FEDERAL EMERGENCY MANAGEMENT AGENCY

 

September, 2000

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 


2 2 State of the Art Report on
0 Past Performance of Steel
2 Moment-Frame Buildings
0 in Earthquakes



 

 

 

Program to Reduce the Earthquake Hazards DISCLAIMER
This document provides practicing engineers and building officials with a resource document for
understanding the behavior of steel moment-frame buildings in earthquakes. It is one of the set of
six State of the Art Reports containing detailed derivations and explanations of the basis for the
design and evaluation recommendations prepared by the SAC Joint Venture. The recommendations
and state of the art reports, developed by practicing engineers and researchers, are based on
professional judgment and experience and supported by a large program of laboratory, field, and
analytical research. No warranty is offered with regard to the recommendations contained
herein, by the Federal Emergency Management Agency, the SAC Joint Venture, the
individual joint venture partners, or the partner’s directors, members or employees. These
organizations and their employees do not assume any legal liability or responsibility for the
accuracy, completeness, or usefulness of any of the information, products or processes
included in this publication. The reader is cautioned to review carefully the material
presented herein and exercise independent judgment as to its suitability for application to
specific engineering projects. This publication has been prepared by the SAC Joint Venture with
funding provided by the Federal Emergency Management Agency, under contract number EMW95-C-4770.

Cover Art. The beam-column connection assembly shown on the cover depicts the standard
detailing used in welded steel moment-frame construction prior to the 1994 Northridge
earthquake. This connection detail was routinely specified by designers in the period 1970-1994
and was prescribed by the Uniform Building Code for seismic applications during the period
1985-1994. It is no longer considered to be an acceptable design for seismic applications.
Following the Northridge earthquake, it was discovered that many of these beam-column
connections had experienced brittle fractures at the joints between the beam flanges and column
flanges.

 State of the Art Report on
Past Performance of Steel Moment-Frame Buildings in
Earthquakes
SAC Joint Venture
A partnership of
Structural Engineers Association of California (SEAOC)
Applied Technology Council (ATC)
California Universities for Research in Earthquake Engineering (CUREe)
Prepared for the SAC Joint Venture Partnership by
Evan Reis
Comartin-Reis

David Bonowitz
Project Oversight Committee
William J. Hall, Chair
James R. Harris
Richard Holguin
Nestor Iwankiw
Roy G. Johnston
Len Joseph

Shirin Ader
John M. Barsom
Roger Ferch
Theodore V. Galambos
John Gross

Duane K. Miller
John Theiss
John H. Wiggins

SAC Project Management Committee
SEAOC: William T. Holmes
ATC: Christoper Rojahn
CUREe: Robin Shepherd

Program Manager: Stephen A. Mahin
Project Director for Topical Investigations:
James O. Malley
Project Director for Product Development:
Ronald O. Hamburger

Topical Investigation Team
Peter Clark
Michael Durkin
James Goltz

Bruce Maison
Peter Maranian
Terrence Paret

Maryann Phipps
Allan Porush

Technical Advisory Panel
Jacques Cattan
Gary C. Hart
Y. Henry Huang
Helmut Krawinkler

Dennis Randall
Andrei Reinhorn
Arthur E. Ross

SAC Joint Venture
SEAOC: www.seaoc.org
ATC: www.atcouncil.org
CUREe: www.curee.org
September 2000

C. Mark Saunders
W. Lee Shoemaker
John Theiss

 THE SAC JOINT VENTURE
SAC is a joint venture of the Structural Engineers Association of California (SEAOC), the Applied
Technology Council (ATC), and California Universities for Research in Earthquake Engineering
(CUREe), formed specifically to address both immediate and long-term needs related to solving
performance problems with welded, steel moment-frame connections discovered following the 1994
Northridge earthquake. SEAOC is a professional organization composed of more than 3,000 practicing
structural engineers in California. The volunteer efforts of SEAOC’s members on various technical
committees have been instrumental in the development of the earthquake design provisions contained in
the Uniform Building Code and the 1997 National Earthquake Hazards Reduction Program (NEHRP)
Recommended Provisions for Seismic Regulations for New Buildings and other Structures. ATC is a
nonprofit corporation founded to develop structural engineering resources and applications to mitigate the
effects of natural and other hazards on the built environment. Since its inception in the early 1970s, ATC
has developed the technical basis for the current model national seismic design codes for buildings; the de
facto national standard for postearthquake safety evaluation of buildings; nationally applicable guidelines
and procedures for the identification, evaluation, and rehabilitation of seismically hazardous buildings;
and other widely used procedures and data to improve structural engineering practice. CUREe is a
nonprofit organization formed to promote and conduct research and educational activities related to
earthquake hazard mitigation. CUREe’s eight institutional members are the California Institute of
Technology, Stanford University, the University of California at Berkeley, the University of California at
Davis, the University of California at Irvine, the University of California at Los Angeles, the University
of California at San Diego, and the University of Southern California. These laboratory, library,
computer and faculty resources are among the most extensive in the United States. The SAC Joint
Venture allows these three organizations to combine their extensive and unique resources, augmented by
subcontractor universities and organizations from across the nation, into an integrated team of
practitioners and researchers, uniquely qualified to solve problems related to the seismic performance of
steel moment-frame buildings.

ACKNOWLEDGEMENTS
Funding for Phases I and II of the SAC Steel Program to Reduce the Earthquake Hazards of Steel
Moment-Frame Structures was principally provided by the Federal Emergency Management Agency,
with ten percent of the Phase I program funded by the State of California, Office of Emergency Services.
Substantial additional support, in the form of donated materials, services, and data has been provided by a
number of individual consulting engineers, inspectors, researchers, fabricators, materials suppliers and
industry groups. Special efforts have been made to maintain a liaison with the engineering profession,
researchers, the steel industry, fabricators, code-writing organizations and model code groups, building
officials, insurance and risk-management groups, and federal and state agencies active in earthquake
hazard mitigation efforts. SAC wishes to acknowledge the support and participation of each of the above
groups, organizations and individuals. In particular, we wish to acknowledge the contributions provided
by the American Institute of Steel Construction, the Lincoln Electric Company, the National Institute of
Standards and Technology, the National Science Foundation, and the Structural Shape Producers Council.
SAC also takes this opportunity to acknowledge the efforts of the project participants – the managers,
investigators, writers, and editorial and production staff – whose work has contributed to the development
of these documents. Finally, SAC extends special acknowledgement to Mr. Michael Mahoney, FEMA
Project Officer, and Dr. Robert Hanson, FEMA Technical Advisor, for their continued support and
contribution to the success of this effort.

In Memory of Egor Popov, Professor Emeritus, University of California at Berkeley

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Table of Contents

TABLE OF CONTENTS
LIST OF FIGURES ...................................................................................................................... vii
LIST OF TABLES......................................................................................................................... ix
1. INTRODUCTION ................................................................................................................. 1-1
1.1 Purpose.......................................................................................................................... 1-2
1.2 Background ................................................................................................................... 1-2
1.3 Overview....................................................................................................................... 1-9
1.4 Approach..................................................................................................................... 1-11
1.5 Limitations .................................................................................................................. 1-11
1.6 Summary ..................................................................................................................... 1-14
2. DESIGN AND CONSTRUCTION OF WSMFs IN SEISMIC AREAS ............................... 2-1
2.1 Cast Iron and Wrought Iron Construction .................................................................... 2-1
2.2 Transition to Steel......................................................................................................... 2-3
2.3 Evolution to Moment Frames with Bolted Connections .............................................. 2-3
2.4 Welded Moment Frames and the “Pre-Northridge” Connection.................................. 2-4
2.5 Optimized Design ......................................................................................................... 2-8
3. TESTING OF STEEL MOMENT-FRAME CONNECTIONS............................................. 3-1
3.1 Early Testing................................................................................................................. 3-4
3.2 Bouwkamp and Clough; Popov and Franklin; Beedle (1965) ...................................... 3-5
3.3 Popov and Pinkney (1969)............................................................................................ 3-6
3.4 Bertero, Popov, and Krawinkler (1972)........................................................................ 3-6
3.5 Popov and Stephen (1972), Popov and Bertero (1973) ................................................ 3-7
3.6 Popov, Amin, Louie, and Stephen (1985)..................................................................... 3-7
3.7 Popov (1987)................................................................................................................. 3-8
3.8 Popov and Tsai (1987); Tsai and Popov (1988) ........................................................... 3-8
3.9 Popov, Tsai, and Engelhardt (1988) ............................................................................. 3-8
3.10 Anderson and Linderman (1991) .................................................................................. 3-9
3.11 Schneider, Roeder, and Carpenter (1993)..................................................................... 3-9
3.12 Engelhardt and Husain (1993) ...................................................................................... 3-9
3.13 Roeder and Foutch (1996) .......................................................................................... 3-10
3.14 Connection Testing Since Northridge......................................................................... 3-11
4. CODES AND STANDARDS FOR STEEL MOMENT FRAMES ...................................... 4-1
4.1 1906-1924 ..................................................................................................................... 4-5
4.2 1925-1932 ..................................................................................................................... 4-6
4.3 1933-1958 ..................................................................................................................... 4-6
4.4 1959-1965 ..................................................................................................................... 4-7
4.5 1966-1985 ..................................................................................................................... 4-7
4.6 1986-1988 ..................................................................................................................... 4-9
4.7 1989-1993 ................................................................................................................... 4-10
iii

