along the West Coast. The
number of stories considered in the archetype design space ranges from one
to five stories.
Minimum and maximum values of design spectral response acceleration, SDS,
used in this example are 0.375g and 1.00g, respectively. The minimum value
of 0.375g is lower than 0.50g specified for SDC D in Table 5-1A, but is used
in this illustrative example to make use of available designs. An SDS of 1.00g
represents the Maximum seismic intensity for regular, short-period SDC D
structures. Design for wind loads is not considered.
The range of building occupancies considered includes residential and
commercial occupancies, the latter including educational and institutional
uses. The primary difference between residential and commercial buildings
is the spacing between shear wall lines, which affects the tributary seismic
mass. Residential buildings generally have more walls and closer spacing. A
typical spacing of 25 feet between shear wall lines and a tributary width of
12.5 feet for seismic mass was used for residential occupancies. Commercial
buildings are more likely to have open configurations with widely spaced
shear walls at the perimeter. A typical spacing of 80 feet between shear wall
lines and a tributary width of 40 feet for seismic mass was used for
commercial occupancies.
Residential occupancies were further split into one- and two-family detached
dwellings and multi-family dwellings, respectively. One- and two-family
detached dwellings were assumed to be one- and two-stories tall, with typical
C-10 C: Development of Index Archetype Configurations FEMA P695
wood-frame floor weights (without topping slabs). Multi-family dwellings
were assumed to be three- to five-stories tall, with floor weights including
gypcrete topping slabs.
Shear wall aspect ratios included high aspect ratio shear walls (height/width
ratios of 2.7 to 3.3) and low aspect ratio shear walls (height/width ratios of
1.5 or less). In accordance with ASCE/SEI 7-05, a capacity reduction was
considered for high aspect ratio shear walls, resulting in the need for
configurations with more dense nailing patterns.
C.2.2 Identifying Index Archetype Configurations and
Populating Performance Groups
A set of 16 index archetype configurations are developed to investigate the
effect of wall aspect ratio, height (one- to five-stories), and lateral loading
intensity, as shown in Table C-5. These configurations are chosen
recognizing the predominant use of high aspect ratio shear walls in
residential buildings and low aspect ratio shear walls in commercial
buildings. These structures are grouped into performance groups according
to the design lateral load intensity (SDC Dmax or Dmin), wall aspect ratio (high
or low) and period domain dominating the response (short- or long-period)
(see Table 9-23). Other configurations were eliminated because they are not
representative of wood light-frame construction. In particular, nominal
gravity loads are considered in all the designs because light wood-frame
archetype design and performance is not influenced by gravity loads.
Likewise, there are few archetype configurations falling in the long-period
domain because these are not representative.
C.2.3 Preparing Index Archetype Designs and Index
Archetype Models
Each of the index archetype configurations in Table C-5 are used to prepare
index archetype designs, which are represented with index archetype models.
In this example, index archetype designs were prepared in accordance with
the requirements of ASCE/SEI 7-05, considering only the contribution of
wood structural panel sheathing, and ignoring the possible beneficial effects
of gypsum wallboard that is likely to be present.
Designs were developed using a response modification coefficient, R, equal
to 6. Index archetype designs, including number of piers, sheathing, and
fasteners for R=6 are provided in Table C-6.
FEMA P695 C: Development of Index Archetype Configurations C-11
Table C-5 Index Archetype Configurations for Wood Light-Frame Shear
Wall Systems
Model
No.
No. of
Stories
Seismic
Design
Coef.
(SDS)
Tributary
Width for
Seismic
Weight (ft)
Floor/Roof
Tributary
Seismic
Weight
(kips)
Shear
Wall
Aspect
Ratio
Occupancy
1 1 1.0 40 41/0 Low Commercial
2 1 1.0 12.5 13.65/0 High 1&2 Family
3 1 0.375 40 41/0 High Commercial
4 1 0.375 12.5 13.65/0 High 1&2 Family
5 2 1.0 40 82/41 Low Commercial
6 2 1.0 12.5 17.3/13.65 High 1&2 Family
7 2 0.375 40 82/41 High Commercial
8 2 0.375 12.5 17.3/13.65 High 1&2 Family
9 3 1.0 40 82/41 Low Commercial
10 3 1.0 12.5 27.3/13.65 High Multi-Family
11 3 0.375 40 82/41 Low Commercial
12 3 0.375 12.5 27.3/13.65 High Multi-Family
13 4 1.0 12.5 27.3/13.65 High Multi-Family
14 4 0.375 12.5 27.3/13.65 High Multi-Family
15 5 1.0 12.5 27.3/13.65 High Multi-Family
16 5 0.375 12.5 27.3/13.65 High Multi-Family
Table C-6 Index archetype designs for wood light-frame shear wall
systems (R = 6)
Model
No.
No. of
Stories
No. of
Piers
Pier Length
(ft)
Sheathing Shear Wall
Nailing
1 1 2 9.0 7/16” OSB 8d at 6"
2 1 4 3.0 7/16“OSB 8d at 6"
3 1 4 3.0 7/16” OSB 8d at 6"
4 1 2 3.0 7/16” OSB 8d at 6"
5
2 3 8.0 7/16” OSB 8d at 4"
1 3 8.0 7/16” OSB 8d at 2"
C-12 C: Development of Index Archetype Configurations FEMA P695
Table C-6 Index archetype designs for wood light-frame shear wall
systems (R = 6) continued
Model
No.
No. of
Stories
No. of
Piers
Pier Length
(ft)
Sheathing Shear Wall
Nailing
6
2 5 3.0 7/16” OSB 8d at 6"
1 5 3.0 7/16” OSB 8d at 3"
7
2 5 3.0 7/16” OSB 8d at 4"
1 5 3.0 7/16” OSB 8d at 2"
8
2 2 3.0 7/16” OSB 8d at 6"
1 2 3.0 7/16” OSB 8d at 4"
9
3 3 10.0 7/16” OSB 8d at 6"
2 3 10.0 7/16” OSB 8d at 2"
1 3 10.0 19/32” PLWD 10d at 2"
10
3 6 3.0 7/16” OSB 8d at 6"
2 6 3.0 7/16” OSB 8d at 2"
1 6 3.0 19/32” PLWD 10d at 2"
11
3 2 7.0 7/16” OSB 8d at 6"
2 2 7.0 7/16” OSB 8d at 3"
1 2 7.0 7/16” OSB 8d at 2"
12
3 3 3.0 7/16” OSB 8d at 6"
2 3 3.0 7/16” OSB 8d at 3"
1 3 3.0 7/16” OSB 8d at 2"
13
4 6 3.3 7/16” OSB 8d at 6"
3 6 3.3 7/16” OSB 8d at 2"
2 6 3.3 19/32” PLWD 10 at 2"
1 6 3.3 19/32” PLWD 10 at 2"
14
4 4 3.0 7/16” OSB 8d at 6"
3 4 3.0 7/16” OSB 8d at 4"
2 4 3.0 7/16” OSB 8d at 3"
1 4 3.0 7/16” OSB 8d at 2"
15
5 6 3.7 7/16” OSB 8d at 6"
4 6 3.7 7/16” OSB 8d at 3"
3 6 3.7 7/16” OSB 8d at 2"
2 6 3.7 19/32” PLWD 10d at 2"
1 6 3.7 19/32” PLWD 10d at 2"
16
5 4 3.3 7/16” OSB 8d at 6"
4 4 3.3 7/16” OSB 8d at 4"
3 4 3.3 7/16” OSB 8d at 3"
2 4 3.3 7/16” OSB 8d at 2"
1 4 3.3 7/16” OSB 8d at 2"
FEMA P695 C: Development of Index Archetype Configurations C-13
C.2.4 Other Considerations for Wood Light-Frame Shear Wall
Systems
The following considerations were not addressed in this example, but are
provided for the purpose of further illustrating the development of index
archetype configurations.
A mix of shear wall aspect ratios could very likely occur within a given
wood light-frame system. Careful thought should be given as to whether
performance groups consisting of all high, all low, or mixed aspect ratio
walls will be most representative or most critical. This could change with the
system being considered.
It is becoming common practice to mix alternative bracing elements with
conventional shear wall bracing in this type of construction. These elements
could be designed to carry a small or large portion of the seismic story shear
forces. Consideration should be given as to whether the seismic story shear
carried by alternative bracing elements needs to be added as another variable
in the index archetype configurations.
For wood light-frame systems in particular, wall finish materials are known
to have a very significant impact on the overall collapse behavior of the
system. The effects of wall finish materials should be considered in the
context of the system being proposed. It may be appropriate to include a
range of wall finish materials as another variable in the index archetype
configurations.
FEMA P695 D: Consideration of Behavioral Effects D-1
Appendix D
Consideration of Behavioral
Effects
This appendix illustrates how seismic behavioral effects are considered in the
development of index archetype configurations for reinforced concrete
moment frame systems conforming to design requirements contained in
ASCE/SEI 7-05, Minimum Design Loads for Buildings and Other Structures
(ASCE, 2006a), and ACI 318-05, Building Code Requirements for Structural
Concrete (ACI, 2005).
Development of structural system archetypes considers both structural
configuration issues and seismic behavioral effects described in Chapter 4.
Consideration of structural configuration issues is discussed in Appendix C.
Index archetype configurations are subsequently used to develop index
archetype designs, which are then used to develop index archetype models.
Development and calibration of nonlinear index archetype models for
reinforced concrete moment frames is described in Appendix E.
D.1 Identification of Structural Failure Modes
Consideration of seismic behavioral effects includes identifying critical limit
states, dominant deterioration modes and collapse mechanisms that are
possible, and assessing the likelihood that they will occur. How a component
or system behaves under seismic loading is often influenced by configuration
decisions, so behavioral effects and configuration issues should be
considered concurrently in the development of index archetype
configurations. This process is illustrated in Figure D-1.
Once all possible failure modes have been identified for a given system, the
list is narrowed to a subset of likely collapse mechanisms by ruling out
failure modes that are unlikely to occur based on system design and detailing
requirements, experimental data, engineering judgment, analytical models, or
observations from past earthquakes. Remaining failure modes will be
assessed through explicit simulation of failure modes through nonlinear
analyses, or through evaluation of non-simulated failure modes using
alternative limit state checks on demand quantities from nonlinear analyses.
Index archetype configurations must address failure modes that will be
explicitly simulated in nonlinear analysis models in Chapter 5.
D-2 D: Consideration of Behavioral Effects FEMA P695
Figure D-1 Consideration of behavioral effects in developing index
archetype configurations.
D.2 System Definition
Possible element deterioration and failure modes are influenced by the
material properties, mechanical properties, and design requirements of the
proposed seismic-force-resisting system.
Reinforced concrete moment frame systems in this example are assumed to
meet design requirements for reinforced concrete moment frames specified in
ASCE/SEI 7-05 and ACI 318-05. For special moment frames, these
standards provide rigorous and specific design and detailing provisions
intended to prevent certain failure modes from occurring, and to promote
ductile failure modes in selected members. Special reinforced concrete
moment frame requirements include, for example, an upper limit on
longitudinal steel reinforcement ratio, seismic hoop detailing, confinement
reinforcing in columns and beam-column joints, exclusion of lap splices from
hinge regions, strong-column-weak-beam provisions, and other capacity
design requirements. Ordinary moment frames are not subject to these
provisions.
Begin with system design
requirements and preliminary
configurations
Identify possible element
deterioration modes
What types of models
are needed to capture
these deterioration
modes?
Identify local and global collapse
scenarios (often combinations of
deterioration modes)
Assess the likelihood of each
collapse scenario
Reflect deterioration modes and
collapse scenarios in index
archetype configurations
FEMA P695 D: Consideration of Behavioral Effects D-3
For a proposed seismic-force-resisting system, design and detailing
requirements must be well-defined and specific before element deterioration
modes and system collapse scenarios can be identified.
D.3 Element Deterioration Modes
In this example, failure modes have been identified for reinforced concrete
moment frame components based on a review of experimental tests, available
published information, and observations from past earthquakes. Potential
deterioration modes for reinforced concrete moment frame components are
listed in Table D-1 and shown in Figure D-2. They are classified into six
groups (A to F) depending on the type of structural element and the physical
behavior associated with deterioration, as shown in Figure D-3. For each
mode, currently available nonlinear element models are rated for their ability
to simulate the deterioration behavior.
Table D-1 Possible Deterioration Modes for Reinforced Concrete
Moment Frame Components.
Deterioration
Mode
Element Behavior
Model Availability1
Description
Simulation Fragility
A Beamcolumn
Flexural 4 NR
Concrete spalling
Reinforcing bar yielding
Concrete core
Reinforcing bar buckling (incl.
Stirrup fracture)
Reinforcing bar fracture
B Beamcolumn
Axial
compression
2 4
Concrete crushing, longitudinal
bar yielding
Stirrup-rupture, longitudinal bar
buckling
C Beamcolumn
Shear
Shear +Axial
1 4
Concrete Shear
Transverse tie pull-out of rupture
Concrete
Loss of aggregate interlock
Possible loss of axial load-carrying
capacity in columns
D Joint Shear 3 2 Panel shear failure
E Reinforcing
bar
Pull-out or
Bond-slip 2 2
Reinforcing bar bond-slip or
anchorage failure at joint
Reinforcing bar lap-splice failure
Reinforcing bar pull-out (in beams
or at footing)
F Slab
connection Shear 2 3
Punching shear or large shear
eccentricity at slab column
connection, in shear wall
structures possible loss of
connection between slab and
wall
Possible vertical collapse of slab
1Model availability ratings: 0 - nonexistent; 1 – low confidence; 5 – high confidence; NR –
not required because behavior can be simulated.
D-4 D: Consideration of Behavioral Effects FEMA P695
Figure D-2 Reinforced concrete frame plan and elevation views, showing
location of possible deterioration modes.
Figure D-3 Illustration and classification of possible deterioration modes for
reinforced concrete moment frame components.1
1 Photo sources: (A) nisee.berkeley.edu/thumbnail/6257_3021_0662/IMG0071.jpg
(B) www.structures.ucsd.edu/Taiwaneq/buildi25.jpg (C) www.disaster.archi.
tohoku.ac.jp/eng/topicse/030726htm/5-28.jpg (D) www.structures. ucsd.edu/
Taiwaneq/ buildi21.jpg (E) www.structures.ucsd.edu/Taiwaneq/buildil9.jpg
(F) www.nbmg.unr.edu/nesc/bobcox/ndx2.php
FEMA P695 D: Consideration of Behavioral Effects D-5
In this example, foundation failure modes have not been included because
they are judged not critical for this system. Foundation failure modes should
be considered if they are judged to have a potentially significant effect on the
collapse performance of a proposed system.
D.3.1 Flexural Hinging of Beam and Columns
Deterioration mode ‘A’ consists of flexural hinging in beams and columns
and associated concrete spalling, concrete core crushing, stirrup fracture, and
reinforcing bar yielding, buckling, and fracture. Reinforcing bar yielding,
concrete core crushing, and the associated strength and stiffness degradation
can be simulated fairly accurately. In contrast, modeling of buckling and
fracture of longitudinal reinforcement, or stirrup fracture is less accurate.
These behaviors are important contributors to the deterioration of strength
and stiffness in reinforced concrete frame elements at large deformations
near collapse.
Fiber-type models capture the spread of plasticity along the length of the
element and the constituent concrete models can be calibrated to adequately
model the behavior associated with concrete deterioration from cracking to
crushing (Haselton et al., 2007). However, currently available steel material
models are not able to replicate the behavior of rebar as it buckles and
fractures. Due to this limitation, current fiber models were judged
inadequate for simulating collapse. While lumped plasticity models do not
have the precision of fiber models, they can be calibrated to capture the
deterioration associated with rebar buckling and stirrup fracture leading to
loss of confinement (Haselton, 2006; Ibarra, 2003).