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Table of Contents

5. PERFORMANCE OF STEEL FRAME BUILDINGS IN PAST EARTHQUAKES ........... 5-1
5.1 San Francisco, 1906 ...................................................................................................... 5-3
5.2 Kanto, Japan, 1923...................................................................................................... 5-11
5.3 Santa Barbara, 1925 .................................................................................................... 5-11
5.4 Long Beach, 1933 ....................................................................................................... 5-12
5.5 Kern County, 1952...................................................................................................... 5-12
5.6 Prince William Sound, Alaska, 1964.......................................................................... 5-12
5.7 San Fernando, 1971 .................................................................................................... 5-14
5.8 Mexico City, 1985 ...................................................................................................... 5-17
5.9 Loma Prieta, 1989....................................................................................................... 5-19
5.10 Landers and Big Bear, 1992 ....................................................................................... 5-22
6. PERFORMANCE OF WSMFs IN THE 1994 NORTHRIDGE EARTHQUAKE ............... 6-1
6.1 Early Findings and Engineering Response ................................................................... 6-2
6.2 New Regulation ............................................................................................................ 6-3
6.3 Social, Economic, and Political Effects........................................................................ 6-4
6.3.1 Changes in Practice........................................................................................... 6-5
6.3.2 Legislation and Public Policy ........................................................................... 6-6
6.3.3 Legal Implications ............................................................................................ 6-9
6.4 Damage Data............................................................................................................... 6-10
6.4.1 W1 Flaws ........................................................................................................ 6-12
6.4.2 Damage Data................................................................................................... 6-12
6.4.3 Using the Damage Data .................................................................................. 6-14
6.5 Case Studies ................................................................................................................ 6-16
6.5.1 Krawinkler et al. ............................................................................................. 6-19
6.5.2 Engelhardt et al. .............................................................................................. 6-19
6.5.3 Hart et al.......................................................................................................... 6-19
6.5.4 Naeim et al. ..................................................................................................... 6-20
6.5.5 Uang et al. ....................................................................................................... 6-20
6.5.6 Kariotis and Eimani ........................................................................................ 6-20
6.5.7 Paret and Sasaki .............................................................................................. 6-21
6.5.8 Santa Clarita City Hall (Green) ...................................................................... 6-21
6.5.9 Anderson, Johnston, and Partridge ................................................................. 6-21
6.5.10 Borax Corporate Headquarters (Hajjar et al.)................................................. 6-22
APPENDIX A. WSMF DATA FROM THE NORTHRIDGE EARTHQUAKE...................... A-1
APPENDIX B. NORTHRIDGE EARTHQUAKE WSMF BUILDING DAMAGE. ................B-1
B.1 Introduction...................................................................................................................B-1
B.2 Connection Component Damage ..................................................................................B-1
B.3 Sources of Building Survey Data..................................................................................B-2
B.4 Seismic Demands..........................................................................................................B-3
B.5 Building Database.........................................................................................................B-4
B.6 Summary Statistics........................................................................................................B-4
B.7 References...................................................................................................................B-11
iv

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Table of Contents

APPENDIX C. OVERVIEW OF DAMAGE TO STEEL BUILDING STRUCTURES
OBSERVED IN THE 1995 KOBE EARTHQUAKE. .....................................C-1
C.1 Summary .......................................................................................................................C-1
C.2 Introduction...................................................................................................................C-1
C.3 Damage to Steel Buildings ...........................................................................................C-2
C.4 General Damage Statistics for Modern Steel Buildings ...............................................C-4
C.5 Damage to Members in Modern Buildings...................................................................C-4
C.6 Design and Construction Practices Before Kobe Earthquake Damage ........................C-6
C.7 Comparison of Building Damage in the U.S. and Japan ..............................................C-8
C.8 Partial Summary of Post-Kobe Japanese Research ....................................................C-10
C.9 Conclusions.................................................................................................................C-13
APPENDIX D. DAMAGE TO STEEL BUILDINGS DUE TO THE
SEPTEMBER 21, 1999 JI JI, TAIWAN EARTHQUAKE ............................. D-1
REFERENCES, FEMA REPORTS, SAC REPORTS, AND ACRONYMS ..............................R-1
SAC PHASE II PROJECT PARTICIPANTS ............................................................................. S-1

v

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
List of Figures

LIST OF FIGURES
Figure 1-1
Figure 1-2
Figure 1-3
Figure 1-4
Figure 1-5
Figure 2-1
Figure 2-2
Figure 2-3
Figure 2-4
Figure 2-5
Figure 2-6
Figure 3-1
Figure 3-2
Figure 5-1
Figure 5-2
Figure 5-3
Figure 5-4
Figure B-1
Figure B-2
Figure B-3
Figure B-4
Figure B-5
Figure B-6
Figure B-7
Figure B-8
Figure B-9
Figure B-10
Figure B-11
Figure C-1

Figure C-2
Figure C-3
Figure C-4
Figure C-5
Figure C-6

Typical Welded Moment-Resisting Connection Prior to 1994 ........................... 1-4
Common Zone of Fracture Initiation in Beam-Column Connection ................... 1-4
Fractures of Beam-to-Column Joints ................................................................... 1-5
Column Fractures................................................................................................. 1-5
Vertical Fracture through Beam Shear Plate Connection.................................... 1-6
Riveted Beam-Column Connection, Pre-1920s................................................... 2-4
Bolted and Riveted Connection, 1930s................................................................ 2-5
Welded and Bolted Moment Connection, 1950-1960 ......................................... 2-5
Welded Moment Connection, 1980s.................................................................... 2-6
Rise in Steel Construction in Los Angeles County.............................................. 2-8
WSMF Building with Setbacks ......................................................................... 2-10
WSMF Connection Test Setup ............................................................................ 3-2
Testing Apparatus for Modeling 16-Story Infilled Frame Building.................... 3-5
Damaged Moment-Frame Column, Prince William Sound
Earthquake, 1964 ............................................................................................... 5-13
Damage to Steel Frame Column, Mexico City Earthquake, 1985..................... 5-19
Typical Undamaged Joint in Torre Latino Americana, Mexico City
Earthquake, 1985 ............................................................................................... 5-20
Location of WSMF Buildings with Known Connection Damage,
Loma Prieta Earthquake, 1989........................................................................... 5-21
Typical Components of Pre-Northridge Moment Connections ...........................B-1
Distribution of Component Damage....................................................................B-2
Spatial Distribution of Screened Buildings .........................................................B-6
Spatial Distribution of Damaged Buildings.........................................................B-7
Distribution of Building Heights .........................................................................B-8
Distribution of Building Areas ............................................................................B-8
Distribution of Total Areas ..................................................................................B-9
Distribution of Peak Ground Accelerations.........................................................B-9
Distribution of Connection Inspection Rates.....................................................B-10
Distribution of Connection Damage Rates ........................................................B-10
Damage Rates Versus PGA ...............................................................................B-11
Market Share for Japanese Steel Building Construction;
(a) Floor Areas with Respect to Structural Material; (b) Floor Areas
with Respect to Number of Stories ....................................................................C-14
Damage to Old Steel Buildings; (a) Collapse; (b) Construction with
Light-Gauged Sections; (c) Corroded Sections .................................................C-15
Damage Level with Respect to Number of Stories............................................C-16
Types of Beam-to-Column Connections Popular in Japan; (a) ThroughDiaphragm Connection; (b) Interior-Diaphragm Connection ...........................C-16
Types of Column Bases Popular in Japan; (a) Base Plate Connection;
(b) Concrete Encased Column Base; (c) Embedded Column Base ...................C-17
Distribution of Damage Level with Respect to Structural Type .......................C-17

vii

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
List of Figures

Figure C-7
Figure C-8
Figure C-9
Figure C-10
Figure C-11
Figure C-12
Figure C-13

Figure C-14
Figure C-15
Figure C-16

Figure C-17

Figure C-18

Figure C-19

Figure D-1
Figure D-2
Figure D-3
Figure D-4

Damage to Structural Members with Respect to Structural Type .....................C-18
Fracture at Cold-Formed Square Tube Column.................................................C-18
Fracture of Square Tube Jumbo Columns; (a) Fracture at Base Metal;
(b) Fracture at Brace-To-Column Connection...................................................C-19
Damage To Brace Connections; (a) Breakage of Bolts; (b) Beam Web
Buckling and Out-of-Plane Deformation...........................................................C-19
Damage Location and Level of Column Base Connections ..............................C-20
Fracture at Beam-To-Column Connections with Fillet Welding of
Small Sizes; (a) Fracture at Column Top; (b) Fracture at Beam End................C-20
Fracture at Beam-To-Column Connections with Full Penetration
Groove Welding; (a) Fracture at Base Metal Initiating from Toe of
Weld Access Hole; (b) Fracture Involving Yielding and Local Buckling ........C-21
Material Properties of Base Metal Near Fractured Surface; (a) Charpy
V-Notch Test; (b) Hardness Test .......................................................................C-21
Distribution of Damage To Beam-To-Column Connections with
Respect to Type and Location ...........................................................................C-22
Design Base Shear of Level-I Japanese Seismic Design Code and Pseudo
Acceleration Response Spectra of Large Ground Motions Recorded in
Kobe Earthquake................................................................................................C-22
Weld Access Hole Details Proposed after Kobe Earthquake;
(a) Pre-Kobe Standard Detail; (b) Modified Detail with Smaller Hole;
(c) No-Hole Detail .............................................................................................C-23
Comparison Between U.S. RBS Connection and Japanese No-Hole
Connection; (a) No-Hole Connection Specimen; (b) RBS Connection
Specimen............................................................................................................C-23
Beam End Moment Versus Beam Rotation Relationships Obtained
from Test; (a) No-Hole Connection; (b)RBS Connection; (c) Japanese
Pre-Kobe Connection.........................................................................................C-24
Permanent Lateral Displacement in Small Steel Frame ..................................... D-3
Typical Taiwanese High-Rise Structure under Construction in
Tiachung ............................................................................................................. D-3
Welded Connection Detail of Beams to Box Column ........................................ D-4
Bolted End Plate Connection in Light Steel Frame Building Showing
Evidence of Local Yielding ................................................................................ D-4

viii

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
List of Tables

LIST OF TABLES
Table 1-1
Table 2-1
Table 2-2
Table 3-1
Table 4-1
Table 4-2
Table 5-1
Table 5-2
Table 5-3
Table 5-4
Table 5-5
Table 6-1
Table 6-2
Table 6-3
Table A-1
Table B-1
Table B-2
Table B-3
Table C-1
Table C-2
Table D-1

Milestones in U.S. Steel Frame Research, Regulation, Practice, and
Performance ....................................................................................................... 1-12
Milestones in the Development of Structural Steel Buildings............................. 2-2
Relative Costs of Moment Connections .............................................................. 2-6
Milestones in the Research on WSMF Connections ........................................... 3-3
Milestones in Code Development for Steel Moment Frames.............................. 4-2
Uniform Building Code Provisions for Steel Moment-Frame Buildings ............ 4-3
Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era ........................................................................................................... 5-4
Damage by Structure Type in Selected North American Earthquakes of
the WSMF Era ................................................................................................... 5-10
Case Studies of Instrumental WSMF Buildings Affected by the 1971
San Fernando Earthquake .................................................................................. 5-15
Performance of Modern Steel Structural Systems, 1985 Mexico City
Earthquake ......................................................................................................... 5-18
Damage to WSMF Buildings in the 1989 Loma Prieta Earthquake.................. 5-20
Source Lists of WSMF Buildings Affected by the 1994 Northridge
Earthquake ......................................................................................................... 6-10
Number of WSMF Buildings with Various Northridge Earthquake
Damage Rates .................................................................................................... 6-14
Case Studies of WSMF Buildings Affected by the 1994 Northridge
Earthquake ......................................................................................................... 6-17
Master List of Northridge WSMF Databases Sorted by SAC Building
ID and L.A. Building ID ..................................................................................... A-1
Definitions Used in Northridge Database (Partial List) ......................................B-4
185 Inspected WSMF Buildings Affected by the 1994 Northridge
Earthquake Sorted By Height (As of 11/99)......................................................B-13
185 Inspected WSMF Buildings Affected by the 1994 Northridge
Earthquake Sorted by Peak Ground Acceleration (As of 11/99).......................B-19
Cross-Sections Used In Damaged Steel Buildings; (a) Columns;
(b) Beams; (c) Braces ........................................................................................C-13
Types of Connections Used in Damaged Buildings; (a) Columns;
(b) Beams; (c) Beam-to-Column Connections; (d) Column Bases ...................C-14
Statistics on Damage Due to September 21, 1999 Taiwan Earthquake,
by Type of Building Material (NCREE, 1999)................................................... D-2

ix

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

1. INTRODUCTION
Purpose

1.1

This report, FEMA-355E – State of the Art Report on Past Performance of Steel MomentFrame Buildings in Earthquakes, presents an overview of the development of the welded
moment-resisting steel frame system as a preferred system for seismic resistance in the United
States and the limited data upon which this reputation was based. This state of the art report was
prepared in support of the development of a series of Recommended Design Criteria documents,
prepared by the SAC Joint Venture on behalf of the Federal Emergency Management Agency
(FEMA) and addressing the issue of the seismic performance of moment-resisting steel frame
structures. These publications include:
•

FEMA-350 – Recommended Seismic Design Criteria for New Steel Moment-Frame
Buildings. This publication provides recommended criteria, supplemental to FEMA 302 –
1997 NEHRP Recommended Provisions for Seismic Regulations for New Buildings and
other Structures, for the design and construction of steel moment-frame buildings and
provides alternative performance-based design criteria.

•

FEMA-351 – Recommended Seismic Evaluation and Upgrade Criteria for Existing Welded
Steel Moment-Frame Buildings. This publication provides recommended methods to
evaluate the probable performance of existing steel moment-frame buildings in future
earthquakes and to retrofit these buildings for improved performance.

•

FEMA-352 – Recommended Postearthquake Evaluation and Repair Criteria for Welded
Steel Moment-Frame Buildings. This publication provides recommendations for performing
postearthquake inspections to detect damage in steel moment-frame buildings following an
earthquake, evaluating the damaged buildings to determine their safety in the postearthquake
environment, and repairing damaged buildings.

•

FEMA-353 – Recommended Specifications and Quality Assurance Guidelines for Steel
Moment-Frame Construction for Seismic Applications. This publication provides
recommended specifications for the fabrication and erection of steel moment frames for
seismic applications. The recommended design criteria contained in the other companion
documents are based on the material and workmanship standards contained in this document,
which also includes discussion of the basis for the quality control and quality assurance
criteria contained in the recommended specifications.

Detailed derivations and explanations of the basis for these design and evaluation
recommendations may be found in a series of State of the Art Report documents prepared by the
SAC Joint Venture in parallel with these design criteria. These reports include:
•

FEMA-355A – State of the Art Report on Base Metals and Fracture. This report
summarizes current knowledge of the properties of structural steels commonly employed
in building construction, and the production and service factors that affect these
properties.
1-1

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

•

FEMA-355B – State of the Art Report on Welding and Inspection. This report
summarizes current knowledge of the properties of structural welding commonly
employed in building construction, the effect of various welding parameters on these
properties, and the effectiveness of various inspection methodologies in characterizing
the quality of welded construction.

•

FEMA-355C – State of the Art Report on Systems Performance of Steel Moment Frames
Subject to Earthquake Ground Shaking. This report summarizes an extensive series of
analytical investigations into the demands induced in steel moment-frame buildings
designed to various criteria, when subjected to a range of different ground motions. The
behavior of frames constructed with fully restrained, partially restrained and fracturevulnerable connections is explored for a series of ground motions, including motion
anticipated at near-fault and soft-soil sites.

•

FEMA-355D – State of the Art Report on Connection Performance. This report
summarizes the current state of knowledge of the performance of different types of
moment-resisting connections under large inelastic deformation demands. It includes
information on fully restrained and partially restrained moment connections in welded
and bolted configurations, based upon laboratory and analytical investigations.

•

FEMA-355E – State of the Art Report on Past Performance of Steel Moment-Frame
Buildings in Earthquakes. This report summarizes investigations of the performance of
steel moment-frame buildings in past earthquakes, including the 1995 Kobe, 1994
Northridge, 1992 Landers, 1992 Big Bear, 1989 Loma Prieta and 1971 San Fernando
events.

•

FEMA-355F – State of the Art Report on Performance Prediction and Evaluation of Steel
Moment-Frame Buildings. This report describes the results of investigations into the
ability of various analytical techniques, commonly used in design, to predict the
performance of steel moment-frame buildings subjected to earthquake ground motion.
Also presented is the basis for performance-based evaluation procedures contained in the
design criteria documents, FEMA-350, FEMA-351, and FEMA-352.

In addition to the recommended design criteria and the State of the Art Reports, a companion
document has been prepared for building owners, local community officials and other nontechnical audiences who need to understand this issue. A Policy Guide to Steel Moment-Frame
Construction (FEMA 354), addresses the social, economic, and political issues related to the
earthquake performance of steel moment-frame buildings. FEMA 354 also includes discussion
of the relative costs and benefits of implementing the recommended criteria.
1.2

Background

For many years, the basic intent of the building code seismic provisions was to provide
buildings with an ability to withstand intense ground shaking without collapse, but potentially
with some significant structural damage. In order to accomplish this, one of the basic principles
inherent in modern code provisions is to encourage the use of building configurations, structural
systems, materials and details that are capable of ductile behavior. A structure is said to behave
in a ductile manner if it is capable of withstanding large inelastic deformations without
1-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

significant degradation in strength, and without the development of instability and collapse. The
design forces specified by building codes for particular structural systems are related to the
amount of ductility the system is deemed to possess. Generally, the building code allows
structural systems with more ductility to be designed for lower forces than less ductile systems,
as ductile systems are deemed capable of resisting demands that are significantly greater than
their elastic strength limit. Starting in the 1960s, engineers began to regard welded steel
moment-frame buildings as being among the most ductile systems contained in the building
code. Many engineers believed that steel moment-frame buildings were essentially invulnerable
to earthquake-induced structural damage and thought that should such damage occur, it would be
limited to ductile yielding of members and connections. Earthquake-induced collapse was not
believed possible. Partly as a result of this belief, many large industrial, commercial and
institutional structures employing steel moment-frame systems were constructed, particularly in
the western United States.
The Northridge earthquake of January 17, 1994 challenged this paradigm. Following that
earthquake, a number of steel moment-frame buildings were found to have experienced brittle
fractures of beam-to-column connections. The damaged buildings had heights ranging from one
story to 26 stories, and a range of ages spanning from buildings as old as 30 years to structures
being erected at the time of the earthquake. The damaged buildings were spread over a large
geographical area, including sites that experienced only moderate levels of ground shaking.
Although relatively few buildings were located on sites that experienced the strongest ground
shaking, damage to buildings on these sites was, in many cases, quite extensive. Discovery of
these unanticipated brittle fractures of framing connections, often with little associated
architectural damage to the buildings, was alarming to all concerned. The discovery also caused
some concern that similar, but undiscovered, damage may have occurred in other buildings
affected by past earthquakes. Later investigations confirmed such damage in a limited number
of buildings affected by the 1992 Landers, 1992 Big Bear and 1989 Loma Prieta earthquakes.
In general, steel moment-frame buildings damaged by the 1994 Northridge earthquake met
the basic intent of the building codes. That is, they experienced limited structural damage, but
did not collapse. However, the structures did not behave as anticipated and significant economic
losses occurred as a result of the connection damage, in some cases, in buildings that had
experienced ground shaking less severe than the design level. These losses included direct costs
associated with the investigation and repair of this damage as well as indirect losses relating to
the temporary, and in a few cases, long-term, loss of use of space within damaged buildings.
Steel moment-frame buildings are designed to resist earthquake ground shaking based on the
assumption that they are capable of extensive yielding and plastic deformation, without loss of
strength. The intended plastic deformation consists of plastic rotations developing within the
beams, at their connections to the columns, and is theoretically capable of resulting in benign
dissipation of the earthquake energy delivered to the building. Damage is expected to consist of
moderate yielding and localized buckling of the steel elements, not brittle fractures. Based on this
presumed behavior, building codes permit steel moment-frame buildings to be designed with a
fraction of the strength that would be required to respond to design level earthquake ground shaking
in an elastic manner.

1-3

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

Steel moment-frame buildings are anticipated to develop their ductility through the
development of yielding in beam-column assemblies at the beam-column connections. This
yielding may take the form of plastic hinging in the beams (or less desirably, in the columns),
plastic shear deformation in the column panel zones, or through a combination of these
mechanisms. It was believed that the typical connection employed in steel moment-frame
construction, shown in Figure 1-1, was capable of developing large plastic rotations, on the order
of 0.015 to 0.02 radians, without significant strength degradation.
Observation of damage sustained by buildings in the 1994 Northridge earthquake indicated
that contrary to the intended behavior, in many cases brittle fractures initiated within the
connections at very low levels of plastic demand, and in some cases, while the structures
remained essentially elastic. Typically, but not always, fractures initiated at the complete joint
penetration (CJP) weld between the beam bottom flange and column flange (Figure 1-2). Once
initiated, these fractures progressed along a number of different paths, depending on the
individual joint conditions.