D.3.2 Compressive Failure of Columns
Deterioration mode ‘B’ corresponds to compressive failure of columns and is
characterized by concrete crushing, buckling of longitudinal reinforcement,
and yielding or fracture of transverse reinforcement. This deterioration mode
occurs when compressive forces in the column exceed the compressive
capacity. In earthquakes, this may occur due to overturning, which increases
axial loads in some of the columns of a moment-resisting frame.
D.3.3 Shear Failure of Beam and Columns
Deterioration mode ‘C’ corresponds to shear failure of beam-columns,
characterized by shear cracking in concrete and yielding and/or pull-out of
transverse stirrups. This mode of deterioration is particularly dangerous for
columns with significant axial load. Shear deterioration and increased
D-6 D: Consideration of Behavioral Effects FEMA P695
displacement demands can lead to subsequent vertical collapse of the column
(Elwood and Moehle, 2005), and possibly progressive collapse of a structure.
Modeling the cyclic response of a reinforced concrete element experiencing
shear deterioration is complex due to interactions between shear, moment,
and axial force, as well as the overall brittle nature of the deterioration mode.
Elwood (2004) and others have simulated deterioration in the lateral strength
and stiffness of a column by adding a shear spring. To date, models for
vertical collapse of columns following shear failure have been challenged by
a lack of experimental data for columns experiencing vertical collapse.
Due to the brittle nature of this failure mode, special and intermediate
moment frames in seismically active areas are required to utilize capacity
design principles that ensure flexural hinging prior to shear failure (ACI,
2005). Capacity design requirements, however, do not apply to ordinary
concrete moment frames. Shear failure, and the possible subsequent loss of
gravity-load carrying capacity, is treated as a non-simulated collapse mode in
evaluating ordinary moment frame systems, due to limitations in the ability
to directly simulate these failure modes.
D.3.4 Joint Panel Shear Behavior
Deterioration Mode ‘D’ is associated with deterioration in shear strength and
stiffness of the joint panel region. In reinforced concrete frame models, shear
panels are modeled with an inelastic rotational spring inserted at the joint
(Lowes et al., 2004; Altoontash, 2004). This joint modeling capability is
available in the Open System for Earthquake Engineering Simulation
(OpenSees, 2006) software framework and in most structural analysis
software.
Several researchers have used a detailed approach that employs the modified
compression field theory (Vecchio and Collins,1986; Stevens et al., 1991) to
develop a monotonic backbone that relates the panel shear force to the shear
deformation angle (Altoontash, 2004; Lowes et al., 2004). Modified
compression field theory has been shown to work well for conforming joints,
but is not as reliable for non-conforming joints with less confinement.
Modeling of joint shear behavior is especially important for joints that are
not protected by capacity design requirements (e.g., ordinary concrete
moment frame systems).
D.3.5 Bond-Slip of Reinforcing Bars
Deterioration Mode ‘E’ corresponds to slip of the reinforcing bar relative to
the surrounding concrete, including anchorage pull-out, bond slip, and splice
FEMA P695 D: Consideration of Behavioral Effects D-7
failure. The extent of slip, and the possible occurrence of bar pull-out,
depends on the embedded length, bar diameter, concrete strength, and the
number and magnitude of load cycles.
Slip without pull-out occurs when the embedment or splice length is
sufficient to prevent the cut end of the rebar from moving relative to the
surrounding concrete, even under cyclic loading. The effects of slip without
pull-out include: (1) decreased pre-yield stiffness; (2) decreased post-yield
stiffness; and (3) increased element plastic deformation capacity. These
effects are included in the calibration of the plastic hinges for reinforced
concrete moment frame models.
Slip with pull-out occurs when the rebar slips relative to the concrete over the
full length of embedment. This occurs when the embedment or lap splice
lengths are relatively short, as is typical of ordinary concrete moment frames.
Slip with pull-out can lead to severe reduction in strength and stiffness, as
well as a significant reduction in plastic rotation capacity. Models for pullout
are moderately well-developed and range from continuum finite element
models (which attempt to model the interface between the rebar and
concrete) to simple rotational spring models.
D.3.6 Punching Shear in Slab-Column Connections
Deterioration Mode ‘F’ corresponds to punching shear in slab-column
connections or in wall-slab connections. Following punching shear failure, a
slab may experience local collapse depending on the gravity load intensity
and continuity of reinforcement in the area of the slab-column connection
(Aslani 2005).
Punching shear deterioration can be modeled with a standard nonlinear
lumped plasticity spring. The resulting vertical collapse is difficult to
directly simulate, but this failure mode can be treated as a non-simulated
collapse mode (Aslani and Miranda 2005).
D.4 Local and Global Collapse Scenarios
Collapse occurs when seismic loading causes element deterioration modes to
combine in a way that forms a sidesway collapse mechanism or a vertical
(local) collapse mechanism. For reinforced concrete frame systems, possible
collapse scenarios and contributing element deterioration modes are
identified and organized as shown in Tables D-2a and D2-b for sidesway and
vertical collapse modes, respectively. These scenarios were established
using engineering judgment based on examination of collapses in previous
earthquakes, experimental test data, and analytical studies.
D-8 D: Consideration of Behavioral Effects FEMA P695
Table D-2a Collapse scenarios for Reinforced Concrete Moment Frames
– Sidesway Collapse
Scenario
Element Deterioration Mode
Description
A B C D E F
FS1 x Beam and column flexural hinging,
forming sidesway mechanism
FS2 x Column hinging, forming soft-story
mechanism
FS3 x x Beam or column flexural-shear failure,
forming sidesway mechanism
FS4 x x Joint-shear failure, possibly with beam
and/or column hinging
FS5 x x
Reinforcing bar pull-out or splice failure
in columns or beams, leading to
sidesway mechanism
Table D-2b Collapse Scenarios For Reinforced Concrete Moment Frames
– Vertical Collapse
Scenario
Element Deterioration Mode
Description
A B C D E F
FV1 x Column shear failure, leading to column
axial collapse
FV2 x x Column flexure-shear failure, leading to
column axial collapse
FV3 x
Punching shear failure, leading to slab
collapse
FV4 Failure of floor diaphragm, leading to
column instability
FV5 x
Crushing of column, leading to column
axial collapse; possibly from overturning
effects
D.5 Likelihood of Collapse Scenarios
Table D-3 includes the range of possible collapse scenarios for reinforced
concrete frame structures. From this list, the scenarios which are most likely
to occur for each type of moment frame defined in ACI 318-05 (ordinary,
intermediate, and special) are identified.
Changes in detailing requirements for different reinforced concrete frame
systems govern which collapse modes are likely to occur for that system.
Design requirements for special concrete moment frames (ACI, 2005) are
designed to promote ductile collapse modes and to prevent the formation of
brittle collapse modes. These requirements serve to limit the likelihood of
several possible collapse modes.
FEMA P695 D: Consideration of Behavioral Effects D-9
Ordinary moment frames are vulnerable to a wider range of possible collapse
modes, due to less stringent design requirements. Several researchers (e.g.
Aycardi et al., 1994; Kunnath et al., 1995; Filiatrault et al., 1998) have noted
the tendency for soft story or column-hinging mechanisms to form in
ordinary concrete moment frames. In addition, ordinary concrete moment
frames might experience lap-splice failure, pull-out of reinforcing bars at
beam-column joints, and column shear failures. The behavior of
intermediate moment frames is anticipated to be somewhere between special
and ordinary moment frames.
The likelihood of each collapse scenario for special, intermediate, and
ordinary moment frames is shown in Table D-3. Depending on the proposed
system and associated design requirements, there may be one or several
collapse modes that are likely to occur.
Table D-3 Likelihood of Column Collapse Scenarios by Frame Type (H:
High, M: Medium, L: Low)
System
Sidesway Collapse Vertical Collapse
FS1 FS2 FS3 FS4 FS5 FV1 FV2 FV3 FV4 FV5
SMF H M L L L L L L L L-M
IMF H M-H L M M L L L M L-M
OMF H H H H H M H M M M-H
D.6 Collapse Simulation
For reinforced concrete moment frame systems, possible collapse modes are
considered in the context of the three-bay variable-height archetype frame
model described in Appendix C and Chapter 9. Index archetype
configurations must account for all collapse scenarios that have been
identified as likely to occur, and will be explicitly simulated in the nonlinear
index archetype models. Critical limit states or failure modes that cannot be
explicitly simulated in index archetype models are evaluated using the
procedure for non-simulated collapse modes.
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-1
Moment Frame Systems
Appendix E
Nonlinear Modeling of
Reinforced Concrete Moment
Frame Systems
E.1 Purpose
This appendix describes the development of nonlinear index archetype
models used for collapse assessment of reinforced concrete special moment
frame and reinforced concrete ordinary moment frame example structures
presented in Chapter 9. It includes an overview of critical modeling
decisions and a description of the calibration procedures used for beamcolumn
element models. Index archetype models represent index archetype
configurations, which are developed with careful consideration of structural
configuration issues (Appendix C) and seismic behavioral effects (Appendix
D).
Although the information is specific to reinforced concrete moment frame
systems, the example is illustrative of the issues and considerations typical of
nonlinear collapse simulation for any structural system type. Similar
procedures were used to create nonlinear models for the wood light-frame
system presented in Chapter 9, and the steel special moment frame system
presented in Chapter 10.
E.2 Structural Modeling Overview
Structural models should be capable of simulating the accrual of structural
damage and the resulting sidesway collapse when a structure is subjected to
severe ground shaking. Identification of key deterioration and collapse
modes, as described in Appendix D, is an important precursor to the choice
of the nonlinear analysis model.
The seismic-force-resisting system for reinforced concrete moment frames is
represented by a two-dimensional, three-bay frame. The destabilizing Pdelta
effects are modeled using a leaning column. At the element level,
frames are modeled with the following features illustrated in Figure E-1:
beam-column elements with concentrated inelastic rotational hinges at each
end and finite size beam-column joints that employ five concentrated
inelastic springs to model joint panel shear distortion and bond slip at each
E-2 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
face of the joint. The collapse behavior is simulated using the Open System
for Earthquake Engineering Simulation (OpenSees, 2006) software.
Figure E-1 Schematic diagram illustrating key elements of nonlinear frame
model for reinforced concrete frame systems.
Although not shown in Figure E-1, the effect of foundation flexibility on the
archetype models is incorporated using elastic, semi-rigid rotational springs
at the base of the column. For shorter frame structures (fewer than 4 stories),
the stiffness of the rotational springs is determined from assumed grade beam
and soil stiffness. For taller buildings, the structure is assumed to have a
basement, exterior columns are assumed to be fixed (connected to the
basement wall), and the properties of the rotational spring attached to interior
columns are computed based on the basement column and beam stiffnesses.
E.3 Beam-Column Element Model
Since the damage is likely to concentrate in the beams and columns of the
reinforced moment frame structures, accurate modeling of the inelastic
effects in beam and column elements is an essential component of collapse
modeling of these structures. As shown in Figure E-1, the beam-column
elements of the lateral resisting frame are modeled with lumped plasticity
elements in the plastic hinge locations. Lumped plasticity elements are
frequently used in structural analysis models, and their use here follows the
precedent established in ASCE/SEI 41-06 Seismic Rehabilitation of Existing
Buildings (ASCE, 2006b) and other guideline documents. When calibrated
properly, they are capable of capturing degradation of strength and stiffness
that is essential to collapse modeling. Their properties can also be easily
modified in a sensitivity analysis to determine the effects of uncertainties in
material modeling.
Researchers have also used a variety of other methods to simulate cyclic
response of reinforced concrete beam-columns, including creating fiber
models which can capture cracking behavior and the spread of plasticity
throughout the element (see e.g., Filippou, 1999) The decision to use a
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-3
Moment Frame Systems
lumped-plasticity approach here was based on simplicity and an inherent
limitation in the fiber element formulation which makes simulation of the
strain softening associated with rebar buckling difficult. The choice of
element model should be carefully evaluated for any given structural system,
and with careful consideration of available simulation technologies.
In this section, the properties of the lumped plasticity elements used to model
the example reinforced concrete special moment frame structures are
illustrated, including descriptions of both the element model used and the
process through which the key modeling parameters were calibrated. The
calibration process described here demonstrates the level of detail and
sophistication that is possible in studies of this type. Depending on the
structural system and type of analytical model, a greater or lesser degree of
detail may be required.
E.3.1 Element and Hysteretic Model
Lumped plasticity element models for the reinforced concrete special
moment frame structures utilize a material model developed by Ibarra,
Medina, and Krawinkler (2005), and implemented in OpenSees. This model
is chosen because it is capable of capturing the important modes of
deterioration that precipitate sidesway collapse of reinforced concrete frames,
but one could imagine using another material model or software platform that
also met these requirements.
Figure E-2 shows the tri-linear monotonic backbone curve and associated
hysteretic rules of the model, which permit versatile modeling of cyclic
behavior. For simulating structural collapse, the most important aspect of
this model is the post-peak response, which enables modeling the strain
softening behavior associated with concrete crushing, rebar buckling and
fracture, and/or bond failure. The model also captures four basic modes of
cyclic deterioration: (1) strength deterioration of the inelastic strain
hardening branch; (2) strength deterioration of the post-peak strain softening
branch; (3) accelerated reloading stiffness deterioration; and (4) unloading
stiffness deterioration. Cyclic deterioration is based on an energy index that
includes normalized energy dissipation capacity (.) and an exponent term to
describe how the rate of cyclic deterioration changes with accumulation of
damage (c).
This element model requires the specification of seven parameters to control
the monotonic and cyclic behavior of the model: My, .y, Mc/My, .cap,pl, .pc, .,
and c. The post-yield and post-capping stiffnesses are quantified by Mc/My
and .pc; Ks and Kc can be easily computed as:
E-4 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
. ... . . Ks . Ke .y .cap,pl Mc .My /My (E-1)
Kc . .Ke .. y . pc ..Mc /My . . (E-2)
Symbols are defined in Figure E-2 and in the list of symbols at the end of the
document.
0
0.2
0.4
0.6
0.8
1
1.2
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Chord Rotation (radians)
Normalized Moment (M/My)
Kc
.cap
p
l
Ks
My
Ke
.y
Mc
.pc
.cap
Figure E-2 Monotonic and cyclic behavior of component model used to
model reinforced concrete beam-column elements.
In order to determine the appropriate values of each of these parameters,
model parameters were carefully calibrated to 255 experimental tests of
reinforced concrete columns. These calibrations were then used to develop
empirical equations relating the design parameters of a beam-column to the
modeling parameters needed for input in the lumped plasticity model.
Calibrations are based on mean values. This type of calibration procedure
l
“strain hardening” “strain softening”
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-5
Moment Frame Systems
can be used in conjunction with any type of material model, provided that
there is sufficient test data available.
E.3.2 Calibration of Parameters for the Reinforced Concrete
Beam-Column Element Model
Experimental Database
Parameters of the material model are calibrated with data from the PEER
Structural Performance Database (PEER, 2006a; Berry et al., 2004),
assembled by Berry and Eberhard. This database includes the results of
cyclic and monotonic tests of 306 rectangular columns and 177 circular
columns. For each column test, the database reports the force-displacement
time history, the column geometry and reinforcement information, the failure
mode, and other relevant information. All data is converted to that of an
equivalent cantilever, regardless of experimental setup.