Figure 1-1

Typical Welded Moment-Resisting Connection Prior to 1994
Column flange
Fused zone
Beam flange

Backing bar

Figure 1-2

Fracture
Common Zone of Fracture Initiation in Beam-Column Connection

1-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

In some cases, the fractures progressed completely through the thickness of the weld, and
when fire protective finishes were removed, the fractures were evident as a crack through
exposed faces of the weld, or the metal just behind the weld (Figure 1-3a). Other fracture
patterns also developed. In some cases, the fracture developed into a crack of the column flange
material behind the CJP weld (Figure 1-3b). In these cases, a portion of the column flange
remained bonded to the beam flange, but pulled free from the remainder of the column. This
fracture pattern has sometimes been termed a “divot” or “nugget” failure.
A number of fractures progressed completely through the column flange, along a nearhorizontal plane that aligns approximately with the beam lower flange (Figure 1-4a). In some
cases, these fractures extended into the column web and progressed across the panel zone (Figure
1-4b). Investigators have reported some instances where columns fractured entirely across the
section.

a. Fracture at Fused Zone
Figure 1-3

a.

b. Column Flange "Divot" Fracture

Fractures of Beam-to-Column Joints

a. Fracture Progresses into Column Web

Fractures through Column Flange
Figure 1-4

Column Fractures

1-5

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

Once such fractures have occurred, the beam-column connection has experienced a
significant loss of flexural rigidity and strength to resist those loads that tend to open the crack.
Residual flexural strength and rigidity must be developed through a couple consisting of forces
transmitted through the remaining top flange connection and the web bolts. However, in
providing this residual strength and stiffness, the bolted web connections can themselves be
subject to failures. These include fracturing of the welds of the shear plate to the column,
fracturing of supplemental welds to the beam web, or fracturing through the weak section of
shear plate aligning with the bolt holes (Figure 1-5).
Despite the obvious local strength impairment resulting from these fractures, many damaged
buildings did not display overt signs of structural damage, such as permanent drifts or damage to
architectural elements, making reliable postearthquake damage evaluations difficult. In order to
determine reliably if a building has sustained connection damage it is necessary to remove
architectural finishes and fireproofing, and perform detailed inspections of the connections.
Even if no damage is found, this is a costly process. Repair of damaged connections is even
more costly. At least one steel moment-frame building sustained so much damage that it was
deemed more practical to demolish the building than to repair it.

Figure 1-5

Vertical Fracture through Beam Shear Plate Connection

Initially, the steel construction industry took the lead in investigating the causes of this
unanticipated damage and in developing design recommendations. The American Institute of
Steel Construction (AISC) convened a special task committee in March, 1994 to collect and
disseminate available information on the extent of the problem (AISC, 1994a). In addition,
together with a private party engaged in the construction of a major steel building at the time of
the earthquake, AISC participated in sponsoring a limited series of tests of alternative connection
details at the University of Texas at Austin (AISC, 1994b). The American Welding Society
(AWS) also convened a special task group to investigate the extent to which the damage was
related to welding practice, and to determine if changes to the welding code were appropriate
(AWS, 1995).
In September 1994, the SAC Joint Venture, AISC, the American Iron and Steel Institute and
National Institute of Standards and Technology jointly convened an international workshop
(SAC, 1994) in Los Angeles to coordinate the efforts of the various participants and to lay the
1-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

foundation for systematic investigation and resolution of the problem. Following this workshop,
FEMA entered into a cooperative agreement with the SAC Joint Venture to perform problemfocused studies of the seismic performance of steel moment-frame buildings and to develop
recommendations for professional practice (Phase I of SAC Steel Project). Specifically, these
recommendations were intended to address the following: the inspection of earthquake-affected
buildings to determine if they had sustained significant damage; the repair of damaged buildings;
the upgrade of existing buildings to improve their probable future performance; and the design of
new structures to provide reliable seismic performance.
During the first half of 1995, an intensive program of research was conducted to explore
more definitively the pertinent issues. This research included literature surveys, data collection
on affected structures, statistical evaluation of the collected data, analytical studies of damaged
and undamaged buildings, and laboratory testing of a series of full-scale beam-column
assemblies representing typical pre-Northridge design and construction practice as well as
various repair, upgrade, and alternative design details. The findings of these tasks formed the
basis for the development of FEMA-267 – Interim Guidelines: Evaluation, Repair, Modification,
and Design of Welded Steel Moment Frame Structures, which was published in August, 1995.
FEMA-267 provided the first definitive, albeit interim, recommendations for practice, following
the discovery of connection damage in the 1994 Northridge earthquake.
In September 1995, the SAC Joint Venture entered into a contractual agreement with FEMA
to conduct Phase II of the SAC Steel Project. Under Phase II, SAC continued its extensive
problem-focused study of the performance of moment-resisting steel frames and connections of
various configurations, with the ultimate goal of developing reliable seismic design criteria for
steel construction. This work has included: extensive analyses of buildings; detailed finite
element and fracture mechanics investigations of various connections to identify the effects of
connection configuration, material strength, and toughness and weld joint quality on connection
behavior; as well as more than 120 full-scale tests of connection assemblies. As a result of these
studies, and independent research conducted by others, it is now known that the typical momentresisting connection detail employed in steel moment-frame construction prior to the 1994
Northridge earthquake, and depicted in Figure 1-1, had a number of features that rendered it
inherently susceptible to brittle fracture. These included the following:
•

The most severe stresses in the connection assembly occurred where the beam joins to the
column. Unfortunately, this is also the weakest location in the assembly. At this location,
bending moments and shear forces in the beam must be transferred to the column through the
combined action of the welded joints between the beam flanges and column flanges and the
shear tab. The combined section properties of these elements, for example the cross sectional
area and section modulus, were typically less than those of the connected beam. As a result,
stresses were locally intensified at this location.

•

The joint between the bottom beam flange and the column flange was typically made as a
downhand field weld, often by a welder sitting on top of the beam top flange, in a so-called
“wildcat” position. To make the weld from this position, each pass was interrupted at the
beam web, with either a start or stop of the weld at this location. Further, the welder often
completed all passes on one side of the beam web rather than alternating from one side to the

1-7

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

other as required. This welding technique often resulted in poor quality welding at this
critical location, with slag inclusions, lack of fusion, and other defects. These defects can
serve as crack initiators, when the connection is subjected to severe stress and strain
demands.
•

The basic configuration of the connection made it difficult to detect hidden defects at the root
of the welded beam-flange-to-column-flange joints. The backing bar, which was typically
left in place following weld completion, restricts visual observation of the weld root.
Therefore, the primary method of detecting defects in these joints was through the use of
ultrasonic testing (UT). However, the geometry of the connection also made it very difficult
for UT to detect flaws reliably at the bottom beam flange weld root, particularly at the center
of the joint, at the beam web. As a result, many of these welded joints had undetected
significant defects that can serve as crack initiators.

•

Although typical design models for this connection assume that nearly all beam flexural
stresses are transmitted by the flanges and all beam shear forces by the web, in reality, due to
boundary conditions imposed by column deformations, the beam flanges at the connection
carry a significant amount of the beam shear. This results in significant flexural stresses on
the beam flange at the face of the column, and also induces large secondary stresses in the
welded joint. Some of the earliest investigations of these stress concentration effects in the
welded joint were conducted by Richard, et al. (1995). The stress concentrations resulting
from this effect resulted in severe strength demands at the root of the complete joint
penetration welds between the beam flanges and column flanges, a region that often includes
significant discontinuities and slag inclusions, which are ready crack initiators.

•

Weld access holes were needed to complete both the top and bottom flange welds.
Depending on their geometry, severe strain concentrations can occur in the beam flange at
the toe of these weld access holes. These strain concentrations can result in low-cycle fatigue
and the initiation of ductile tearing of the beam flanges after only a few cycles of moderate
plastic deformation. Under large plastic flexural demands, these ductile tears can quickly
become unstable and propagate across the beam flange.

•

The center of the beam-flange-to-column-flange joint is restrained from movement,
particularly in connections of heavy sections with thick beam flanges. This condition of
restraint inhibits the development of yielding at this location, resulting in locally high stresses
on the welded joint, which exacerbates the tendency to initiate fractures at defects in the
welded joints.

•

Design practice in the period from 1985 to 1994 encouraged connections with relatively
weak panel zones. In connections with excessively weak panel zones, inelastic behavior of
the assembly is dominated by shear deformation of the panel zone. This panel zone shear
deformation results in a local kinking of the column flanges adjacent to the beam-flange-tocolumn-flange joint, and further increases the stress and strain demands in this sensitive
region.

In addition to the above, additional conditions contributed significantly to the vulnerability of
connections constructed prior to 1994.

1-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

•

In the mid-1960s, the construction industry moved to the use of the semi-automatic, selfshielded, flux-cored arc welding process (FCAW-SS) for making the joints of these
connections. The specific welding consumables that building erectors most commonly used
under this process inherently produced welds with very low notch toughness. The weld
quality and notch toughness of this material could be further compromised by excessive
deposition rates, which unfortunately were commonly employed by welders. As a result,
brittle fractures could initiate in welds with large defects, at stresses approximating the yield
strength of the beam steel, precluding the development of ductile behavior.

•

Early steel moment frames tended to be highly redundant and nearly every beam-column
joint was constructed to behave as part of the lateral-force-resisting system. As a result,
member sizes in these early frames were small and much of the early acceptance testing of
this typical detail were conducted with specimens constructed of small framing members. As
the cost of construction labor increased, the industry found that it was more economical to
construct steel moment-frame buildings by moment-connecting a relatively small percentage
of the beams and columns and by using larger members for these few moment-connected
elements. The amount of strain demand placed on the connection elements of a steel moment
frame is related to the span-to-depth ratio of the member. Therefore, as member sizes
increased, strain demands on the welded connections also increased, making the connections
more susceptible to brittle behavior.

•

In the 1980s, many steel mills adopted modern production processes, including the use of
scrap-based production. Steels produced by these more modern processes tended to include
micro-alloying elements that increased the yield strength of the materials so that despite the
common specification of A36 material for beams, many beams actually had yield strengths
that approximated or exceeded that required for grade 50 material. As a result of this
increase in base metal yield strength, the weld metal in the beam-flange-to-column-flange
joints became under-matched, potentially contributing to its vulnerability.