From this database, rectangular columns failing in a flexural mode (220 tests)
or in a combined flexure-shear mode (35 tests) were selected, for a total of
255 tests. These tests cover the typical range of column design parameters:
0.0 < . < 0.7, 0.0 < P/Pb < 2.0, 1.5 < Ls/H < 6.0, 20 < f’c (MPa) < 120, 340 <
fy (MPa) < 520, 0.015 < . < 0.043, 0.1 < s/d < 0.6, 0.002 < .sh < 0.02. These
symbols are defined in the list of symbols at the end of this document.
Calibration Procedure
In order to calibrate the element model parameters, each column test in the
database was modeled as a cantilever column in OpenSees, idealized using
an elastic element and a zero-length Ibarra model plastic hinge at the column
base. The properties of the plastic hinge are the subject of this calibration
effort. This appendix provides only a brief description of the calibration
process and the interested reader is referred to Haselton et al. (2007) for more
details.
The calibration of the beam-column element model to the data from each
experimental test was done systematically and, referring to Figure E-3, each
test was calibrated according to the following standardized procedure. First,
the yield shear force (1) was estimated visually from the experimental results.
Then, the “yield” displacement (2) was calibrated as the point of a significant
observed change in lateral stiffness, i.e., where rebar yielding or significant
concrete crushing occurs. Calibration of this point often required judgment,
since concrete response is nonlinear well before rebar yielding. In addition,
displacement at 40% of the yield force (3) was calibrated to provide an
estimate of initial stiffness. In step (4), the post-yield hardening stiffness was
visually calibrated.
E-6 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
-0.1 -0.05 0 0.05 0.1
-300
-200
-100
0
100
200
300
Shear Force (kN)
Column Drift (displacement/height)
Test 170 (kN, mm, rad):
.y = 0.0131
.stf 40 = 0.0012
My = 3.462725e+005 kN-mm
Mc/My = 1.21
.cap,pl = 0.060 (LB = 0)
.pc = 0.225
. = 81, c = 1.00
isPDeltaRemoved = 1
Experimental Results
Model Prediction
Figure E-3 Example of calibration procedure; calibration of reinforced
concrete beam-column element to experimental test by
Saatcioglu and Grira, specimen BG-6.
The next step (5) was to calibrate the normalized cyclic energy dissipation
capacity, .. This parameter is defined such that the total energy dissipated in
a cycle is given by Et = .My.y. The parameter c describes how energy
dissipation changes in subsequent cycles (Ibarra et al., 2005). The element
model allows cyclic deterioration coefficients . and c to be calibrated
independently for each of the four cyclic deterioration modes. However,
based on a pilot study of 20 columns, c = 1.0 was found to be acceptable for
columns failing in flexure and flexure-shear modes. The deterioration rate, .,
was set to be equal for the basic strength and post-capping strength
deterioration modes, following recommendations by Ibarra (2003). Based on
observations of the hysteretic response of the reinforced concrete columns,
no accelerated stiffness deterioration was used. Unloading stiffness
deterioration was neglected to avoid an error in the OpenSees
implementation of the element model; in future, unloading stiffness
deterioration should be employed, as it leads to better modeling of cyclic
response. These simplifications reduce the calibration of cyclic energy
dissipation capacity to one parameter (.). When calibrating ., the aim was to
match the average deterioration for the full displacement history, but with a
slightly higher emphasis on matching the deterioration rate of the later, more
damaging, cycles.
The final step of the calibration process (6) involved quantification of the
capping point and the post-capping deformation capacity. It has been shown
(e.g., Haselton et al., 2007) that incorrect calibration of strength deterioration
can have a significant impact on structural response prediction, and the
(1)
(2)
(3)
(4)
(5)
(6)
(6)
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-7
Moment Frame Systems
calibration procedure carefully distinguished between in-cycle and cyclic
strength deterioration. The capping point and post-capping stiffness were
only calibrated when a negative post-peak stiffness was clearly observed in
the data, such that strength loss occurs within a single cycle (sometimes
called “in-cycle strength deterioration”). Often the test specimen did not
undergo sufficient deformations for a capping point to be observed. In such
cases where the data do not reveal the capping point, a lower-bound value of
the capping point was determined.
It should be noted that the hinge model is based on the definition of a
monotonic backbone and cyclic deterioration rules. This calibration process
relied primarily on cyclic tests with many cycles to calibrate both the
monotonic backbone parameters and the cyclic deterioration rules. As a
result, the monotonic backbone and the cyclic deterioration rules are
interdependent, and the approximation of the monotonic backbone depends
on cyclic deterioration rules assumed and, to some extent, the displacement
pattern used in experimental tests. This approximation of the monotonic
backbone from cyclic data is not ideal, but is necessary because sufficient
data are not available to calibrate the monotonic and cyclic behavior
separately.
A full table of calibrated model parameters for each of the 255 experimental
tests used can be found in the extended report on this study (Haselton et al.,
2007).
Interpretation of Calibration Results and Regression Analysis
Calibrated model parameters from the 255 column tests are used to create
empirical equations that predict model parameters based on the column
design parameters. The functional form used in regression analysis was
carefully determined based on trends in the data and isolated effects of
individual variables, previous research and existing equations, and judgment
based on mechanics and expected behavior. The regression analysis was
performed using the natural logarithm of the model parameter, and the
logarithmic standard deviation quantifies the uncertainty. More details on
how the data were dissected to create empirical regression equations to
predict each model parameter is available in Haselton et al. (2007).
As noted previously, those test specimens in which post-capping behavior
was not observed were calibrated with a lower-bound value for deformation
capacity. In the regression analysis, this lower-bound calibration data were
given special consideration. In order to take advantage of lower-bound data,
without unnecessarily biasing the results, the deformation capacity equations
presented here are based on all data, and the prediction uncertainties are
E-8 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
reported based on only the data with an observed capping point in order to
avoid reporting the artificially high uncertainty associated with the lowerbound
data. Note that this approach is still appropriate for elements with
high deformation capacity because the lower-bound data underestimates the
true deformation capacity.
Proposed Equations
Predictive equations were developed for each of the parameters of the
element model for reinforced concrete columns. Each equation includes all
statistically significant parameters, unless otherwise noted. The extended
report on this study also includes more simplified equations (Haselton et al.,
2007).
Effective Initial Stiffness (EIy/EIg and EIstf/EIg)
Due to nonlinearities in reinforced concrete behavior associated with
cracking, the definition of the stiffness of a reinforced concrete element
depends on the load and deformation level. In this work, two values of
effective stiffness are defined: (1) the secant stiffness to the yield point of the
component (termed EIy or Ke) and (2) the secant stiffness to 40% of the yield
force of the component (termed EIstf). EIstf is used as the initial stiffness in
the creation of the structural models for reinforced concrete frames in
Chapter 9, based on a study by Haselton et al. (2007). The component
stiffness includes all modes of deformation including flexure, shear, and
bond-slip.
The equation for secant stiffness to yield depends on both axial load ratio and
shear span ratio of the column, and is given as follows:
' 0.07 0.59 0.07 y s
g g c
EI P L
EI A f H
. . . . . . . . . . . . .. .. . .
(E-3)
where 0.2 0.6 y
g
EI
EI
. .
This equation represents the mean value of effective stiffness to yield. The
prediction uncertainty, assuming the residuals are lognormally distributed, is
given by the logarithmic standard deviation: sLN = 0.28. R2, a measure of the
extent to which the proposed equation explains the data, is 0.80. For this and
all equations to follow, these values are reported below in Table E-1. The
upper and lower limits on the stiffness were imposed because there is limited
data for columns with very low axial loads and, at high levels of axial load,
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-9
Moment Frame Systems
the positive trend diminishes and the scatter in the data is large. The limits
were chosen based on a visual inspection of data.
The effective initial stiffness, defined as the secant stiffness to 40% of the
yield force of a reinforced concrete column, can be predicted as follows:
' 0.02 0.98 0.09 stf s
g g c
EI P L
EI A f H
. . . . . . . . . . . . .. .. . .
(E-4)
where 0.35 0.8 stf
g
EI
EI
. .
For a typical column, Equation (E-4) predicts the initial stiffness to be
approximately 1.7 times stiffer than the secant stiffness to yield (Equation
(E-3)).
The effective stiffness of reinforced concrete columns has been the subject of
much research; for comparison, selected studies are presented here. The
guidelines in FEMA 356 Prestandard and Commentary for Seismic
Rehabilitation of Buildings (FEMA, 2000) permit the use of standard
simplified values based on the level of axial load for linear analysis: 0.5EcIg
when P .Ag f 'c . . 0.3 , and 0.7EcIg when P .Ag f 'c . . 0.5 . More recently,
Elwood and Eberhard (2006) proposed an equation for effective stiffness that
includes all components of deformation (flexure, shear, and bond-slip),
where the effective stiffness is defined as the secant stiffness to the yield
point of the component. Their equation proposes 0.2EcIg when
P .Ag f 'c . . 0.2 and 0.7EcIg when P .Ag f 'c . . 0.5 , with a linear transition
between these two extremes.
The equations proposed here for EIy are similar to those recently proposed by
Elwood and Eberhard (2006), but also includes an Ls/H term. Due to this
additional term, the proposed Equation (E-3) has a lower prediction
uncertainty of sLN = 0.28, as compared to the coefficient of variation of 0.35
reported by Elwood et al. The stiffness predictions in FEMA 356 are much
higher than the values of EIy predicted in this study, and slightly larger than
the predicted values of EIstf. Elwood and Eberhard (2006) found that most of
this difference can be explained if it is assumed that the FEMA 356 values
only include flexural deformation, and do not account for the bond-slip
deformations, which can account for a significant proportion of an element’s
flexibility.
E-10 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
Flexural Strength (My)
Panagiotakos and Fardis (2001) have published equations to predict flexural
strength. Their equation works well, so for this study their method is used to
predict My. When comparing the calibrated values to flexural strength for the
255 columns to predictions by Panagiotakos and Fardis (2001) for these
columns, the median ratio of My / My,(Fardis) is 0.97, and the coefficient of
variation is 0.36.
Plastic rotation capacity (.cap,pl)
The following equation is proposed for predicting plastic rotation capacity,
including all parameters that are statistically significant:
. .. .. .
. . . . . . '
0.43
,
0.01 0.1 10.0
0.12 1 0.55 0.16 0.02 40
0.54 units c 0.66 n 2.27
v
cap pl sl sh
c f s
a
.
. . . . .
(E-5)
Possible correlations between sh . and sn were verified to be small to
eliminate concerns regarding co-linearity in regression. This equation is
based on all data, including the lower-bound data for plastic rotation
capacity, as discussed previously. The shear-span ratio (Ls/H) is notably
absent from the predictive equations for rotation capacity. The stepwise
regression process consistently showed the shear-span ratio to be statistically
insignificant. These findings differ from those of Panagiotakos and Fardis
(2001).
The experimental data used in this study is limited to tests of columns with
symmetrical arrangements of reinforcement and, as a result, Equation (E-5)
applies only to columns with symmetric reinforcement. To eliminate the
symmetric reinforcement limitation from the rotation capacity equation, it is
proposed to multiply the rotation capacity obtained from these equations by
the term proposed by Fardis and Biskini (2003), which accounts for the ratio
between the areas of compressive and tensile steel.
It is also useful as verification to compare the predicted rotation capacity to
the ultimate rotation capacity predicted by Fardis and Biskini based on their
study of more than 700 reinforced concrete elements including beams,
columns, and walls (Fardis and Biskini ,2003; Panagiotakos and Fardis,
2001). The equation based on these studies consistently predicts higher
values, which is expected since Equation (E-5) predicts the capping point
where the ultimate rotation capacity equation is based on the ultimate (20%
strength loss) point. To make a more consistent comparison, the predictions
of the ultimate rotation (at 20% strength loss) are combined with the
calibrated values of post-capping slope (.pc) to back-calculate a prediction of
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-11
Moment Frame Systems
.cap,pl from the equation in Fardis and Biskini. This comparison shows that
the proposed equation here predicts lower deformation capacity, with a mean
ratio of 0.94 and a median ratio of 0.69. Based on this comparison, Equation
(E-5) may still include some conservatism. This conservatism exists even
though the predicted deformation capacities are already much higher than
what is typically used; for example, values in ASCE/SEI 41 are typically less
than one-half of those shown in Table E-2. Note: The recent supplement to
ASCE/SEI 41 for concrete elements, particularly those with non-ductile
detailing, has reduced some of the conservatism in ASCE/SEI provisions
(Elwood et al., 2007).
Post-capping Rotation Capacity (.pc)
Previous research on predicting post-capping rotation capacity has been
limited, despite its important impact on predicting collapse capacity. The
key parameters considered in the development of an equation for postcapping
response are those that are known to most impact deformation
capacity: axial load ratio (.), transverse steel ratio (.sh), rebar buckling
coefficient (sn), stirrup spacing, and longitudinal steel ratio. The equation is
based on only those tests where a post-capping slope was observed. The
proposed equation for post-capping rotation capacity is as follows:
. .. . . .1.02 0.76 0.031 0.02 40 0.10 v
. pc . . .sh . (E-6)
The upper bound imposed on Equation (E-6) is due to lack of reliable data
for elements with shallow post-capping slopes. The upper limit of 0.10 may
be conservative for well-confined, conforming elements, but existing test
data does not justify using a larger value.
Post-Yield Hardening Stiffness
Regression analysis shows that axial load ratio and concrete strength
statistically impact the prediction of hardening stiffness (Mc/My). Even so,
inclusion of both of the parameters scarcely improved the regression
analysis, so a constant value of Mc My is recommended. Equation (E-7) led
to an acceptably low prediction uncertainty of sLN = 0.10.
Mc My .1.13 (E-7)
Cyclic Energy Dissipation Capacity
Based on the observed trends in the data, the following equation is proposed
for the mean energy dissipation capacity, including all statistically significant
predictors:
E-12 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
. . (131.0)(0.18). (0.26)s d (0.57)Vp Vn (61.4).sh,eff (E-8)
As discussed previously, this value is calibrated for both types of strength
deterioration and should be used with c = 1.0.
Discussion of Proposed Equations
The proposed equations can be evaluated to determine (a) how well they
reflect the mean tendencies in the data, and (b) whether they provide suitably
small prediction errors. Table E-1 reports the median ratio of predicted to
observed values (from model calibrations) for each proposed equation. The
regression analyses (e.g., Equations (E-3) – (E-8)) assumed the prediction
uncertainty is lognormally distributed, so this median ratio should be close to
1.0; this ratio varies between 1.00 and 1.06 for the proposed equations,
showing that the predictive equations have little bias.
Table E-1 Prediction Uncertainties and Bias in Proposed Equations
Equation
Median
(predicted/
observed)
s LN R2
Effective Stiffness to Yield (E-1) 1.03 0.28 0.80
Effective Stiffness to 40% Yield (E-2) 1.02 0.33 0.59
Plastic Rotation Capacity (E-3) 1.02 0.54 0.60
Post-Capping Rotation Capacity (E-4) 1.00 0.72 0.51
Post-Yield Hardening Stiffness (E-5) 1.01 0.10 n/a
Cyclic Energy Dissipation Capacity (E-6) 1.06 0.47 0.51
Table E-1 also shows that the prediction uncertainty is large for many of the
important parameters. For example, the prediction uncertainty (sLN) for
plastic deformation capacity is 0.54. Previous research has shown that these
large uncertainties in element deformation capacity cause similarly large
uncertainties in collapse capacity (Ibarra, 2003; Goulet et al., 2006; Haselton
et al., 2007). These large uncertainties are associated with wide variability in
the physical phenomena and limitations in the available test data. With the
availability of more test data it may be possible to reduce these uncertainties.
Due to its particularly large effect on collapse assessment, it is useful to also
examine the prediction bias for selected subsets of the data for .cap,pl as given
in Equation (E-5). For non-conforming elements (i.e., those with .sh <
0.003) the prediction is unbiased, with a median ratio of predicted to
observed values of 1.02. The plastic rotation capacity equation tends to
overpredict the plastic rotation capacity by approximately 12% for
conforming elements (.sh > 0.006), and underpredicts the plastic rotation
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-13
Moment Frame Systems
capacity by approximately 10% for elements with extremely high axial load
(i.e., axial load ratios above 0.65). Considering the large uncertainty in the
prediction of plastic rotation capacity and the small number of datapoints in
some of the subsets, these computed biases seem reasonable. With more
data, the prediction error and biases could be further reduced.
To illustrate the impact of column design variables on the model parameters
predicted by Equation (E-3) through (E-8), modeling parameters obtained for
an 8-story reinforced concrete special moment frame perimeter system are
shown in Table E-2. When the proposed equations are used, a typical
column at the base should be modeled with an effective stiffness to yield of
35% of EIg, a post-yield hardening ratio of 1.13, a plastic rotation capacity of
0.085, a post-capping rotation capacity of 0.10 and . equal to 154.
Table E-2 Predicted Model Parameters for an 8-Story Reinforced Concrete
Special Moment Frame Perimeter System (Interior Column, 1st-
Story Location)
Design
Parameter
Value EIstf/EIg Mc /My .cap,pl .pc .
All Baseline1 0.35 1.13 0.085 0.100 154
v
0.0 0.35
1.13
0.095 0.100 170
0.3 0.52 0.055 0.100 104
.sh
0.002
0.35 1.13
0.042 0.059
0.010 0.078 0.100 154
0.020 0.105 0.100
Ls/h
2 0.35
1.13 0.085 0.100 154
6 0.58
s/d
0.1
0.35 1.13 0.085 0.100
163
0.4 106
0.6 80
.
0.01
0.35 1.13
0.080
0.100 154
0.03 0.094
1Baseline values: . = 0.06, .sh = 0.0121, Ls/h = 2.7, s/d = 0.14, . = 0.018, f'c
= 35 MPa, asl = 1, sn = 7.5.
Table E-2 also illustrates how changes in each design parameter impact
predicted model parameters. For example, suppose various design
parameters were changed such that the column under consideration in the
first row of Table E-2 has a different axial load ratio, amount of transverse
reinforcement, or column reinforcement ratio; these changes would lead to
modifications in the parameters used to model that column. Within the range
of column parameters considered in Table E-2, the predicted plastic rotation
E-14 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
capacity varies from 0.042 to 0.105 radians. Axial load ratio (.) and the
lateral confinement ratio (.sh) have the largest effect on the predicted value of
.cap,pl, while concrete strength (f’c), rebar buckling coefficient (sn), and
longitudinal reinforcement ratio (.) have less dominant effects (f’c and sn are
not shown here). Table E-2 also shows the effects of the design parameters
on initial stiffness, hardening stiffness ratio, post-capping stiffness, and
cyclic deterioration parameters.
Summary and Limitations
The calibration process provided a comprehensive set of equations capable of
predicting the parameters of a lumped plasticity element model for a
reinforced concrete beam-column, with the empirical predictive equations
having the ability to capture the effects of design aspects such as detailing
and level of axial load. These equations are proposed for use with the
element model developed by Ibarra et al. (2003, 2005), and can be used to
model cyclic and in-cycle strength and stiffness degradation to capture
element behavior up to the point of structural collapse. Even so, these
equations are general and can be used with slight modifications with most
lumped plasticity models currently used in research.
The predictive empirical equations proposed here provide a critical link
between column design parameters and element modeling parameters,
facilitating the creation of nonlinear structural models of reinforced concrete
frame elements. The empirical equations predict the mean value of each
parameter, so the prediction uncertainty associated with each equation is also
quantified and reported. This provides an indication of the uncertainty in the
prediction of model parameters, and can be used in sensitivity analyses and
propagation of structural modeling uncertainties.
As with any study, there are limitations in terms of the applicability of these
equations, which are discussed briefly here.
The equations developed here are based on a comprehensive database
assembled by Berry et al. (Berry et al., 2004; PEER, 2006a). Even so, the
range of column parameters included in the database is limited, and the
derived equations may not be applicable outside the range of column
parameters considered.
The equations are also limited more generally by the number of test
specimens available that have an observed capping point. There are only a
small number of experimental tests with clearly observable negative postcapping
stiffness. For model calibration and understanding of element
behavior, it is important that future testing continue to deformation levels
FEMA P695 E: Nonlinear Modeling of Reinforced Concrete E-15
Moment Frame Systems
large enough to clearly show the negative post-capping stiffness. In addition,
there is very limited data that show post-peak cyclic deterioration behavior.
With additional data, it may be possible to further remove some of the
conservatism still in the proposed equations and reduce the prediction
uncertainties. A further limitation is associated with the testing protocols;
virtually all of the available test data are based on a cyclic loading protocol
with many cycles and 2-3 cycles per deformation level. This type of loading
may not be representative of the type of earthquake loading that typically
causes structural collapse, which would generally only contain a few large
displacement cycles before collapse occurs. Tests conducted under a variety
of loading histories will lead to a better understanding of how load history
impacts cyclic behavior, and provide a basis for better development and
calibration of the element model cyclic rules.
In addition to the above mentioned issues, it is also important to remember
that the empirical equations developed in this work are all based on
laboratory test data where the test specimen was constructed in a controlled
environment, and thus indicative of a high quality of construction. Actual
buildings are constructed in a less controlled environment, so we expect the
elements of actual buildings to have a lower level of performance than that
predicted using the equations presented here. This work does not attempt to
quantify this difference in performance coming from construction quality.
E.4 Joint Modeling
Nonlinear models of reinforced concrete structures employ a twodimensional
joint model developed and implemented by Lowes et al. (2004).
This model accounts for the finite joint size, and includes rotational springs
and systems of constraints for direct modeling of the shear panel and bondslip
behavior. Figure E-4 shows a schematic diagram of this model.
Figure E-4 Schematic diagram of joint model.
E-16 E: Nonlinear Modeling of Reinforced Concrete FEMA P695
Moment Frame Systems
The joint model and parameters are based on a careful review of previous
research. This approach is different than the detailed calibration study
presented earlier to determine the parameters for beam-column elements.
E.4.1 Shear Panel Spring
Current building code provisions require that the joint shear capacity be
based on capacity design principles for special moment frames, so if properly
implemented, the joint shear demand should never exceed the capacity. A
review of available research finds that the joint shear capacity designs
provisions are conservative, and should be able to prohibit joint shear failure
in all properly designed reinforced concrete special moment frame structures.
For example, Brown and Lowes (2006) reviewed results of 45 experimental
tests of conforming joints, finding that not one of these conforming joints
exhibited damage requiring joint replacement. As a result of observations
like these, the joint shear modeling is not judged to be a critical part of the
overall behavior of the frame; damage in the joint shear panel will increase
the flexibility of the frame, but the joint shear panel is not a dominant
damage/failure mode (see also Appendix D).
For reinforced concrete special moment frame structures, joint models are
developed to accurately reflect system stiffness. For simplicity, joints are
modeled as elastic elements with the cracked stiffness based on simple
mechanics (Umemura and Aoyama, 1969). Reinforced concrete ordinary
moment frame structures are modeled to account for the deterioration that
may occur in the joints of those structures. For readers interested in a more
detailed approach, the modified compression field theory (MCFT) (Vecchio
and Collins, 1986; Stevens et al., 1991; Altoontash, 2004, Lowes et al., 2004;
and Lowes and Altoontash, 2003) is often used to develop panel shear
models.
E.4.2 Bond-Slip Spring Model
The effect of bond-slip in joints and column footings is to decrease the preyield
and post-yield stiffnesses and increase the plastic rotation capacity of
reinforced concrete elements. In this study, bond-slip deformations are
directly included in the plastic hinge calibration and thus do not need to be
separately defined.
By lumping the effects of bond-slip in the plastic hinge model, it is
inherently assumed that the bond-slip component of deformation is linear
from zero to the yield load. This is a simplification, and a quadrilinear model
could instead be used to capture the nonlinear bond-slip behavior prior to
rebar yielding.
FEMA P695 F: Collapse Evaluation of Individual Buildings F-1
Appendix F
Collapse Evaluation of
Individual Buildings
F.1 Introduction
Although developed as a tool to establish seismic performance factors for
generic seismic-force-resisting systems, the Methodology could be readily
adapted for collapse assessment of an individual building system. As such, it
could be used to demonstrate adequate collapse performance for a new
building designed using performance-based design methods (as permitted by
ASCE/SEI 7-05), or for collapse safety evaluation of an existing building.
This appendix describes an adaptation of the Methodology for use in collapse
evaluation of an individual building system.
F.2 Feasibility
Buildings designed or evaluated using performance-based methods are often
large or important structures. Such projects typically utilize detailed models
for analysis of the building, and peer review is often required. In this way,
they are already set up to utilize many aspects of the Methodology.
In contrast to index archetype models, analytical models for individual
buildings might have more elements based on the actual configuration and
specific geometry of the building. Collapse evaluation can be performed
using two-dimensional or three-dimensional models of the building.
Depending on the needs of the project, individual building models may, or
may not include a high level of sophistication in modeling nonlinear
behavior.
F.3 Approach
The Methodology is based on the concept of collapse level ground motions,
defined as the level of ground motions that cause median collapse (i.e., onehalf
of the records in the set cause collapse). For a building to meet the
collapse performance objectives of this Methodology, the median collapse
capacity must be an acceptable ratio above the Maximum Considered
Earthquake (MCE) ground motion demand level (i.e., the adjusted collapse
margin ratio, ACMR, must exceed acceptable values).
F-2 F: Collapse Evaluation of Individual Buildings FEMA P695
By starting with an acceptable collapse probability for MCE ground motions,
and working backwards through the Methodology, values of the spectral
shape factor, SSF, and collapse margin ratio, CMR, can be calculated to
determine the ground motion intensity corresponding to median collapse.
The Methodology can be “reverse engineered” to determine the level of
ground motions for which not more than one-half of the records should cause
collapse.
By scaling the record set to this level, trial designs for a subject new building
can be evaluated. If the analytical model of the trial design survives one-half
or more of the records without collapse, then the building has a collapse
probability that is equal to (or less than) the acceptable collapse probability
for MCE ground motions, and meets the collapse performance objectives of
the Methodology.
Similarly, an existing building can be evaluated for records scaled to the
intensity corresponding to median collapse. If the analytical model of the
existing building survives one-half or more of the records, then the existing
building meets the collapse performance objectives of the Methodology. If
different seismic criteria are needed, as is often the case with existing
buildings, the process can begin with any probability of collapse deemed
acceptable for the project.
F.4 Collapse Evaluation of Individual Building
Systems
The process for collapse evaluation of an individual building system is
summarized in the following steps.
F.4.1 Step One: Develop Nonlinear Model(s)
Development of a representative model (or models) of the building must
incorporate the nonlinear behavioral characteristics of building-specific
components.
. Two-Dimensional Versus Three-Dimensional Models. It is likely that
the project has developed a three-dimensional (3-D) model of the
building (for design), which could be used for collapse evaluation. Twodimensional
(2-D) models could be used, but such models would need to
address response in both horizontal directions (e.g., two 2-D models) and
account for torsion and potential coupling of bi-directional response.
. Gravity System. Since the construction of the building is known,
elements of the gravity system can be incorporated into the buildingspecific
model. Whether or not they are directly simulated, potential
FEMA P695 F: Collapse Evaluation of Individual Buildings F-3
collapse modes of the gravity system must be included in the collapse
assessment process for an individual building. This is necessary because
displacement compatibility requirements do not prohibit gravity system
collapse from becoming a controlling behavior mode; they only address
compatibility of displacements at the MCE ground motion level, not at
collapse level ground motions. P-delta effects must consider the full
weight of the building.
Inclusion of gravity system collapse modes is a departure from the
formal application of the Methodology, which does not account for
gravity system collapse. Accounting for collapse modes of the gravity
system is typically accomplished by using limit state criteria for nonsimulated
collapse modes. Practically speaking, there is little
experimental data on which to base limit state criteria for many gravity
system collapse modes. In these cases, a peer review team could assist in
determining appropriate criteria for use in the collapse assessment.
. Nonlinear Elements. Model(s) should incorporate nonlinear properties
for all elements of a seismic force-resisting system (and gravity system)
that cannot be shown to remain fully elastic in a near-collapse condition.
F.4.2 Step Two: Define Limit States and Acceptance Criteria
. Limit States. Accurately simulating sidesway collapse requires
sophisticated modeling of element degradation and strain softening.
Depending on the availability of component test data, and ability to
develop high-fidelity representations of individual building components,
use of non-simulated collapse limit states (Chapter 5) can be considered
in lieu of direct simulation of collapse.
. Acceptable Collapse Probability. The maximum acceptable probability
of collapse should be determined. An acceptable collapse probability of
10% is consistent with the collapse performance objectives of this
Methodology.
F.4.3 Step Three: Determine Total System Uncertainty and
Acceptable Collapse Margin Ratio
Total System Uncertainty (.TOT). The value of total system collapse
uncertainty, .TOT, is determined in accordance with Section 7.3, based on
quality of the design requirements, quality of the test data used to
develop nonlinear properties, and quality of the nonlinear model.
The quality of design requirements (Chapter 3) could be “Superior” for
buildings generally conforming to materials and detailing requirements
of ASCE/SEI 7-05. In the case of an existing building, a judgment
F-4 F: Collapse Evaluation of Individual Buildings FEMA P695
would need to be rendered on the quality of the code under which the
building was originally constructed. The quality of test data (Chapter 3)
could range from “Good” to “Fair” depending on the extent to which
available test data accurately capture degrading behavior of existing
building components. For an individual building, the extent to which a
model captures the full range of the “design space” would be considered
“High,” because the building configuration is a known quantity. In this
case, when determining model quality in accordance with Chapter 5, the
first row in Table 5-3 should be used.
. Acceptable Value of Adjusted Collapse Margin Ratio (e.g.,
ACMR10%). The acceptable value of the adjusted collapse margin ratio
should be determined. Based on an acceptable collapse probability of
10%, ACMR10% is consistent with the collapse performance objectives of
this Methodology.
F.4.4 Step Four: Perform Nonlinear Static Analysis (NSA)
. Nonlinear Static Analysis. A nonlinear static (pushover) analysis is
performed to check nonlinear behavior of the model, and to verify that
all elements assumed to be essentially elastic have not yielded at the
point that a collapse mechanism develops in the structure. Pushover
analyses should be completed in both horizontal directions.
. Structure Period-Based Ductility, .T. The period-based ductility, .T, is
determined from pushover analysis results, for both horizontal directions,
in accordance with Chapter 6 and including non-simulated limit states as
appropriate.
. Spectral Shape Factor, SSF. The spectral shape factor, SSF, for both
horizontal directions, is determined based on the period-based ductility,
.T, the building Seismic Design Category, and the fundamental period, T,
in the direction of interest, in accordance with Section 7.2.2.
F.4.5 Step Five: Select Record Set and Scale Records
. Record Set Selection. If the building is located within 10 km of an
active fault, then the Near-Field record set should be selected for collapse
evaluation, otherwise the Far-Field record set should be used.
. Record Set Scaling. Both components of each record in the record set
should be scaled up to the required collapse level intensity at which onehalf
or more of the records must not cause collapse.