At this time, it is clear that in order to obtain reliable ductile behavior of steel moment-frame
construction, a number of changes to past practices in design, materials, fabrication, erection and
quality assurance are necessary. The recommendations contained in this document, and the
companion publications, are based on an extensive program of research into materials, welding
technology, inspection methods, frame system behavior, and laboratory and analytical
investigations of different connection details.
1.3

Overview

Dynamic. Engineers know this word describes building behavior in earthquakes. Ground
motions impart energy to the elements of a structure, which interact in complex ways. Yielding
components redirect forces, resulting in a cascade of change that pushes the structure into a
different state. When the elements work together and complement each other, a building can
survive an earthquake with little damage. When they do not, the results can be devastating.
“Dynamic” is an appropriate descriptor for the behavior of the engineering and construction
communities as well. Powerful forces drive change in one sector, imparting energy and impetus
that soon affects the whole industry. Interests are varied, and their interactions complex.

1-9

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

Sometimes they work together, and the result is positive change. Sometimes they compete and
the result can be movement in the wrong direction or perpetuation of the status quo.
The history of welded steel moment frame (WSMF) buildings offers a case study of the
dynamic nature of the engineering community. There is an interwoven relationship of research,
regulation, practice, and, significantly, of nature. A careful study of this structural system and its
past earthquake performance traces a path that started in the 1850s, when steel became a mass
produced material, and led to unexpected results in the Northridge earthquake.
This report explores some of the forces that cut this path. It also illuminates several events
that shaped its direction. It is clear that the two most important of these forces have been “need”
and “nature.” Our need to build and improve our homes and workplaces led to the explosion of
the metropolis and demanded buildings mass produced in manners unimaginable in previous
centuries. The first “skyscrapers” rose over a hundred feet at the turn of the century. The tallest
buildings in any city had been churches; they became offices and apartments. New materials and
structural systems were needed. Steel alone could meet the need and so became the material of
choice for tall buildings.
As steel construction flourished, the steel industry sponsored research and contributed to
code development efforts. This sponsorship over the past forty years has allowed researchers to
investigate the anticipated seismic performance of WSMF connections. Their overall
expectations of good dynamic behavior were enthusiastic. But in many cases, researchers noted
that critical conditions at the beam-column interface could lead to premature brittle failures. In
retrospect, these concerns deserved more attention than they now appear to have received. All
who endorsed steel moment frames for seismic resistance—engineers, inspectors, building
officials, contractors, material suppliers, and the researchers themselves—share responsibility for
these oversights.
Nature has been more obvious in its impact. No other impetus, whether political, economic
or scientific, has had the ability to move the engineering and construction industries forward like
an earthquake. Steel construction has probably been the greatest beneficiary of earthquakes’
effects on buildings. After nearly every major event this century the steel building has been
hailed as an excellent performer and has been compared with examples of disastrous
performance of concrete and masonry structures. This perceived performance led engineers and
code writers to encourage the use of steel frames in seismic regions.
Although steel has generally outperformed other structural materials, the WSMF as a seismic
force-resisting system developed a glowing reputation that was perhaps undeserved. There has
not been conclusive evidence to substantiate the “excellent” performance of modern WSMFs. In
fact, there is a decided lack of evidence about the performance of WSMFs prior to 1994. The
term “steel frame” was used in numerous postearthquake reports to encompass all manner of
steel construction, not just WSMFs. There were only a handful of well-documented examples of
WSMF performance before Northridge. This is in part because the number of true WSMFs
shaken by earlier earthquakes was small. It is also due to our own tendency to focus on damage
that might be more obvious in concrete and masonry buildings and not on the more subtle
behavior of steel frames.
1-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

Table 1-1 summarizes some of the important milestones in the development of WSMF
construction, considering practice, regulation, research, and nature. Each perspective is
described in more detail in a separate chapter of this report.
1.4

Approach

In order to understand the performance of any structural system, it is essential to know where
that system came from. McGuire (1988) put it nicely: “As in all long-established branches of
technology, there is found in current practice a residual influence of decisions made and
directions taken long ago, before the underlying sciences were well understood.” Therefore,
before reviewing the record of WSMFs in earthquakes, this report presents a brief history of the
system. It offers both a review of past practice and an historical context in which to understand
current thinking. The report describes what the state of the art in steel frame design has been over
the past half century and how that state evolved finally to produce the current SAC Guidelines.
Specifically, the report reviews the following:
•

The development of WSMF connections, following the transition from iron to steel framing,
and tracing the use of riveted, bolted, and ultimately welded connections.

•

Research from the past thirty years on the performance and design of WSMF connections.

•

Milestones in the development of building code provisions related to WSMF construction.

•

The performance of steel moment frame connections in past earthquakes, including
Northridge. Summaries of the available raw data from Northridge are included in Appendices
A and B. Related data from the 1995 Kobe (Japan) and 1999 Ji-Ji (Taiwan) earthquakes is
provided in Appendix C.

There are two ways to present and assess the “past performance” of a given structural system.
One is to compare its performance with other systems. The other is to compare its performance
with the intentions and expectations of its own proponents. For the most part, this report takes
the latter approach, which the authors consider to be more useful, more reliable, and more
enlightening.
1.5

Limitations

This report focuses narrowly on topics that may be useful to the users of the SAC Guidelines.
The broader topic of steel construction and even of steel moment frame construction has not
been addressed in detail, nor is it the intent of this report to make a detailed comparison of steel
construction with other structural systems.
Information for this report has been gathered from previously published material and from
unpublished test reports and postearthquake data. No original research was performed. The data
gathered for this report and contained in the appendices is largely the original data. No effort has
been made to synthesize this information into recommendations for design or construction.
Researchers are encouraged to review the original data and reports produced by SAC and others,
many of which are cited in this report and listed in the References.

1-11

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

Table 1-1
Year
1850s

Milestones in U.S. Steel Frame Research, Regulation, Practice, and
Performance
Practice

Research

Bessemer process allows
mass production of steel.

1895
1900s

Regulation

First steel material specifications.
Steel “skyscrapers” reach
over 100 feet. Iron phased
out in favor of structural
steel.

1906

San Francisco earthquake
No WSMF structures, but buildings with steel frames perform well.

1906

1920s

San Francisco adopts 30 psf wind
and seismic load provisions.
Welding popular for
mechanical equipment.

1925

Santa Barbara earthquake
No WSMF structures, but buildings with steel frames perform well compared to those of other
materials.

1927

First seismic provisions written
into the UBC.

1933

Long Beach earthquake
No WSMF structures, but buildings with steel frames perform well compared to those of other
materials.

1933

California regulates design of state
buildings.

1937

Base shear as functions of soil and
height.

1920s–
1950s

Masonry infill phased out
in seismic regions.

1948,
1950s

High strength bolts replace
rivets.

UBC building K factor introduced.
Base shear a function of period.

1-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 1-1

Early
1960s

FEMA-355E
Chapter 1: Introduction

Milestones in U.S. Steel Frame Research, Regulation, Practice, and
Performance (continued)

Welding in building
construction becomes
popular.

1964

1959: SEAOC publishes first Blue
Book. K factor a function of
structural material and system.
Steel frame, or its equivalent,
required for tall buildings.
Prince William Sound earthquake

Considerable damage to all material types, including steel. Only a few true WSMF structures,
however, and performance is inconclusive.
Late
1960s

FCAW popular for high
production work.
Introduction of E70T-4
flux core wire.

AISI sponsored research on
moment frames. Potential
ductility demonstrated but weld
fractures noted.

1968: Blue Book defines K factor
for ductile WSMFs and defines
properties for ductile steel and
concrete systems.

San Fernando earthquake

1971

Considerable damage to concrete and masonry. Steel perceived to perform well. Only a few true
WSMF structures. Brittle fractures repaired in two buildings under construction. Completed
buildings not thoroughly inspected.
1972

Research focuses on panel zone
design.

1975

1980s

1985

Panel zone and continuity plate
requirements in Blue Book. Web
welds recommended.
Increased WSMF
construction. Detailing
requirements and
computer-aided design
lead to use of larger
sections and less redundant
frames.
Mexico City earthquake
Considerable damage to all material types. Some poorly configured steel braced frames perform
poorly. Only a few WSMFs. Weld damage is noted but overshadowed by other steel issues.

19851988

W18 and similar sized beams
tested. Poor ductility observed.
Weld fractures not raised as a
major concern.

1-13

UBC requires supplemental web
welds and strong column-weak
beam, relaxes panel zone
requirements, defines
“prequalified” WSMF connection.
Rw replaces K factor.

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

Table 1-1

Milestones in U.S. Steel Frame Research, Regulation, Practice, and
Performance (continued)

1989

Loma Prieta earthquake
Highlights poor performance of URMs and non-ductile concrete. Steel buildings, including WSMFs,
perceived to perform well. At least five buildings with connection damage are discovered, most upon
inspection after 1994 Northridge earthquake.

Early
1990s

1992: AISC seismic
specifications published.

1994

Research into repair of weld
cracks: initial specimens fail at
low plastic rotation. Tests of
supplemental web welds on
large sections: nearly 80% fail at
bottom flange welds.
Northridge earthquake

Unexpected fractures in some WSMF connections, but none cause death or serious injury. Expected
poor performance of outdated systems is realized.
1994

1.6

Engineers await approved
details for repair,
strengthening, and new
construction.

Northridge data collection and
testing of alternative WSMF
details begin.

ICBO enacts emergency code
change requiring cyclic testing of
moment frame joint designs.

Summary

This report reviews the past performance of welded steel moment frames as seismic forceresisting systems. The historical context in which this system developed is described as three
interrelated streams of activity: research, regulation, and practice. Each stream at times lagged
and at times led the others. All three responded to the redirecting forces of actual earthquakes.
The 1994 Northridge earthquake was a benchmark event for welded steel moment frames
(WSMFs). This report offers a post-Northridge view of their use in the United States and their
evolution from earlier steel frame systems. From a post-Northridge perspective, some
compelling lessons include:
•

The WSMF is a young structural system. Its essential components were not in place until
about 1970, and it evolved substantially over the next two decades. In Los Angeles County,
more structural steel was erected in the 1980s than in the previous two decades combined.

•

A quarter century of WSMF connection tests consistently demonstrated the potential for
outstanding seismic performance. But a broad view now reveals a pattern of fractures and
inconsistent acceptance criteria. Testing could not keep pace with practice, missing key
developments in member sizing, material properties, and erection procedures.