To scale the records to this level, all normalized records are multiplied by
the same scale factor, SF. The scaling factor, SF, should be computed
FEMA P695 F: Collapse Evaluation of Individual Buildings F-5
for both horizontal directions and the average value used to scale both
components of each record.
10%
3
MT
D NRT
SF ACMR S
C SSF S
. .
. . .
. .
(G-1)
where:
SF = record (and component) scale factor in the direction of interest,
required for collapse evaluation of an individual building system.
ACMR10% = acceptable value of the adjusted collapse margin ratio from
Section 7.4, corresponding to an acceptable collapse probability of 10%.
SSF = spectral shape factor in the direction of interest, as defined in
Chapter 7.
C3D = three-dimensional analysis coefficient, taken as 1.2 for threedimensional
analysis, and 1.0 for two-dimensional analysis.
SMT = MCE, 5% damped, spectral response acceleration at the
fundamental period, T, of the building in the direction of interest, as
defined in Section 11.4.3 of ASCE/SEI 7-05, including Site Class.
NRT S = median value of normalized record set, 5% damped, spectral
response acceleration at the fundamental period, T, of the building in the
direction of interest.
T = the fundamental period of the building in the direction of interest,
based on the limits of Section 12.8.2 of ASCE/SEI 7-05, computed in
accordance with Equation 5-5 in Chapter 5.
F.4.6 Step Six: Perform Nonlinear Dynamic Analysis (NDA)
and Evaluate Performance
. Nonlinear Dynamic Analysis. A nonlinear dynamic analysis of the
model(s) is performed separately for each scaled record of the record set.
The response is classified as either “collapse” or “non-collapse,” based
on sidesway collapse of the analytical model or through the use of nonsimulated
collapse component acceptance criteria. Ground motion
record pairs should be applied to two-dimensional and three-dimensional
models in accordance with Chapter 6.
. Collapse Evaluation. If less than one-half of the records cause collapse,
then the trial design (or the existing building) meets the collapse
performance objective, and the building has an acceptably low
probability of collapse for MCE ground motions. If one-half or more of
F-6 F: Collapse Evaluation of Individual Buildings FEMA P695
the records cause collapse, then the design does not meet the collapse
performance objective, and re-design and re-evaluation are required.
FEMA P695 Symbols G-1
Symbols
A = force normalized by effective seismic weight, W, corresponding to
arbitrary post-yield displacement, D, of the isolation system in the
horizontal direction under consideration, used to define bi-linear
spring properties of isolators
Ag = gross cross-sectional area of an element
asl = indicator variable (0 or 1) to signify possibility of longitudinal
rebar slip past the column end, where asl = 1 if slip is possible
(Panagiotakos, 2001)
Ay = force normalized by effective seismic weight, W, corresponding to
idealized yield displacement, Dy, of the isolation system in the
horizontal direction under consideration, used to define bi-linear
spring properties of isolators
ACMR = adjusted collapse margin ratio
ACMR10% = acceptable value of the adjusted collapse margin ratio (ACMR),
on average, for the performance group of interest
ACMR20% = acceptable value of the adjusted collapse margin ratio (ACMR)
for an individual archetype of the performance group of interest
b = element width
B1E = numerical coefficient as set forth in Table 18.6-1 of ASCE/SEI 7-
05 for the effective damping equal to .I
+ .V1 and period equal to
T
c = cyclic deterioration calibration term (exponent); describes the
change in the rate of cyclic deterioration as the energy dissipation
capacity is exhausted
C3D = 3-dimensional analysis coefficient, 1.2 for 3-dimensional analysis
(and 1.0 for 2-dimensional analysis)
Cd = deflection amplification factor (current values given in Table
12.2-1 of ASCE/SEI 7-05)
Cs = seismic response coefficient as determined in Section 12.8.1.1 of
ASCE/SEI 7-05)
G-2 Symbols FEMA P695
Ct = approximate period coefficient as determined in Table 12.8-2 of
ASCE/SEI 7-05)
Cu = upper-limit period coefficient as determined in Table 12.8-1 of
ASCE/SEI 7-05)
cunits = a units conversion variable that equals 1.0 when f’c is in MPa
units and 6.9 for ksi units
CI = confidence interval
CMR = collapse margin ratio
C0 = coefficient relating fundamental-mode (SDOF) displacement to
roof (MDOF) displacement of an index archetype model, as
defined in Section 6.3
d = column depth
D = effect of dead load for use in load combinations of Section 12.4 of
ASCE/SEI 7-05
D = arbitrary post-yield displacement, in inches, of the isolation
system in the horizontal direction under consideration used to
define bi-linear spring properties of isolators (Chapter 10)
DM = maximum displacement, in inches, at the center of rigidity of the
isolation system in the direction under consideration, as
prescribed by Eq. 17.5-3 of ASCE/SEI 7-05
DTM = total maximum displacement, in inches, of the isolation system in
the direction under consideration considering both translational
and torsional displacement, as prescribed by Eq. 17.5-6 of
ASCE/SEI 7-05
Dy = idealized yield displacement, in inches, of the isolation system in
the horizontal direction under consideration used to define bilinear
spring properties of isolators
EIg = gross cross-sectional moment of inertia
EIstf = effective cross-sectional moment of inertia such that the secant
stiffness is defined to 40% of the yield moment/force of the
component
EIy = effective cross-sectional moment of inertia that provides a secant
stiffness through the yield point
Et = total energy dissipation capacity
EXP = exponential
FEMA P695 Symbols G-3
Fa = short-period site coefficient (at 0.2-second period) as given in
Section 11.4.3 of ASCE/SEI 7-05
Fc = maximum strength (at capping point)
Fr = residual strength
Fv = long-period site coefficient (at 1.0-second period) as given in
Section 11.4.3 of ASCE/SEI 7-05
Fx = nonlinear static (pushover) analysis force at level x
Fy = yield strength of material
f’
c = compressive strength of unconfined concrete, based on standard
cylinder test
fy = yield stress of longitudinal reinforcement (Appendix E).
g = constant acceleration due to gravity
H = foundation loads due to lateral earth pressure, ground water
pressure, or pressure of bulk materials, Chapter 2 of ASCE/SEI 7-
05
h = element height
hn = height in feet above the base to the highest level of the structure
(ASCE/SEI 7-05)
hr = height above the base to the roof level of the archetype building
I = the importance factor in Section 11.5.1 of ASCE/SEI 7-05
K = residual strength present in material model, defined as a ratio of
My
Kc = post-capping stiffness, i.e., stiffness beyond .cap,pl
Ke = effective elastic secant stiffness to the yield point
Kpc = post-capping tangent stiffness
keff = effective stiffness of the isolation system, in kips/in., at
displacement, D, in the horizontal direction under consideration
kMmin = minimum effective stiffness, in kips/in., of the isolation system at
the maximum displacement in the horizontal direction under
consideration, as prescribed by Eq. 17.8-6 of ASCE/SEI 7-05
Ks = hardening stiffness, i.e., stiffness between .y
and .cap,pl
L = effect of live load for use in load combinations of Section 12.4 of
ASCE/SEI 7-05
G-4 Symbols FEMA P695
Ls = shear span, distance between column end and point of inflection
LN = natural logarithm
Mc = moment capacity at the capping point; used for prediction of
hardening stiffness
My = yield moment for Ibarra material model (nominal moment
capacity of the column)
My (Fardis) = yield moment as calculated based on predictive equations
(Panagiotakos and Fardis, 2001)
Mp = plastic moment capacity of element
mx = mass of index archetype model at level x
N = number of levels (stories) in an index archetype model
NMi = normalization factor of the ith record of the set of interest
NTH1,i = normalized ith record, horizontal component 1
NTH2,i = normalized ith record, horizontal component 2
P = perimeter frame system (Chapter 9)
P = axial load
Pb = axial load at the balanced condition
PGVPEER = peak ground velocity of the ith record based on the geometric
mean of the two horizontal components considering different
record orientations, based on the PEER-NGA database
QE = the effect of horizontal seismic force from total design base shear,
V, for use in load combinations of Section 12.4 of ASCE/SEI 7-
05
R = response modification coefficient (current values given in Table
12.2-1 of ASCE/SEI 7-05)
Reff = effective R factor which accounts for the additional strength
caused by the minimum base shear requirement
RI = numerical coefficient related to the type of seismic force-resisting
system above the isolation system (i.e., three-eights of the R value
given in Table 12.2.-1 of ASCE/SEI 7-05, not to exceed 2.0, need
not be taken as less than 1.0)
RC = reinforced concrete
s = spacing of transverse reinforcement in the column hinge region
FEMA P695 Symbols G-5
S = snow load, Chapter 2 of ASCE/SEI 7-05
S = space frame system (Chapter 9)
S1 = mapped MCE, 5-percent damped, spectral response acceleration
parameter at a period of 1 second as defined in Section 11.4.1 of
ASCE/SEI 7-05
SCT = random variable representing collapse level earthquake, 5-percent
damped, spectral response acceleration at the fundamental period,
T, of the building (Site Class D)
SCT(SC) = collapse level earthquake, 5-percent damped, spectral response
acceleration at the fundamental period, T, of the building (Site
Class D) obtained from simulated collapse failure modes.
SCT(NSC) = collapse level earthquake, 5-percent damped, spectral response
acceleration at the fundamental period, T, of the building (Site
Class D) obtained from non-simulated collapse failure modes.
ˆ
CT S = median value of collapse level earthquake, 5-percent damped,
spectral response acceleration at the fundamental period, T, of the
building (Site Class D)
SCT1 = collapse level earthquake, 5-percent damped, spectral response
acceleration at the fundamental period, T1, of the building (Site
Class D)
SDS = design, 5-percent damped, spectral response acceleration
parameter at short periods as defined in Section 11.4.4 of
ASCE/SEI 7-05
SD1 = design, 5-percent damped, spectral response acceleration
parameter at a period of 1 second as defined in Section 11.4.4 of
ASCE/SEI 7-05
Smax = maximum lateral force of the fully-yielded seismic force-resisting
system normalized by the effective seismic weight of the
building, W
SMS = the MCE, 5-percent damped, spectral response acceleration
parameter at short periods adjusted for site class effects as defined
in Section 11.4.3 of ASCE/SEI 7-05
SMT = MCE, 5-percent damped, spectral response acceleration at the
fundamental period, T, of the building, as defined in Section
11.4.3 of ASCE/SEI 7-05 (Site Class D)
G-6 Symbols FEMA P695
SM1 = the MCE, 5-percent damped, spectral response acceleration
parameter at a period of 1 second adjusted for site class effects as
defined in Section 11.4.3 of ASCE/SEI 7-05 (Site Class D)
sn = rebar buckling coefficient, (s/db)(fy/100)0.5, where fy is in MPa
(similar to a term proposed by Dhakal and Maekawa (2002))
NRT Sˆ
= median value of normalized record set, 5-percent damped,
spectral response acceleration at the fundamental period, T, of the
building
SS = mapped MCE, 5-percent damped, spectral response acceleration
parameter at short periods as defined in Section 11.4.1 of
ASCE/SEI 7-05
T S = median value of normalized record set, 5-percent damped,
spectral response acceleration at the fundamental period, T, for
record set scaled to an arbitrary intensity. This parameter is also
the anchor point for individual ground motion records of the
record set at this intensity
Sa = spectra acceleration, g
Sa(T) = the spectral acceleration at the period, T
SCWB = ratio of flexure strengths of column and beams framing into a
joint, computed as per ACI 318-05 (ACI 2005)
CT SD = median value of collapse level earthquake, 5-percent damped,
spectral response displacement at the fundamental period, T, of
the building corresponding to spectral response acceleration, CT Sˆ
SDMT = MCE, 5-percent damped, spectral response displacement at the
fundamental period, T, of the building, corresponding to spectral
response acceleration, SMT
SF = record (and component) scale factor required for collapse
evaluation of an individual building
SMF = special moment frame
SSF = spectral shape factor
SSFi = spectral shape factor of ith index archetype analysis model
T = the fundamental period of the building, in seconds, based on the
limits of Section 12.8.2 of the ASCE/SEI 7-05 and the
approximate fundamental period, Ta
T1 = the fundamental period of the building, as determined by
eigenvalue analysis of the structural model (seconds)
FEMA P695 Symbols G-7
Ta = the approximate fundamental period of the building, in seconds,
as determined in Section 12.8.2.1 of ASCE/SEI 7-05
Teff = effective period of isolation system, in seconds, at displacement,
D, in the horizontal direction under consideration.
TH1,i = record i, horizontal component 1, PEER database
TH2,i = record i, horizontal component 2, PEER database
TL = long-period transition period of the building, in seconds, as
defined in Section 11.4.5 of ASCE/SEI 7-05
TM = effective period, in seconds, of the seismically isolated structure
at the maximum displacement in the direction under
consideration, as prescribed by Eq. 17.5-4 of ASCE/SEI 7-05
Ts = short-period transition period of the building, in seconds, equal to
SD1/SDS, as defined by Section 11.3 of ASCE/SEI 7-05
V = total design lateral force or shear at base (ASCE/SEI 7-05)
Vc = shear capacity of concrete, as per ACI 318-05
VE = lateral force that would be developed in the seismic forceresisting
system if the system remained entirely elastic for design
earthquake ground motions (FEMA, 2004b)
Vmax = maximum lateral force of the fully-yielded seismic force-resisting
system (same as VY parameter, FEMA, 2004b)
Vmpr = capacity shear demand, caused by fully plastic moments at each
end of the element, and including 25% overstrength, as per ACI
318-05 (ACI 2005)
Vn = nominal shear capacity including contributions from concrete and
steel, as per the ACI 318-05
Vs = total lateral seismic design force or shear on elements above the
isolation system, as prescribed by Eq. 17.5-8 of ASCE/SEI 7-05
Vu = shear demand
W = effective seismic weight of the structure above the isolation
interface, as defined in Section 17.5.3.4 of ASCE/SEI 7-05
W = effective seismic weight of the building as defined in Section
12.7.2 of ASCE/SEI 7-05
wx = portion of W that is located at or assigned to Level x
G-8 Symbols FEMA P695
x = parameter of Equation (12.8-7) given in Table 12.8-2 of
ASCE/SEI 7-05
.0
= regression coefficient from the equation LN[SCT1] = . 0 + . 1e
. 1 = regression coefficient from the equation LN[SCT1] = . 0 + . 1e
.DR = design requirements-related collapse uncertainty
.eff = effective damping of the isolation system, percent of critical
damping, at displacement, D, in the horizontal direction under
consideration..
.F = uncertainty associated with ductile fracture of steel reduced beam
sections, logarithmic standard deviation
.I
= component of effective damping of the structure due to the
inherent dissipation of energy by elements of the structure, at or
just below the effective yield displacement of the seismic forceresisting
system, Section 18.6.2.1 of ASCE/SEI 7-05
.M = effective damping, percent of critical, of the isolation system at
the maximum displacement in the direction under consideration,
as prescribed by Eq. 17.8-8 of ASCE/SEI 7-05
.MDL = modeling-related collapse uncertainty
.RTR = record-to-record collapse uncertainty
.TD = test data-related collapse uncertainty
.TOT = total system collapse uncertainty
.V1 = component of effective damping of the fundamental mode of
vibration of the structure in the direction of interest due to viscous
dissipation of energy by the damping system, at or just below the
effective yield displacement of the seismic force-resisting system,
Section 16.6.2.3 of ASCE/SEI 7-05
. = roof drift of the seismic force-resisting system for design
earthquake ground motions (FEMA, 2004b)
.c
= displacement associated with maximum strength, Fc (at the
capping point)
.E = roof drift of the seismic force-resisting system if the system
remained entirely elastic for design earthquake ground motions
(FEMA, 2004b)
FEMA P695 Symbols G-9
.u
= roof displacement used to approximate the ultimate displacement
capacity of the seismic force-resisting system, as derived from
pushover analysis
.p
= pre-capping plastic deformation capacity
.y,eff = effective roof displacement used to approximate full yield of the
seismic force-resisting system, as derived from pushover analysis
.. .. number of standard deviations between observed spectral value
and the median prediction from an attenuation function
..(T) = the epsilon value of a ground motion, evaluated at period, T
. (T) ,records = the mean epsilon value of the Far-Field ground motion set,
evaluated at period, T
0 . (T) = the expected e value for the site and hazard-level of interest
. = strength reduction factor
.1,x... ... ordinate of the fundamental mode in the direction of interest at
level x
.1,r.. ... ordinate of the fundamental mode in the direction of interest at the
roof.