•

Building codes and standards embraced the connection tests and adopted their results
broadly, standardizing the connection detail even for untested sizes and unprecedented frame
configurations. Design codes also failed to address reliability issues raised by the tests.

1-14

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 1: Introduction

•

The modern WSMF had never been tested in meaningful numbers until the Northridge
earthquake. It missed the destructive San Fernando earthquake in 1971, and, perhaps because
other structural systems fared worse, it was largely unscrutinized after Loma Prieta in 1989.

•

Early estimates of WSMF damage in the Northridge earthquake were high. In the end, no
WSMFs collapsed, only a handful lost significant lateral capacity, and half of all inspected
WSMF buildings had no connection damage at all.

Nevertheless, the Northridge earthquake exposed a faulty detail that performed in the field
much as it had in the lab for twenty years. Most of the connections survived. Too many failed. In
many ways, this report merely updates a 1991 study of the past performance of steel structures in
earthquakes (Yanev et al., 1991). That report captured the pre-Northridge thinking of the entire
design and construction community—both the right and wrong of it—in a single short paragraph:
When failures of steel structures occur, connection failures are the most common cause.
No advantage can be derived from the strength and ductility of a steel member if its
connections fail prematurely. However, use of industry-standard details generally
provides acceptable performance.

1-15

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

2. DESIGN AND CONSTRUCTION OF WSMFs IN SEISMIC AREAS
The history of steel as a modern building construction material traces its evolution in large
part to its ease of manufacture and erection. The Bessemer process and the open hearth furnace,
developed in the mid 19th century, allowed steel to be mass produced. Fifty years later it was
already an essential material for “high-rise” buildings. By proportioning the alloying elements,
especially the carbon content, manufacturers could control the strength and ductility of steel to a
degree not possible with iron.
This chapter summarizes some milestones in steel frame design and construction. Much of
the information presented is taken from FEMA-274 (1997). An excellent brief history of steel
frame construction is given by McGuire (1988). The information presented here is not
exhaustive, but is intended to give the reader insight into how steel became such a widely used
and hailed material. With respect to the use of steel moment frames in seismic areas, this brief
history offers a number of lessons in hindsight:
•

The growth of cities and the need for denser development required practical structural
systems for tall buildings. Steel fit the bill. Good performance of steel buildings relative to
masonry ones in earthquakes made steel even more attractive.

•

The rigidity provided by large riveted connections in steel frames and the development of
curtain wall systems after World War II made the use of brittle infill masonry as a lateral
force-resisting element unnecessary.

•

The introduction of bolted and later welded connections allowed engineers to distinguish
moment and non moment-resisting connections. This allowed the use of fewer but larger
sections in discrete plane frames, offering an economic advantage over the more traditional
distributed system, but focusing seismic demands on fewer members and connections. The
ability to optimize structural systems with computer technology heightened this effect.

•

Increased production of steel frame buildings called for faster and more economical welding
processes. When higher deposition rates became available, engineers, fabricators, and
inspectors adopted them without much regard for weld toughness and appropriate quality
control.
Table 2-1 summarizes some of the milestones and developments discussed in this section.

2.1

Cast Iron and Wrought Iron Construction

Cast iron has been used in construction for over two thousand years and was popular in the
United States up until about 1900. Cast iron is stronger than wood or masonry and can support
taller buildings with relatively slender structural columns. However, cast iron is brittle and
unable to resist large bending or tensile demands. Therefore, it was not an ideal material for
moment frame structures, nor was it a reliable choice for resisting dynamic loads.

2-1

 FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

Table 2-1

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Milestones in the Development of Structural Steel Buildings

Date

Milestone

1850s

Bessemer process allows mass production of steel

1895

First specification for structural steel

1900s

Iron phased out as structural steel becomes easy to manufacture.

1906, 1925

Performance of steel in earthquakes increases its popularity.

1920s

Welding becomes popular for mechanical equipment manufacture.

1920s – 1950s

Steel frames with masonry infill phased out in seismic regions as steel moment frames are
constructed with built-up joint sections and as curtain wall systems are introduced. Moment
connections “partially restrained.”

1950s

High strength bolts replace rivets in moment frame joints. Connections become more compact.
Slip critical connections create “fully restrained” joints.

Early 1960s

Welding in building construction becomes popular. Moment connections become more efficient
as beam flanges can be directly welded to columns without cover plates or clips. Webs are
typically bolted to shear tabs welded to the columns.

Late 1960sEarly 1970s

Semiautomatic Flux Core Arc Welding (FCAW) becomes popular for high speed, high
production work, replacing the slower Shielded Metal Arc Welding (SMAW) in WSMF
construction. Lincoln Electric introduces the NS-3M (E70T-4) flux cored wire.

1980s

Steel use increases as building boom in western U.S. demands more efficient systems for large
structures. Codes requiring costly doubler and continuity plates encourage the use of “jumbo”
column sections with thick webs and flanges.

Mid 1980s

Development of PC based computer-aided design software allows engineers to “tune” the
design of WSMFs. Space frames are eliminated and replaced by plane frames, often only one or
two bays wide, with larger members.

1994

Northridge earthquake highlights defects in WSMF performance at welded beam-to-column
joints. WSMF design in high seismic regions slows as damage is assessed and research is
begun.

Wrought iron, used primarily in the U.S. after the 1850s, has a greater ductility than cast iron
and so could be used more reliably for beams and other bending elements. This permitted the
more efficient construction of frame structures, although the amount of ductility in these
elements was still limited. Taller buildings of wrought iron that were designed to resist lateral
loads—wind was the main consideration—often used exterior infill walls of unreinforced
masonry or hollow clay tile. These acted as shear walls or, more accurately, like braced frames,
with the infill acting as a compression strut between beam-column joints. Diagonal iron bars
were also commonly used to resist lateral loads.
Wrought iron frames probably did develop some moment frame action by virtue of their
connections. It is nearly impossible to create a truly pinned joint that will resist no bending.
Riveting the webs or flanges of the beams to the columns with cleats necessarily added some
rigidity.

2-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

No specific accounts of the earthquake performance of iron frame buildings could be
identified for this report, and it is unlikely that present-day engineers would learn much from
them about modern construction, as iron is no longer used as a primary building material.
2.2

Transition to Steel

Steel was not typically employed in building construction until the last decades of the 19th
century (McGuire, 1988). The first specification for structural steel was published in 1894-95
(FEMA 274, 1997). The lower proportion of carbon in steel versus iron and the use of other
alloys give it better ductility and tensile strength. This made steel an ideal material for frame
structures, and led to the first skyscrapers over 100 feet in height. The metallurgical properties of
steel have continued to change over the past 100 years, with properties such as carbon content,
yield/tensile strength ratios, and elongation specified by ASTM to meet evolving needs.
The transition from masonry bearing wall buildings to steel frames began with the
construction of infilled frame structures, in which lateral loads were primarily resisted by the
masonry. Infill frame buildings are still built in non-seismic regions. The relative ease with
which steel shapes could be rolled permitted the use of more complex beam-column connections
capable of resisting relatively large moments. This allowed beam spans to increase and permitted
the design of true moment frames able to resist cyclic lateral forces in a moderately ductile
manner, without infill.
The usefulness of the moment frame was still limited by the strength of its connections.
Riveted, and later bolted, connections required heavy plates to join the beam flanges and webs to
the column. These plates were usually structural shapes of their own (I-, T- or L-sections). They
often created an eccentric load path between the beams and columns, resulting in high stresses.
Furthermore, the low ductility of rivets limited the overall capacity to absorb the energy of
inelastic cyclic loading. Figure 2-1 shows an example of a riveted connection. Ironically, one
advantage of these connections was that configurations dominated by yielding of connection
clips or angles in bending actually created a semi ductile mechanism, as long as yielding was
kept out of the rivets. Still, moment frames were not typically designed with the intent that the
frames would resist large lateral forces. Infill walls and even concrete fireproofing were still
considered the primary stiffening elements. Engineers designing for lateral wind loads even
through the 1920s considered it “unrealistic and uneconomical—indeed poor engineering—to
disregard” the lateral strength and stiffness of infill (McGuire, 1988).
2.3

Evolution to Moment Frames with Bolted Connections

Between the 1920s and 1950s high strength bolts became an alternative to rivets. Much
stronger than rivets, high-strength bolts are also faster to install. High-strength bolts permitted
very large clamping forces, which led to the development of the slip-critical connection. Slipcritical connections rely on pressure between the mating surfaces, not solely on the shear
strength of the bolt itself. Figure 2-2 shows an example of a combination bolted and riveted
connection.

2-3

 FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

Figure 2-1

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Riveted Beam-Column Connection, Pre-1920s
Source: Preece and Collin, 1991

High strength bolts allowed connections to become smaller. Because of the high clamping
forces, frame joints also became more rigid, reducing distortion of the frames under lateral loads.
The first specification for high strength bolts was available in 1949 (Beedle, 1963), and by 1950
“high strength bolts were being given strong consideration as a replacement for rivets in highrise buildings” (Preece and Collin, 1991). By 1970, riveting was largely discontinued. All-bolted
connections were still somewhat bulky and did not typically achieve a fully restrained
connection even though moment frames were being designed to resist larger lateral forces, and
masonry infill was being relied upon less.
2.4

Welded Moment Frames and the “Pre-Northridge” Connection

The use of welding, while popularized for steel frame construction mainly in the last forty
years, has been in wide use since the 1920s for the mass production of electric motors and other
mechanical equipment (Blodgett, 1963). Its use in building construction allowed smaller and
more efficient beam-to-column connections. No longer were bent plates or structural sections
required to join the members. With welds, the connection could be made purely by fusion of the
materials. First used to attach shear tabs to columns, welds eventually formed the beam flange
connections as well. This allowed for “fully-restrained” joints that reduced mid-span beam
moments under gravity loading and increased building stiffness under lateral loading.
In 1959, AISC researchers studied fully-welded beam-to-column connections under gravity
loading to establish force-deflection relationships for use with plastic design methods (see
Graham et al. in Table 3-1). They noted that the “direct-welded connection” imposes more
severe loads on the column but that it “has certain advantages and may eventually come into
more general use.” By 1963, Beedle described welded moment connections as “familiar” and
“used extensively,” although the accompanying photograph shows a “top plate” detail, not a

2-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

directly welded beam flange (Beedle, 1963). Figures 2-3 and 2-4 show examples of welded
moment connections.