. = normalized energy dissipation capacity; defined such that Et =
.My.y (Ibarra et al. 2005)
.DR = random variable representing design requirements-related
collapse uncertainty
.MDL = random variable representing modeling-related collapse
uncertainty
.RTR = random variable representing record-to-record collapse
uncertainty
.TD = random variable representing test data-related collapse uncertainty
.T = period-based ductility of an index archetype model
. = axial load ratio (P/Agf’c)
.cap = total chord rotation at capping, sum of elastic and plastic
deformations (radians)
.cap,pl (or .p
) = plastic chord rotation from yield to cap (radians)
.p
= plastic hinge rotation (radians)
.pc = post-capping plastic rotation capacity, from the cap to point of
zero strength (radians)
G-10 Symbols FEMA P695
p
ˆ.
= median value of plastic hinge rotation at which ductile fracture in
steel RBS is initiated (radians)
.y
= chord rotation at yielding, taken as the sum of flexural, shear and
bond-slip components; yielding is defined as the point of
significant stiffness change, i.e., steel yielding or concrete
crushing (radians)
... = redundancy factor based on the extent of structural redundancy
present in a building as defined in Section 12.3.4 of ASCE/SEI
7-05
. (or .tot) = ratio of total area of longitudinal reinforcement (for columns) or
ratio of tensile longitudinal reinforcement (for beams)
.’ = ratio compressive longitudinal reinforcement (for beams)
.sh = area ratio of transverse reinforcement in column hinge region
.sh,eff = effective ratio of transverse reinforcement in column hinge region
(.shfy,w/f’c..
.LN = logarithmic standard deviation for the prediction uncertainty
. = calculated overstrength of an index archetype analysis model
.O = overstrength factor appropriate for use in the load combinations of
Section 12.4 of ASCE/SEI 7-05 (current values of .O are given in
Table 12.2-1 of ASCE/SEI 7-05)
FEMA P695 Glossary H-1
Glossary
Definitions
Archetype: A prototypical representation of a seismic-force-resisting
system.
Archetype Design Space: The overall range of permissible configurations,
structural design parameters, and other features that define the
application limits for a seismic-force-resisting system.
Base: The level at which the horizontal seismic ground motions are
considered to be imparted to the structure (ASCE/SEI 7-05).
Base Shear: Total design lateral force or shear at the base (ASCE/SEI 7-05).
Building: Any structure whose intended use includes shelter of human
occupants (ASCE/SEI 7-05).
Collapse Level Earthquake Ground Motions: The level of earthquake
ground motions that cause collapse of the seismic force-resisting system
of interest.
Component: A part or element of an architectural, electrical, mechanical or
structural system (ASCE/SEI 7-05).
Damping Device: A flexible structural element of the damping system that
dissipates energy due to relative motion of each end of the device
(ASCE/SEI 7-05).
Damping System: The collection of structural elements that includes all
individual damping devices, all structural elements or bracing required to
transfer forces from damping devices to the base of the structure, and the
structural elements required to transfer forces from damping devices to
the seismic force-resisting system (ASCE/SEI 7-05).
Design Earthquake Ground Motions: The earthquake ground motions that
are two-thirds of the corresponding MCE ground motions (ASCE/SEI 7-
05).
Design Requirements-Related Uncertainty: Collapse uncertainty
associated with the quality of the design requirements of the system of
interest.
H-2 Glossary FEMA P695
Displacement Restraint System: A collection of structural elements that
limits lateral displacement of seismically isolated structures due to the
maximum considered earthquake (Chapter 17, ASCE/SEI 7-05).
Effective Damping: The value of equivalent viscous damping corresponding
to energy dissipated during cyclic response of the isolation system
(Chapter 17, ASCE/SEI 7-05).
Effective Stiffness: The value of the lateral force in the isolation system, or
element thereof, divided by the corresponding lateral displacement
(Chapter 17, ASCE/SEI 7-05).
Importance Factor: A factor assigned to each structure according to its
Occupancy Category, as prescribed in Section 11.5.1 of ASCE/SEI 7-05.
Incremental Dynamic Analysis: Series of nonlinear response history
analyses using an input ground motion that is incrementally scaled to
increasing intensities until collapse is detected in the analysis.
Index Archetype Configuration: A prototypical representation of a
seismic-force-resisting system configuration that embodies key features
and behaviors related to collapse performance when subjected to
earthquake ground motions.
Index Archetype Design: An index archetype configuration that has been
proportioned and detailed using the design requirements of the system of
interest.
Index Archetype Model: An idealized mathematical representation of an
index archetype design used to simulate collapse using nonlinear static
and dynamic analyses.
Isolation Interface: The boundary between the upper portion of the
structure, which is isolated, and the lower portion of the structure, which
moves rigidly with the ground (Chapter 17, ASCE/SEI 7-05).
Isolation System: The collection of structural elements that includes all
individual isolator units, all structural elements that transfer force
between elements of the isolation system, and all connections to other
structural elements. The isolation system also includes the windrestraint
system, energy-dissipation devices, and/or the displacement
restraint system, if such systems and devices are used to meet the design
requirements of Chapter 17 (ASCE/SEI 7-05).
Isolator Unit: A horizontally flexible and vertically stiff structural element
of the isolation system that permits large lateral deformations under
design seismic loads. An isolator unit is permitted to be used either as
FEMA P695 Glossary H-3
part of, or in addition to, the weight-supporting system of the structure
(Chapter 17, ASCE/SEI 7-05).
Maximum Considered Earthquake (MCE) Ground Motions: The most
severe earthquake effects considered, as defined by Section 11.4 of
ASCE/SEI 7-05.
Modeling Uncertainty: Collapse uncertainty associated with the quality of
the index archetype models.
Nonbuilding Structure: A structure, other than a building, constructed of a
type included in Chapter 15 of ASCE/SEI 7-05.
Non-Simulated Collapse: Structural collapse caused by collapse modes that
are not represented in the analytical model. Non-simulated collapse
occurs when a component limit state is exceeded, as defined by
component fragility functions.
Occupancy: The purpose for which a building or other structure, or part
thereof, is used or intended to be used (ASCE/SEI 7-05).
Occupancy Category: A classification assigned to a structure based on
occupancy as defined in Table 1-1 of ASCE/SEI 7-05.
Performance Group: A subset of the archetype design space containing a
group of index archetype configurations that share a set of common
features or behavioral characteristics, binned for statistical evaluation of
collapse performance.
Record-to-Record Uncertainty: Collapse uncertainty due to variability in
response to different ground motions.
Record-to-Record Variability: Variation in the response of a structure
under multiple input ground motions that are scaled to a consistent
ground motion intensity.
Seismic Design Category: A classification assigned to a structure based on
Occupancy Category and the severity of design earthquake ground
motions at the site, as defined in Section 11.4 of ASCE/SEI 7-05.
Seismic Force-Resisting System: That part of the structural system that is
considered to provide the required resistance to seismic forces prescribed
in ASCE/SEI 7-05.
Sidesway Collapse: Structural collapse due to excessive story drift
associated with loss of lateral strength and stiffness due to material and
geometric nonlinearities.
H-4 Glossary FEMA P695
Simulated Collapse: Structural collapse caused by collapse modes that are
directly represented in the analytical model.
Site Class: A classification assigned to a site based on the types of soils
present and their engineering properties, as defined in Chapter 20 of
ASCE/SEI 7-05.
Structure: That which is built or constructed and limited to buildings and
nonbuilding structures, as defined in ASCE/SEI 7-05.
Test Data-Related Uncertainty: Collapse uncertainty associated with the
quality of the test data for the system of interest.
Total Maximum Displacement: The maximum considered earthquake
lateral displacement, including additional displacement due to actual and
accidental torsion, required for verification of the stability of the
isolation system or elements thereof, design of structure separations, and
vertical load testing of isolator unit prototypes (Chapter 17, ASCE/SEI 7-
05).
Vertical Collapse: Structural collapse due to the loss of vertical-loadcarrying
capacity of a critical component.
FEMA P695 References I-1
References
Abrahamson, N.A., and Silva, W.J., 1997, “Empirical spectral response
attenuation relations for shallow crustal earthquakes,” Seismological
Research Letters, 68 (1), pp. 94-126.
ACI, 2005, Building Code Requirements for Structural Concrete (ACI 318-
05) and Commentary, (ACI 318R-05), American Concrete Institute,
Farmington Hills, Michigan.
ACI, 2002b, Building Code Requirements for Masonry Structures (ACI
530/ASCE 5/TMS 402), Masonry Standards Joint Committee of the
American Concrete Institute, Farmington Hills, Michigan; Structural
Engineering Institute of the American Society of Civil Engineers,
Reston, Virginia; and The Masonry Society, Boulder, Colorado.
ACI, 2002a, Building Code Requirements for Structural Concrete (ACI 318-
02) and Commentary (ACI 318R-02), American Concrete Institute,
Farmington Hills, Michigan.
ACI, 2001, Acceptance Criteria for Moment Frames Based on Structural
Testing (ACI T1.1-01) and Commentary (ACI T1.1R-01), Innovation
Task Group 1 and collaborators, American Concrete Institute,
Farmington Hills, Michigan.
AISC, 2005, Seismic Provisions for Structural Steel Buildings, ANSI/AISC
341-05, American Institute for Steel Construction, Chicago, Illinois.
Altoontash, A., 2004, Simulation and Damage Models for Performance
Assessment of Reinforced Concrete Beam-Column Joints, Ph.D.
Dissertation, Department of Civil and Environmental Engineering,
Stanford University, Stanford, California.
ANSI/AF&PA, 2005, National Design Specification for Wood Construction
(ANSI/AF&PA NDS-2005), American National Standards Institute
and American Forest and Paper Association, Washington, D.C.
ASCE, 2006b, Seismic Rehabilitation of Existing Buildings, ASCE Standard
ASCE/SEI 41-06, American Society of Civil Engineers, Reston,
Virginia.
I-2 References FEMA P695
ASCE, 2006a, Minimum Design Loads for Buildings and Other Structures,
ASCE Standard ASCE/SEI 7-05, including Supplement No. 1,
American Society of Civil Engineers, Reston, Virginia.
ASCE, 2005, Minimum Design Loads for Buildings and Other Structures,
ASCE 7-05, American Society of Civil Engineers, Reston, Virginia.
ASCE, 2003, Minimum Design Loads for Buildings and Other Structures.
ASCE Standard ASCE 7-02, American Society of Civil Engineers,
Washington, D.C.
ASCE, 2002, Minimum Design Loads for Buildings and Other Structures,
ASCE 7-02, American Society of Civil Engineers, Reston, Virginia.
ASTM, 2003, Standard Test Method for Cyclic (Reversed) Load Test for
Shear Resistance of Walls for Buildings, ASTM ES 2126-02a,
American Society for Testing and Materials, West Conshohocken,
Pennsylvania.
ATC, 2009, Guidelines for Seismic Performance Assessment of Buildings -
50% Draft, Report No. ATC-58, prepared by the Applied
Technology Council for the Federal Emergency Management
Agency, Washington, D.C.
ATC, 2008, Interim Guidelines on Modeling and Acceptance Criteria for
Seismic Design and Analysis of Tall Buildings, 90% Draft, Report
No. ATC-72-1, Applied Technology Council, Redwood City,
California.
ATC, 1992, Guidelines for Cyclic Seismic Testing of Components of Steel
Structures, Report No. ATC-24, Applied Technology Council,
Redwood City, California.
ATC, 1978, Tentative Provisions for the Development of Seismic Regulations
for Buildings, Report No. ATC 3-06, Applied Technology Council,
Redwood City, California; also NSF Publication 78-8 and NBS
Special Publication 510.
Aslani, H., 2005, Probabilistic Earthquake Loss Estimation and Loss
Deaggregation in Buildings, Ph.D. Dissertation, Department of Civil
and Environmental Engineering, Stanford University, Stanford,
California.
Aslani, H., and Miranda, E., 2005, “Fragility assessment of slab-column
connections in existing non-ductile reinforced concrete buildings,”
Journal of Earthquake Engineering, 9 (6), pp. 777-804.
FEMA P695 References I-3
Aycardi, L.E., Mander, J., and Reinhorn, A., 1994, “Seismic resistance of
reinforced concrete frame structures designed only for gravity loads:
experimental performance of subassemblages.” ACI Structural
Journal, 91 (5), pp. 552-563.
Baker, J.W., 2007, “Quantitative classification of near-fault ground motions
using wavelet analysis,” Bulletin of the Seismological Society of
America, 97 (5), pp. 1486-1501.
Baker, J.W., 2005, Vector-Valued Ground Motion Intensity Measures for
Probabilistic Seismic Demand Analysis, Ph.D. Dissertation,
Department of Civil and Environmental Engineering, Stanford
University, Stanford, California.
Baker, J.W. and Cornell, C.A., 2006, “Spectral shape, epsilon and record
selection,” Earthquake Engineering and Structural Dynamics, 34
(10), pp. 1193-1217.
Benjamin, J.R., and Cornell, C.A, 1970, Probability, Statistics, and Decision
for civil engineers, McGraw-Hill, New York, New York, 684 pp.
Berry, M., Parrish, M., and Eberhard, M., 2004, PEER Structural
Performance Database User’s Manual, Pacific Earthquake
Engineering Research Center, University of California, Berkeley,
California, 38 pp. Available at http://nisee.berkeley.edu/spd/ and
http://maximus.ce.washington.edu/~peera1/.
Boore, D.M., Joyner, W.B., and Fumal, T.E., 1997, “Equations for
estimating horizontal response spectra and peak accelerations from
western North America earthquakes: a summary of recent work,”
Seismological Research Letters, 68 (1), pp. 128-153.
Brown, P.C., and Lowes, L.N., 2006, “Fragility functions for modern
reinforced concrete beam-column joints,” Earthquake Spectra, 23
(2), pp. 263-289.
Campbell, K.W., and Borzorgnia, Y., 2003, “Updated near-source groundmotion
(attenuation) relations for the horizontal and vertical
components of peak ground acceleration and acceleration response
spectra,” Bulletin of the Seismological Society of America, Vol. 93,
No. 1, pp. 314-331.
Chatterjee, S., Hadi, A.S., and Price, B., 2000, Regression Analysis by
Example, Third Edition, John Wiley and Sons Inc., New York,
ISBN: 0-471-31946.
I-4 References FEMA P695
Chopra, A.K., Goel, R.K., and De la Llera, J.C., 1998, “Seismic code
improvements based on recorded motions of buildings during
earthquakes,” SMIP98 Seminar Proceedings, California Geological
Survey, Sacramento, California.
Clark, P., Frank, K., Krawinkler, H., and Shaw, R., 1997, Protocol for
Fabrication, Inspection, Testing, and Documentation of Beam-
Column Connection Tests and Other Experimental Specimens, SAC
Steel Project Background Document, Report No. SAC/BD-97/02.
CoLA, 2001, Report of a Testing Program of Light-Framed Walls with
Wood-Sheathed Panels, Final Report to the City of Los Angeles
Department of Building and Safety, by Structural Engineers
Association of Southern California and COLA-UCI Light Frame
Test Committee, Department of Civil Engineering, University of
California, Irvine, California, 93 pp.
Dhakal, R.P., and Maekawa, K., 2002, “Modeling of postyielding buckling
of reinforcement,” Journal of Structural Engineering, Vol. 128, No.
9, pp. 1139-1147.
Ekiert, C., and Hong, J., 2006, Framing-to-Sheathing Connection Tests in
Support of NEESWood Project, Network of Earthquake Engineering
Simulation; host institution: University at Buffalo, Buffalo, New
York, 20 pp.
Elghadamsi, F.E., and Mohraz, B., 1987, “Inelastic earthquake spectra,”
Earthquake Engineering and Structural Dynamics, Vol. 15, pp. 91-
104.
Ellingwood, B., Galambos, T.V., MacGregor, J.G., and Cornell, C.A., 1980,
Development of a Probability-Based Load Criterion for American
National Standard A58, National Bureau of Standards, Washington,
DC, 222 pp.
Elwood, K.J., 2004, “Modeling failures in existing reinforced concrete
columns,” Canadian Journal of Civil Engineering, 31 (5), pp. 846-
859.
Elwood, K.J., and Eberhard, M.O., 2006, “Effective stiffness of reinforced
concrete columns,” PEER Research Digest 2006-1, pp. 1- 5.
Elwood, K. J., and Moehle, J., 2005, “Axial capacity model for sheardamaged
columns,” ACI Structural Journal, 102 (4).
FEMA P695 References I-5
Engelhardt, M., Winneberger, T., Zekany, A.J., and Potyraj, T.J., 1998,
“Experimental investigation of dogbone moment connections,”
Engineering Journal of Steel Construction, Vol. 4, pp. 128-139.
Fardis, M.N., and Biskini, D.E., 2003, “Deformation capacity of RC
members, as controlled by flexure or shear,” Otani Symposium,
University of Tokyo, Japan, pp. 511- 530.
FEMA, 2009, Effects of Strength and Stiffness Degradation on Seismic
Response, FEMA P440A, prepared by the Applied Technology
Council for the Federal Emergency Management Agency,
Washington, D.C.
FEMA, 2007, Interim Testing Protocols for Determining the Seismic
Performance Characteristics of Structural and Nonstructural
Components, FEMA 461, prepared by the Applied Technology
Council for the Federal Emergency Management Agency,
Washington, D.C.
FEMA, 2005, Improvement of Nonlinear Static Seismic Analysis Procedures,
FEMA 440, Federal Emergency Management Agency, Washington,
D.C.
FEMA, 2004b, NEHRP Recommended Provisions for Seismic Regulations
for New Buildings and Other Structures, FEMA 450-2/2003 Edition,
Part 2: Commentary, Federal Emergency Management Agency,
Washington, D.C.
FEMA, 2004a, NEHRP Recommended Provisions for Seismic Regulations
for New Buildings and Other Structures, FEMA 450-1/2003 Edition,
Part 1: Provisions, Federal Emergency Management Agency,
Washington, D.C.
FEMA, 2001, NEHRP Recommended Provisions for Seismic Regulations for
New Buildings and Other Structures, FEMA 369, Federal
Emergency Management Agency, Washington, D.C.
FEMA, 2000, Prestandard and Commentary for Seismic Rehabilitation of
Buildings, FEMA 356, Prepared by the American Society of Civil
Engineers for the Federal Emergency Management Agency,
Washington, D.C.
Filiatrault, A., Wanitkorkul, A., and Constantinou, M.C., 2008, Development
and Appraisal of a Numerical Cyclic Loading Protocol for
Quantifying Building System Performance, Technical Report
MCEER-08-0013, MCEER, University at Buffalo, Buffalo, New
York.
I-6 References FEMA P695
Filiatrault, A., Lachapelle, E., and Lamontagne, P., 1998, “Seismic
performance of ductile and nominally ductile reinforced concrete
moment resisting frames I: experimental study,” Canadian Journal
of Civil Engineering, 25 (2), pp. 342-358.
Filippou, F.C., 1999, “Analysis platform and member models for
performance-based earthquake engineering,” U.S.-Japan Workshop
on Performance-Based Earthquake Engineering Methodology for
Reinforced Concrete Building Structures, PEER Report 1999/10,
Pacific Earthquake Engineering Research Center, University of
California, Berkeley, California, pp. 95-106.
Folz, B., and Filiatrault, A., 2004b, “Seismic analysis of woodframe
structures II: model implementation and verification,” ASCE Journal
of Structural Engineering, Vol. 130, No. 8, pp. 1361-1370.
Folz, B., and Filiatrault, A., 2004a, “Seismic analysis of woodframe
structures I: model formulation,” ASCE Journal of Structural
Engineering, Vol. 130, No. 8, pp. 1353-1360.
Folz, B., and Filiatrault, A., 2001, “Cyclic analysis of wood shear walls,”
ASCE Journal of Structural Engineering, 127 (4), pp. 433-441.
Fonseca, F., Rose, S., and Campbell, S., 2002, Nail, Wood Screw, and Staple
Fastener Connections, CUREE Report No. W-16, Consortium of
Universities for Earthquake Engineering Research, Richmond,
California.
Gatto, K., and Uang, C.M., 2002, Cyclic Response of Woodframe
Shearwalls: Loading Protocol and Rate of Loading Effects, CUREE
Report No. W-16, Consortium of Universities for Earthquake
Engineering Research, Richmond, California.
Goulet, C., Haselton, C.B., Mitrani-Reiser, J., Beck, J., Deierlein, G.G.,
Porter, K.A., and Stewart, J., 2006, “Evaluation of the seismic
performance of a code-conforming reinforced-concrete frame
building - from seismic hazard to collapse safety and economic
losses,” Earthquake Engineering and Structural Dynamics, 36 (13),
pp. 1973-1997.
Goulet, C., Haselton, C.B., Mitrani-Reiser, J., Stewart, J., Taciroglu, E., and
Deierlein, G.G., 2006, “Evaluation of seismic performance of a codeconforming
reinforced-concrete frame buildings - part I, ground
motion selection and structural collapse simulation,” Proceedings,
8th National Conference on Earthquake Engineering, San Francisco,
California.
FEMA P695 References I-7
Harmsen, S.C., 2001, “Mean and Modal e in the Deaggregation of
Probabilistic Ground Motion,” Bulletin of the Seismological Society
of America, 91 (6), pp. 1537-1552.
Harmsen, S.C., Frankel, A. D., and Petersen, M. D., 2003, “Deaggregation of
U.S. Seismic Hazard Sources: The 2002 Update,” U.S. Geological
Survey Open-File Report 03-440, Available at
http://pubs.usgs.gov/of/2003/ofr-03-440/.
Haselton, C.B., 2006, Assessing Seismic Collapse Safety of Modern
Reinforced Concrete Moment-Frame Buildings, Ph.D. Dissertation,
Department of Civil and Environmental Engineering, Stanford
University, Stanford, California.
Haselton, C.B., Baker, J.W., Liel, A.B. and Deierlein, G.G., 2009,
“Accounting for expected spectral shape (epsilon) in collapse
performance assessment,” American Society of Civil Engineers
Journal of Structural Engineering, Special Publication on Ground
Motion Selection and Modification (submitted).
Haselton, C.B., Mitrani-Reiser, J., Goulet, C., Deierlein, G.G., Beck, J.,
Porter, K.A., Stewart, J., and Taciroglu, E., 2008, An Assessment to
Benchmark the Seismic Performance of a Code-Conforming
Reinforced-Concrete Moment-Frame Building, PEER Report
2007/12, Pacific Earthquake Engineering Research Center,
University of California, Berkeley, California.
Haselton, C.B., and Deierlein, G.G., 2007, Assessing Seismic Collapse Safety
of Modern Reinforced Concrete Frame Buildings, John A. Blume
Earthquake Engineering Center Technical Report No. 156, Stanford
University, Stanford, California.
Haselton, C.B., Liel, A., Taylor Lange, S., and Deierlein, G.G., 2007, Beam-
Column Element Model Calibrated for Predicting Flexural Response
Leading to Global Collapse of RC Frame Buildings, PEER 2007/03,
Pacific Earthquake Engineering Research Center, University of
California, Berkeley, California.
Haselton, C.B., and Baker, J.W., 2006, “Ground motion intensity measures
for collapse capacity prediction: Choice of optimal spectral period
and effect of spectral shape,” Proceedings, 8th National Conference
on Earthquake Engineering, San Francisco, California.
Ibarra, L.F., and Krawinkler, H., 2005b, Global Collapse of Frame
Structures Under Seismic Excitations, PEER Report 2005/06, and
John A. Blume Earthquake Engineering Center Technical Report No.
I-8 References FEMA P695
152, Department of Civil Engineering, Stanford University,
Stanford, California.
Ibarra, L., and Krawinkler, H., 2005a, “Effect of uncertainty in system
deterioration parameters on the variance of collapse capacity,”
Proceedings, ICOSSAR'05, Rome, Italy, pp. 3583-3590, Millpress,
Rotterdam, ISBN 90 5966 040 4.
Ibarra, L.F., Medina, R.A., and Krawinkler, H., 2005, “Hysteretic models
that incorporate strength and stiffness deterioration,” International
Journal for Earthquake Engineering and Structural Dynamics, Vol.
34, No.12, pp. 1489-1511.
Ibarra, L., Medina, R., and Krawinkler, H., 2002, “Collapse assessment of
deteriorating SDOF systems,” Proceedings, 12th European
Conference on Earthquake Engineering, London, Elsevier Science
Ltd, paper #665.
ICBO, 1997, Uniform Building Code, 1997 Edition, International Conference
of Building Officials, Whittier, California.
ICC, 2009, Acceptance Criteria for Prefabricated Wood Shear Panels,
AC130, International Code Council, Available at www.icc-es.org.
ISO/IEC, 2005, General Requirements for the Competence of Testing and
Calibration Laboratories, ISO/IEC 17025, International
Organization for Standardization, Geneva, Switzerland.
Isoda, H., Furuya, O., Tatsuya, M., Hirano, S. and Minowa, C., 2007,
“Collapse behavior of wood house designed by minimum
requirement in law,” Journal of Japan Association for Earthquake
Engineering (under review).
Isoda, H., Folz, B., and Filiatrault, A., 2001, Seismic Modeling of Index
Woodframe Buildings, CUREE Report No. W-12, Consortium of
Universities for Research in Earthquake Engineering, Richmond,
California, 144 pp.
Kircher, C.A., 2006, “Seismically isolated structures,” NEHRP
Recommended Provisions: Design Examples, Chapter 11 of FEMA
451, prepared by the Building Seismic Safety Council for the Federal
Emergency Management Agency, Washington, D.C.
Krawinkler, H., 1996, “Cyclic loading histories for seismic experimentation
on structural components,” Earthquake Spectra, Vol. 12, Number 1,
pp. 1-12.
FEMA P695 References I-9
Krawinkler, H., 1978, “Shear design of steel frame joints,” Engineering
Journal, AISC, Vol. 15, No. 3.
Krawinkler, H., and Zareian, F., 2007, “Prediction of collapse – how realistic
and practical is it, and what can we learn from it?,” The Structural
Design of Tall and Special Buildings, Vol. 16, No. 5, pp. 633-653.
Krawinkler, H., Parisi, F., Ibarra, L., Ayoub, A., and Medina, R., 2000,
Development of a Testing Protocol for Woodframe Structures,
CUREE Publication No. W-02, Consortium of Universities for
Earthquake Engineering Research, Richmond, California, 74 pp.
Krawinkler, H., Bertero, V.V., and Popov, E.P., 1971, Inelastic behavior of
steel beam-to-column sub-assemblages, Report No. UCB/EERC-
71/07, Earthquake Engineering Research Center, University of
California, Berkeley, California.
Kunnath, S.K., Hoffmann, G., Reinhorn, A.M., and Mander, J.B., 1995,
“Gravity-load-designed reinforced concrete buildings -- part I:
seismic evaluation of existing construction,” ACI Structural Journal,
92 (3).
Lai, S.P., and Biggs, J.M., 1980, “Inelastic response spectra for aseismic
buildings,” Journal of the Structural Division, ASCE, Vol. 106, No.
ST6, pp. 1295-1310.
Liel, A. B., 2008, Assessing the Collapse Risk of California’s Existing
Reinforced Concrete Frame Structure: Metrics for Seismic Safety
Decisions, Ph.D. Dissertation, Department of Civil and
Environmental Engineering, Stanford University, Stanford,
California.
Lignos, D.G., and Krawinkler, H., 2007, “A database in support of modeling
of component deterioration for collapse prediction of steel frame
structures,” ASCE Structures Congress, Long Beach, California.
Line, P., Waltz, N., and Skaggs, T., 2008, “Seismic equivalence parameters
for engineered wood frame wood structural panel shear walls,” Wood
Design Focus, 18 (2).
Lowes, L.N., Mitra, N., and Altoontash, A., 2004, A Beam-Column Joint
Model for Simulating the Earthquake Response of Reinforced
Concrete Frames, PEER 2003/10, Pacific Earthquake Engineering
Research Center, University of California, Berkeley, California.
I-10 References FEMA P695
Lowes, L.N., and Altoontash, A., 2003, “Modeling of reinforced-concrete
beam-column joints subjected to cyclic loading,” Journal of
Structural Engineering, 129 (12).
Miranda, E. and Bertero, V.V., 1994, “Evaluation of strength reduction
factors for earthquake-resistant design,” Earthquake Spectra, Vol.
10, No. 2, pp. 357-379.
Nassar, A.A. and Krawinkler, H., 1991, Seismic Demands for SDOF and
MDOF Systems, The John A. Blume Earthquake Engineering Center
Technical Report No. 95, Stanford University, Stanford, California.
Newmark, N.M. and Hall, W.J., 1973, “Seismic design criteria for nuclear
reactor facilities,” Building Practices for Disaster Mitigation, Report
No. 46, National Bureau of Standards, U.S. Department of
Commerce, pp. 209-236.
OpenSees, 2006, Open System for Earthquake Engineering Simulation,
Pacific Earthquake Engineering Research Center, University of
California, Berkeley, Available at http://opensees.berkeley.edu/.
Panagiotakos, T.B., and Fardis, M.N., 2001, “Deformations of reinforced
concrete members at yielding and ultimate,” ACI Structural Journal,
98 (2).
PEER, 2006a, PEER NGA Database, Pacific Earthquake Engineering
Research Center, University of California, Berkeley, California,
Available at http://peer.berkeley.edu/nga/.
PEER, 2006b, PEER Structural Performance Database, Pacific Earthquake
Engineering Research Center, University of California, Berkeley,
Available at http://nisee.berkeley.edu/spd/ and
http://maximus.ce.washington.edu/~peera1/.
Ricles, J.M., Zhang, X., Lu, L.W., Fisher, J., 2004, Development of seismic
guidelines for deep-column steel moment connections, ATLSS
Report No. 04-13, Center for Advanced Technology for Large
Structural Systems, Lehigh University, Bethlehem, Pennsylvania.
Riddell, R., Hidalgo, P., and Cruz, E., 1989, “Response modification factors
for earthquake-resistant design of short-period structures,”
Earthquake Spectra, Vol. 5, No. 3, pp. 571-590.
Sivaselvan M., and Reinhorn, A.M., 2000, “Hysteretic models for
deteriorating inelastic structures,” ASCE Journal of Engineering
Mechanics, Vol. 126 (6), pp. 633-640.