Figure 2-2

Bolted and Riveted Connection, 1930s
Source: Preece and Collin, 1991

Figure 2-3

Welded and Bolted Moment Connection, 1950-1960
Source: Preece and Collin, 1991

2-5

 FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

Figure 2-4

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Welded Moment Connection, 1980s
Source: Preece and Collin, 1991

By the mid-1970s, the standard connection in California WSMFs (SEAOC, 1975) joined the
beam to the column by welding the beam flange and bolting the beam web to a shear tab (Figure
2-4). The alternatives were simply more expensive, as shown in Table 2-2. Elimination of
flange continuity plates and use of a bolted shear tab were proposed as ways to reduce internal
stresses induced by weld cooling and shrinkage. Daniels and Collin (1972) cited this standard
detail as both economical and capable of relieving residual fabrication stresses, but they also
cautioned against the use of untested member sizes and restraint conditions.
Table 2-2
Beam Flange Connection

Relative Costs of Moment Connections

Beam Web Connection

Relative Cost
1979

1983

1986

Full penetration field weld

Bolted to shear tab

1.00

1.00

1.00

Full penetration field weld

Fillet welded to shear tab

1.07

1.07

1.07

Full penetration field weld

Full penetration welded to
column

1.25

1.18

1.25

Bolted flange plate (shop welded
to column)

Bolted to shear tab

2.00

1.75

2.00

Note: The actual cost, represented by the relative cost of 1.00 in one year is not equal to the actual cost, represented
by the relative cost of 1.00 in other years.
Source: Steel Committee of California, 1979, 1983 and 1986

With the popularity of WSMFs on the rise, new welding processes were also developed. The
most common and oldest shielded arc welding process is Shielded Metal Arc Welding. SMAW
uses an electrode “stick,” surrounded by flux, which is melted and fed into the weld area. This
2-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

has been a very reliable welding method and through the 1980s more people were qualified to
perform SMAW than other welding processes (Preece and Collin, 1991).
The self-shielded flux-core process (FCAW-SS), introduced by Lincoln Electric in 1958,
eventually proved itself equally versatile and far less costly than SMAW welding. FCAW uses a
continuous consumable wire electrode fed by machine into the welding “gun” in a
“semiautomatic” process that avoids the starts and stops of stick welding. The nature of flux-core
welding also allows for very high deposition rates. By 1967, high deposition “fast-fill” electrodes
capable of welding in all positions were available (Procedure Handbook, 1973). The faster
FCAW processes were quickly adopted for structural steel erection (Cary, 1970), although most
fabricators continued to use separate electrodes and equipment for flat and vertical welds (Ferch,
2000).
The construction cost savings relative to SMAW welding were substantial. In 1970, Cary
compared SMAW with gas-shielded FCAW and found that SMAW took more than three times
as long to complete a 12-inch vertical weld. A 1973 material and labor cost comparison by
Lincoln (Procedure Handbook) found that FCAW-SS cut the cost of a ¼-inch fillet weld in half
relative to stick welding. A 1997 comparison suggested that FCAW total costs per pound of weld
metal could be as little as one third those of SMAW (Fabricators’ and Erectors’ Guide). Though
economical, the flux-core electrodes most commonly specified for WSMFs in the 1970s and
1980s are now considered to have lacked sufficient fracture resistance (as measured by so-called
“notch toughness”) for reliable seismic performance. The most commonly used electrode, E70T4, never promised any notch toughness in its specifications.
Not all FCAW welding is the same. Self-shielded processes are different from gas-shielded
FCAW, and even within FCAW-SS there are many different electrode classifications and
proprietary products. While some FCAW electrodes provide a specified notch toughness, some
do not. Readers are referred to the FEMA Background Reports (FEMA-288, 1997) and other
SAC reports (Johnson, 2000) for more on weld processes and their application to WSMFs.
Nevertheless, it is fair to say that FCAW-SS E70T-4 welds are found in the vast majority of
field-welded steel moment frames erected in the western United States since 1970 (Goltz and
Weinberg, 1998). Figure 2-5 charts the rise in steel construction over the decades of the
twentieth century. A California building boom in the early 1980s (Seligson and Eguchi, 1999)
put tens of thousands of FCAW-welded-flange, bolted-web connections into service. This
economical and ubiquitous detail is now known as the “pre-Northridge” connection.
Post-Northridge building inspections revealed that, at least in Los Angeles, welded
connections in the 1980s employed a number of practices that may have contributed to poor
performance. Among these were such non-conforming practices as weaving of weld beads, poor
fit-up, and improper weld dams.
Backing bars were commonly left in place after completing the beam-to-column groove
welds. This practice was in compliance with governing codes and standards, including those of
AWS (Ferch, 2000). The practice came under scrutiny after the earthquake because backgouging
of the weld root, which requires removal of the backing bar, has been considered necessary to
2-7

 FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

1990 1993

1980 1989

1970 1979

1960 1969

1950 1959

1940 1949

1930 1939

1920 1929

1910 1919

100
90
80
70
60
50
40
30
20
10
0
1900 1909

MSF of Steel Frame Construction

ensure complete fusion between the beam flange, weld, and column flange (Blodgett, 1963). In
response to Northridge findings, AWS (1995) would later recommend backing removal and
backgouging for WSMFs. FEMA-288 and AWS (1995) offer more on this and related subjects.

Decade

Figure 2-5

Rise in Steel Construction in Los Angeles County
Source: Seligson and Eguchi, 1999

2.5

Optimized Design

In the early 1970s, tall frames were trending toward deep girders of short spans in order to
develop more efficient tube-like behavior. The resulting member and weld sizes had never been
tested, and available codes did not address potential fabrication and erection problems unique to
deep members with thick flanges (Daniels and Collin, 1972).
In the 1980s, expensive detailing discouraged moment connections to the weak axis of wideflange columns (Krawinkler, 1997). Thus, the three-dimensional frame that engaged nearly all of
a building’s columns in both directions was replaced in practice by a number of discrete plane
frames. Elimination of the “space” frame from the Blue Book in 1988 (Zsutty, 1989) sanctioned
the plane frame and encouraged its use.
Steel industry publications encouraged designers to reduce fabrication costs by sizing
members, through careful application of code requirements, so as to avoid continuity plates and
web doubler plates (United States Steel, 1980; Thornton, 1982). In practice, this preference for a
“clean column” resulted in larger column sections with thicker flanges. The resulting frames,
however, were cheaper due to reduced fabrication cost. They were also bigger and stronger, so
fewer were needed for the building to meet overall stiffness requirements.
Thornton (1992) published examples of the relative savings obtained by avoiding doubler
and continuity plates. The effective cost of a doubler and a pair of stiffener plates, including the
erection and fabrication costs, is equivalent to about 700 pounds of steel (material cost only).

2-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

Even if an average column needed to be about 35% heavier to avoid these plates, the net savings
for an average building was demonstrated to be on the order of $1,000 per column.
In extreme cases, these developments led to an “optimal” configuration of one- or two-bay
plane frames, with deep beams and heavy columns, spaced around the building perimeter. Given
the changes in standard material strengths, weld properties, member sizes, and frame
configuration, a typical 1990 WSMF should not be expected to provide the same seismic
performance as a 1970 WSMF (Krawinkler, 1997).
Changing trends in building massing also affected WSMF design. As post-Northridge
building surveys reveal (see the Appendices and Northridge sections of this report), most of the
WSMFs built in greater Los Angeles during the 1980s were not skyscrapers but three- to fivestory office buildings. Garreau (1991) described the “laws” that governed design of these semiurban structures: heights to optimize floor area ratio relative to the cost of a deep foundation,
floor plates to maximize capacity of a single central core and to facilitate corporate management,
and setbacks to maximize corner offices. More atria, corners, and setbacks mean less opportunity
for an uninterrupted space frame. Figure 2-6 shows an example of typical 1980s WSMF
architecture.
Discussion of advancement in steel frame design must consider concurrent advancements in
the technology that allowed engineers to analyze increasingly complex structural systems with
continually greater speed. The personal computer, introduced in the early 1980s, was common in
larger design offices by the end of that decade, and ubiquitous by the mid-1990s. Software was
developed to suit the new computers (Habibullah, 1984; Wilson, 1984). Pre-processors allowed
rapid creation of two- and three-dimensional models of unprecedented complexity. Postprocessors summarized the results and checked code requirements. Alternative designs could be
studied and refined within hours, not days. The result was a design optimized for code
compliance, but not necessarily for performance. The fine-tuned frames met stress and deflection
requirements with a minimum of wasted material or extra cost.
But the software did not account for fabrication or construction sequences. Nor did it analyze
connections. Even today, analysis programs typically neglect connection details—and not just in
steel. Interestingly, this has been the case throughout the history of steel frame analysis and
design. From simplified cantilever and portal methods through the plastic design methods
standardized in the 1960s, emphasis has always been more on the behavior and proportioning of
members than on the essential connections (McGuire, 1988). Steel frame connection sizing and
detailing in most of the United States is still left to the fabricator. In California and some other
areas, engineers have shown connection details on their own drawings for at least as long as the
WSMF has been a viable structural system. But even in California, the WSMF connection was
essentially prescriptive by the mid-1970s (SEAOC, 1974), so it is not surprising that engineers
and programmers focused on the member and assumed the connection would work.

2-9

 FEMA-355E
Chapter 2: Design and Construction
of WSMFs in Seismic Areas

Figure 2-6

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

WSMF Building with Setbacks
Source: Engineering News Record, February 23, 1998.