FEMA P695 References I-11
Stevens, N., Uzumeri, S., and Collins, M., 1991, “Reinforced concrete
subjected to reversed cyclic shear,” ACI Structural Journal, 88 (2),
pp. 135-146.
Takeda, T., Hwang, H.H.M., and Shinozuka, M., 1988, “Response
modification factors for multiple-degree-of-freedom systems,”
Proceedings, 9th World Conference on Earthquake Engineering,
Tokyo-Kyoto, Japan, Vol. V, pp. 129-134.
Uang, C.M., Yu, K., and Gilton, C., 2000, Cyclic Response of RBS Moment
Connections: Loading Sequence and Lateral Bracing Effects,
Structural Steel Research Project Report No. SSRP-99/13.
Umemura, H., and Aoyama, H., 1969, “Empirical evaluation of the behavior
of structural elements,” Proceedings, 4th World Conference on
Earthquake Engineering, Santiago, Chile.
Vamvatsikos, D., and Cornell, C.A., 2002, “Incremental Dynamic Analysis,”
Earthquake Engineering and Structural Dynamics, Vol. 31, Issue 3,
pp. 491-514.
Vecchio, F.J., and Collins, M.P., 1986. “Modified compression-field theory
for reinforced concrete elements subjected to shear,” ACI Journal, 83
(2), pp. 219-231.
Vidic, T., Fajfar, P., and Fischinger, M., 1992, “A procedure for determining
consistent inelastic design spectra,” Proceedings, Workshop on
Nonlinear Seismic Analysis of RC Structures, Bled, Slovenia.
White, T.W. and Ventura, C., 2007, “Seismic behavior of residential woodframe
construction in British Columbia: Part I – modeling and
validation,” Proceedings, Ninth Canadian Conference on Earthquake
Engineering, Ottawa, Canada, 10 pp.
Zareian, F., 2006, Simplified Performance-Based Earthquake Engineering,
Ph.D. Dissertation, Dept. of Civil and Environmental Engineering,
Stanford University, Stanford, California.
FEMA P695 Project Participants J-1
Project Participants
ATC Management and Oversight
Christopher Rojahn (Project Executive)
Applied Technology Council
201 Redwood Shores Parkway, Suite 240
Redwood City, California 94065
William T. Holmes (Project Technical Monitor)
Rutherford & Chekene
55 Second Street, Suite 600
San Francisco, California 94105
Jon A. Heintz (Project Manager)
Applied Technology Council
201 Redwood Shores Parkway, Suite 240
Redwood City, California 94065
FEMA Project Officer
Michael Mahoney
Federal Emergency Management Agency
500 C Street, SW
Washington, DC 20472
FEMA Technical Monitor
Robert D. Hanson
Federal Emergency Management Agency
2926 Saklan Indian Drive
Walnut Creek, California 94595
Project Management Committee
Charles Kircher (Project Technical Director)
Kircher & Associates, Consulting Engineers
1121 San Antonio Road, Suite D-202
Palo Alto, California 94303
Michael Constantinou
University at Buffalo
Dept. of Civil, Structural & Environ. Engineering
132 Ketter Hall
Buffalo, New York 14260
Gregory Deierlein
Stanford University
Dept. of Civil & Environmental Engineering
240 Terman Engineering Center
Stanford, California 94305
James R. Harris
J.R. Harris & Company
1776 Lincoln Street, Suite 1100
Denver, Colorado 80203
John Hooper
Magnesson Klemencic Associates
1301 Fifth Avenue, Suite 3200
Seattle, Washington 98101
Allan R. Porush
URS Corporation
915 Wilshire Blvd., Suite 700
Los Angeles, California 90017
Christopher Rojahn (ex-officio)
William Holmes (ex-officio)
Jon A. Heintz (ex- officio)
J-2 Project Participants FEMA P695
Working Group on Nonlinear Static Analysis
Michael Constantinou, Group Leader
University at Buffalo
Dept. of Civil, Structural & Environ. Engineering
132 Ketter Hall
Buffalo, New York 14260
Assawin Wanitkorkul
The University at Buffalo Foundation
Suite 211, The UB Commons
520 Lee Entrance
Amherst, New York 14228
Working Group on Nonlinear Dynamic Analysis
Gregory Deierlein, Group Leader
Stanford University
Dept. of Civil & Environmental Engineering
Blume Earthquake Engineering Center
Stanford, California 94305
Curt Haselton
California State University, Chico
475 East 10th Avenue
Chico, California 95926
Abbie Liel
University of Colorado at Boulder
1793 Yellow Pine Avenue
Boulder, Colorado 80304
Jason Chou
University of California, Davis
11713 New Albion Drive
Gold River, California 95670
Stephen Cranford
Stanford University
119 Quillen Court, Apt. 417
Stanford, California 94305
Brian Dean
Stanford University
159 Melville Ave.
Palo Alto, California 94301
Kevin Haas
Stanford University
218 Ayrshire Farm Lane #108
Stanford, California 94305
Jiro Takagi
Stanford University
322 College Ave., #C
Palo Alto, California 94306
Working Group on Wood-frame Construction
Andre Filiatrault, Group Leader
University at Buffalo
Dept. of Civil, Structural & Environ. Engineering
134 Ketter Hall
Buffalo, New York 14260
Jiannis Christovasilis
The University at Buffalo Foundation
Suite 211, The UB Commons
520 Lee Entrance
Amherst, New York 14228
Kelly Cobeen
Wiss Janney Elstner Associates, Inc.
2200 Powell St., Ste 925
Emeryville, California 94608
Working Group on Autoclaved Aerated Concrete
Helmut Krawinkler, Group Leader
Stanford University
Dept. of Civil & Environmental Engineering
380 Panama Mall
Stanford, California 94305
Farzin Zareian
University of California, Irvine
Dept. of Civil & Environmental Engineering
E/4141 Engineering Gateway
Irvine, California 92697
FEMA P695 Project Participants J-3
Project Review Panel
Maryann T. Phipps (Chair)
Estructure
8331 Kent Court, Suite 100
El Cerrito, California 94530
Amr Elnashai
Mid-America Earthquake Center
University of Illinois at Urbana-Champaign
Department of Civil and Environmental
Engineering 1241 Newmark
Civil Engineering Lab
Urbana, Illinois 61801
S.K. Ghosh
S.K. Ghosh Associates Inc.
334 East Colfax Street, Unit E
Palatine, Illinois 60067
Ramon Gilsanz
Gilsanz Murray Steficek LLP
129 W. 27th Street, 5th Floor
New York, New York 10001
Ronald O. Hamburger
Simpson Gumpertz & Heger
The Landmark @ One Market, Suite 600
San Francisco, California 94105
Jack Hayes
National Institute of Standards and Technology
100 Bureau Drive, MS 8610
Gaithersburg, Maryland 20899
Richard E. Klingner
University of Texas at Austin
10100 Burnet Rd
Austin, Texas 78758
Philip Line
American Forest and Paper Association (AFPA)
1111 19th Street, NW, Suite 800
Washington, DC 20036
Bonnie E. Manley, P.E.
AISI Regional Director
4 Canvasback Way
Walpole, Maryland 02081
Andrei M. Reinhorn
University at Buffalo
Dept. of Civil, Structural & Envir. Engineering
231 Ketter Hall
Buffalo, New York 14260
Rafael Sabelli
Walter P. Moore
595 Market Street, Suite 950
San Francisco, California 94105
Workshop Participants – Chicago, Illinois
John Abruzzo
Thornton Tomasetti
1617 JFK Boulevard, Suite 545
Philadelphia, Pennsylvania 19103
Victor Azzi
Consulting Structural Engineer
1100 Old Ocean Boulevard
Rye, New Hampshire 03870
William Baker
Skidmore Owings & Merrill LLP
224 South Michigan Avenue, Suite 1000
Chicago, Illinois 60604
Charles Carter
American Institute of Steel Construction
One East Wacker Drive Suite 700
Chicago, Illinois 60601-1802
Finley Charney
Virginia Polytechnic Institute
Dept. of Civil & Environmental Engineering
200 Patton Hall, MS 0105
Blacksburg, Virginia 24061
Peter Cheever
LeMessurier Consultants, Inc.
675 Massachusetts Avenue
Cambridge, Massachusetts 02139
J-4 Project Participants FEMA P695
Helen Chen
American Institute of Steel Construction
One East Wacker Drive Suite 700
Chicago, Illinois 60601-1802
Ned Cleland
Blue Ridge Design, Inc.
19 West Cork Street, Suite 300
Winchester, Virginia 22601-4749
Bruce Ellingwood
Georgia Institute of Technology
790 Atlantic Drive
School of Civil and Environmental Engineering
Atlanta, Georgia 30332-0355
Perry Green
Steel Joint Institute (SJI)
3127 Mr. Joe White Avenue
Myrtle Beach, South Carolina 29577-6760
Greg Greenlee
United Steel Products Company
2150 Kitty Hawk Road
Livermore, California 94551-9522
Kirk Grundahl
QUALTIM, Inc.
6300 Enterprise Lane
Madison, WI 53719
Kurt Gustafson
AISC
One East Wacker Drive Suite 700
Chicago, Illinois 60601-1802
Jerome Hajjar
University of Illinois
2129b Newmark Civil Engineering Lab.
05 N. Mathews Ave., MC 250
Urbana, Illinois 61801
Bob Hannen
Wiss, Janney, Elstner Associates
330 Pfingsten Road
Northbrook, Illinois 60062
Joe Jun
Allied Tube & Conduit
16100 South Lathrop Avenue
Harvey, Illinois 60426
Mike Kempfert
Computerized Structural Design
8989 N. Port Washington Rd. Suite 101
Milwaukee, Wisconsin 53217
Bonghwan Kim
Skidmore Owings & Merrill LLP
14 Wall Street
New York, New York 10005
Vladimir Kochkin
NAHB Research Center
400 Prince George's Blvd.
Upper Marlboro, Maryland 20774
Jay Larson
American Iron and Steel Institute
3810 Sydna Street
Bethlehem, Pennsylvania 18107-1048
Judy Liu
Purdue University
School of Civil Engineering
550 Stadium Mall Drive
West Lafayette, Idiana 47907
Douglas Nyman
DJ Nyman & Associates
12337 Jones Road, Suite 232
Houston, Texas 77070-4844
Marjan Popovski
FPInnovations
2665 East Mall
Vancouver, British Columbia C V6T 2G1 0
Randy Poston
Whitlock Dalrymple Poston & Associates
10621 Gateway Boulevard Suite 200
Manassas, Virginia 20110
Lawrence Reaveley
University of Utah
160 S. Cventral Camus Dr., Rm 104
Salt Lake City, Utah 84112-6931
Brett Schneider
Guy Nordenson & Associates
225 Varick, 6th Floor
New York, New York 10014
FEMA P695 Project Participants J-5
Lee Shoemaker
Metal Building Manufacturers Association
1300 Sumner Avenue
Cleveland, Ohio 44115-2851
Mike Tong
Federal Emergency Management Agency
500 C Street SW
Washington, DC 20472
Workshop Participants – San Francisco, California
Daniel Abrams
University of Illinois
205 N. Mathews Ave., MC 250
Urbana, Illinois 61801
Don Allen
Steel Framing Alliance (SFA)
1201 15th Street, N.W., Suite 320
Washington, DC 20005-2842
Robert Bachman
RE Bachman Consulting Structural Engrs.
25152 La Estrada Drive
Laguna Niguel, California 92677
Yousef Bozorgnia
PEER Center
University of California, Berkeley
325 Davis Hall - MC1792
Berkeley, California 94720
Edwin T. Dean
Nishkian Dean
425 SW Start Street, Second Floor
Portland, Oregon 97204
Ronald DeVall
Read Jones Christoffersen
201-628 W. 13th Avenue
Vancouver, British Columbia V6T 1N9 Canada
Ken Elwood
University of British Columbia
Dept. of Civil Engineering
6250 Applied Science Lane
Vancouver, British Columbia V6T 1Z4 Canada
Johanna Fenten
FEMA Region IX
1111 Broadway, Suite 1200
Oakland, California 94607-4052
David Garza
Garza Structural Engineers
1134 Shady Mill Road
Corona, California 92882
Avik Ghosh
Borm
5161 California Suite 250
Irvine, California 92617
David S. Gromala
Weyerhaeuser; WTC-2B2
P.O. Box 9777
Federal Way, Washington
Bernadette Hadnagy
Applied Technology Council
201 Redwood Shores Pkwy., Suite 240
Redwood City, California 94065
Gary Hart
Weidlinger Associates
4551 Glencoe Ave., Suite 350
Marina del Rey, California 90292
Ayse Hortacsu
Applied Technology Council
201 Redwood Shores Pkwy., Suite 240
Redwood City, California 94065
Ronald Klemencic
Magnusson Klemencic Associates
1301 Fifth Avenue, Suite 3200
Seattle, Washington 98101-2699
Peter L. Lee
Skidmore, Owings, & Merrill LLP
1 Front Street
San Francisco, California 94111
J-6 Project Participants FEMA P695
Roy Lobo
State of California Facilities Development
Division
1600 9th Street, Room 420
Sacramento, California 95814
Suikai Lu
Wienerberger AG
A-1100 Wien
Wienerberg City, Austria
Ray Lui
Department of Building Inspection
1660 Mission Street, Second Floor
San Francisco, California 94103
James A. Mahaney
Wiss, Janney, Elstner Associates, Inc.
104 El Dorado Street
Auburn, California 95603
Steven Mahin
University of California, Berkeley
Dept. of Civil Engineering
777 Davis Hall
Berkeley, California 94720
James O. Malley
Degenkolb Engineers
235 Montgomery Street, Suite 500
San Francisco, California 94104
Michael Mehrain
URS Corporation
915 Wilshire Blvd., Suite 700
Los Angeles, California 90017
Farzad Naeim
John A. Martin & Associates, Inc.
950 S. Grand Avenue, 4th Floor
Los Angeles, California 90015
Steven Pryor
Simpson Strong-Tie
5956 W. Las Positas Blvd.
Pleasanton, California 94588
Josh Richards
KPFF Consulting Engineers
111 SW Fifth Avenue, Suite 2500
Portland, Oregon 97204
Charles Roeder
Unversity of Washington
Dept. of Civil and Environmental Engineering
201 More Hall
Seattle, Washington 98195-2700
James E. Russell
Building Codes Consultant
3654 Sun View Way
Concord, California 94520-1346
Reynaud Serrette
Santa Clara University
Department of Civil Engineering
Santa Clara, California 96053-0563
Benson Shing
University of California, San Diego
409 University Center
9500 Gilman Drive, MC0085
La Jolla, California 92093
Thomas Skaggs
APA
7011 S. 19th Street
Tacoma, Washington 98466
William Staehlin
Division of the State Architect,
1102 Q Street, Suite 5200
Sacramento, California 95814
Kurt Stochlia
International Code Council
5360 Workman Mill Road
Whittier, California 90601-2298
Steve Tipping
Tipping Mar & Associates
1906 Shattuck Avenue
Berkeley, California 94704
Chia-Ming Uang
University of California, San Diego
Dept. of Structural Engineering
409 University Center
La Jolla, California 92093-0085
Williston L. Warren , IV
SESOL, Inc.
1001 West 17th St. Suite K
Costa Mesa, California 92627
FEMA P695 Project Participants J-7
Nabih Youssef
Nabih Youssef & Associates
800 Wilshire Blvd., Suite 200
Los Angeles, California 90017
Victor Zayas
Earthquake Protection Systems, Inc.
451 Azuar Drive, Building 759
Vallejo, California 94592