2-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

3. TESTING OF STEEL MOMENT-FRAME CONNECTIONS
Over the past thirty years, academic and industry researchers have tested a variety of welded
steel moment frame connections to gauge their expected performance under cyclic lateral loads.
These efforts have involved both physical testing of beam-column joints and theoretical
analytical modeling. This research has been instrumental to the advancement of the WSMF,
discussed in the previous section, and to the development of building codes, summarized in the
following section.
Much of the research showed the potential pitfalls of welded construction in severe
earthquake conditions. Interestingly, the observed failures often were considered insignificant
aberrations, or were addressed in the authors’ conclusions as avoidable simply through proper
field inspection. From Popov’s work in 1969 to Engelhardt’s in 1993, fractures at welded joints
have been shown to limit the ductility of steel moment frames. This summary and review is
intended not to criticize the historical research but to draw lessons from the overall pattern that
may not have been apparent from individual studies. Indeed, when the first U.S. code provisions
for WSMFs were being written in the late 1960s (see the next section of this report), there was
evidence that considerable weld defects could be tolerated with no loss of capacity: “[B]uckling
is more of a problem than weld defects are in plastic design” (Couch and Olsson, 1965).
The test programs described below involved over a hundred individual beam-column
specimens tested over thirty years. With a few exceptions, the published papers reached
encouraging conclusions about the expected seismic performance of WSMF beam-column
connections. And when caveats or limitations were stated, they read mostly like legal
disclaimers, cautioning readers against careless extrapolation. The leading California steel
researcher of the time, Egor Popov, would later regret not speaking up: “My flaw was that I
wasn’t sufficiently loudmouthed about how bad they were” (ENR, February 10, 1997).
In 1993, Engelhardt and Husain took advantage of broad hindsight and sounded a warning.
The Northridge earthquake struck a month after their article was published.
This brief description represents most of the important WSMF testing done prior to the
Northridge earthquake. Taken as a whole, and viewed from a post-Northridge perspective, the
historical testing offers some broad conclusions perhaps not obvious from any of the individual
studies:
•

Performance can not be usefully gauged unless demand is well-defined. This is an obvious
statement today, now that codes and standards explicitly acknowledge inelastic seismic
demands, and now that tools to model those demands are available. When the first of these
tests were being performed, practicing engineers did not work in inelastic terms.

•

The beam-column weld and the weld access hole have always been critical locations for
fracture initiation in the pre-Northridge connection. Many of the brittle fracture patterns
observed after Northridge had been seen before in the laboratory.

3-1

 FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

•

Researchers and practitioners have consistently attributed failures in the lab to poor
construction quality. Viewed with hindsight, the low level of quality appears to be systemic.
Fracture of welded joints should perhaps have been a focus of detailed study on its own.

•

High reliability was never attained. Variable results were the rule, and premature brittle
fractures accounted for some portion of nearly every test program. In statistical terms, a small
program can not demonstrate the kind of reliability that engineers expect.

•

Engineers extrapolated unreasonably from test results. None of the pre-Northridge research
programs tested beams deeper than 24 inches or heavier than 76 lbs/ft, but many of the
fractures found after Northridge involved W30 and W36 sections up to 300 lbs/ft. Design
“beyond the precedent established by research” was identified as a concern as early as 1972
(Daniels and Collin, 1972).

Figure 3-1 shows an example of a WSMF connection test setup. Table 3-1 lists significant
research programs that preceded the 1994 Northridge earthquake.

Figure 3-1

WSMF Connection Test Setup
Source: Popov et al., 1985.

3-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 3-1
Ref.
Date

Researcher

FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Milestones in Research on WSMF Connections

Research
sponsors

# of
tests

Description of tests and results

1959

Graham,
Sherbourne,
and Khabbaz

AISC

13

Monotonic gravity tests on fully-welded specimens with
16WF beams and 8WF or 12WF columns, with emphasis
on moment-rotation curves. Did not address lateral
loading. Overall program also included 11 direct pull tests
to simulate the tension beam flange pulling on the column
flange. In all of these, yielding progressed into the column
web, and column flanges were visibly bent before cracks
developed at the mid-length of the butt weld. As late as the
1989 Ninth Edition, the AISC Specification cited these
tests (and Popov and Pinkney, 1969) in support of welded
beam flanges.

1965

Bouwkamp &
Clough

AISI

NA

In-situ vibration studies of actual steel moment frame
buildings to calibrate actual and theoretical period and
damping calculations.

Popov and
Franklin

AISI

4

Tested beam-column joints with welded flanges or flange
plates (8WF20 beams). All showed good ductility with
ultimate failure of welded flanges in weld.

Beedle

not reported

1

Constructed and tested three story mockup under two
cycles of reverse loading.

1965

Bertero and
Popov

NSF

10

Tested the ductility and fatigue resistance of 4-inch deep
beams without beam-column connections. Identified
ductility ratios at beam flange buckling and at beam web
tearing. Led to recommendations for compact section
requirements.

1969

Popov and
Pinkney

AISI

24

Static cyclic tests on a variety of beam-column joints
(8WF20 beams). Most achieved good ductility, but plastic
rotations were not calculated. Failures in welds at beamcolumn joints were observed. Warning made to provide
high quality joints to avoid premature weld failures. As
late as the 1989 Ninth Edition, the AISC Specification
cited these tests in support of welded beam flanges.

1972

Bertero, Popov
and
Krawinkler

AISI

2

Tests focused on panel zone deformation (8WF column,
10B and 14B beams). Results led to code change to
require stiff panel zones. Noted that excessive panel zone
deformations can lead to kinking at beam flanges and
subsequent weld failures.

3-3

 FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Table 3-1
1972

1973

Popov and
Stephen

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Milestones in Research on WSMF Connections (continued)
AISI, Hyatt
Corp.

8

Evaluated plastic rotation ductility in beam-column joints
(W12 column, W18 and W24 beams) with FCAW E70T-4
welds. All specimens achieved plastic rotation of .02 to .06
radians, but four of five with bolted webs failed with
abrupt fracture. Authors remark on the overall excellent
ductility of the WSMF connection.

Popov and
Bertero

1985

Popov, Amin,
Louie, and
Stephen

Norland
Properties,
Trade/Arbed,
Herrick, SOM

8

Tested one-half size mockups of joints designed for 47story building with emphasis on panel zone behavior (W18
beams). Of eight tests only two did not fail in a brittle
manner and only three exceeded 2% plastic rotation.
Authors conclude that the adequacy of the connection has
been validated.

1987

Popov and
Tsai

NSF, AISI

18

Tests evaluated a number of connection details (W18 and
W21 beams). Of the eight specimens typical of preNorthridge practice, only four achieved beam plastic
rotations greater than .01 radians. Two worst performers
used E70T-4 electrodes made by an inexperienced welder.
Authors note that weld and fabrication quality are
important factors in overall performance.

1988

Tsai and
Popov

1991

Anderson and
Linderman

NSF, California
Field
Ironworkers
Trust

15

Focus on repair of nominal weld cracks associated with
expected ductile performance. Seven initial specimens
tested to failure and eight tests of various repairs (W16
beams). Initial specimens all failed at less than .03 radians
total (elastic + plastic) rotation, several with Northridgetype fractures.

1993

Schneider,
Roeder and
Carpenter

NSF

5

Tested weak column-strong girder joints (W12 and W14
beams). Concluded that high ductility can be achieved and
that the anticipated excellent performance of WSMFs is
justified.

1993

Engelhardt and
Husain

Steel Comm. of
CA, Nucor and
Bethlehem
Steel

8

Focus on supplementary web welds with W21 and W24
beams. Only one of eight tests exceeded 1.5% plastic
rotation, with failures initiating in bottom flange welds.
Authors conclude that large variability in performance was
of much greater concern than the web weld issue.

3.1

Early Testing

Monotonic tests of steel beam-column connections date back to 1917 (McGuire, 1988). Table
3-1 describes similar tests with welded specimens from 1959. Beedle (1963) described a number
of monotonic tests of welded, bolted T-stub, and end-plate beam-to-column moment
connections. Photographs of the welded specimens indicate fully welded webs and monotonic
loading that put the beam bottom flange in compression. Wind loads, but not cyclic or seismic
demands, were mentioned. Beedle concluded that steel moment connections of different types
could all be expected to provide a ductility factor of 8 to 10. If not, “it is because some detail has
been underdesigned” (Beedle, 1963).

3-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Dynamic and cyclic behavior of frame structures was a topic of research at least as far back
as 1938, when a shake table test was performed to model a 16-story steel frame and masonry
infill building in San Francisco (Jacobsen and Ayre, 1938). The tests were performed to
determine the elastic modal properties of the structure, not to predict inelastic performance. The
complex model is shown in Figure 3-2.

Figure 3-2

Testing Apparatus for Modeling 16-Story Infilled Frame Building
Source: Jacobsen and Ayre, 1938

3.2

Bouwkamp and Clough; Popov and Franklin; Beedle (1965)

These three papers were cited as the basis for the ductile steel-frame provisions introduced in
the 1968 Blue Book, and later adopted for the 1970 Uniform Building Code. The tests were
intended to evaluate steel moment frames under cyclic loading. Bouwkamp and Clough
subjected a real building in various stages of construction to induced sinusoidal vibrations to
determine the fundamental period and critical damping ratio. Popov and Franklin tested four
beam column assemblies, one with girder flanges groove welded to the column flange. The
sections were small: 8WF48 columns and 8WF20 beams. Test results led the authors to
conclude, “it is possible to expect strains on the order of 1-½% in extreme cases, which
corresponds to a fiber ductility factor µ of 12” (Popov and Franklin). Plastic deformation was
measured in strain not rotation. All four specimens developed bottom flange local buckling. The
authors concluded that the hysteresis loops produced by the tests were remarkably stable and that
these connections can be depended on to achieve a “practically constant” amount of strain energy
absorption.
Beedle tested a full height two-bay, three-story frame with light beam and column sections
(6WF25 columns and 12B16.5 beams). He applied vertical loads to simulate actual gravity
loading, then applied horizontal forces to simulate earthquake motion. He was able to apply one
3-5

 FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

or two cycles of reverse loading to obtain a hysteresis curve for the frame. The goal of the study
was to see how well the actual stress-strain relationships compared with theoretical modeling.
The frames were pushed until the beam compression flanges buckled. The author noted that in
the second cycle of loading, the energy absorption increased due possibly to strain hardening and
kinematic effects. No study was done to determine the rotation capacity of the connections.
3.3

Popov and Pinkney (1969)

Beginning in 1965, Popov and Pinkney tested 24 specimens, evaluating the total energy
absorbed by the system under cyclic, quasi-static loading and measuring the ductility of the
elements and joints. Their interest was in “the manner of failure due to exceptionally high
loads… and the longevity of a connection under substantial overloads.”
The 8WF20 beam specimens were able to achieve plastic rotations between .046 and .069
radians before failure (Popov and Bertero, 1973). No demand estimates were available for
comparison with the measured capacities. The authors acknowledged this, but nevertheless
concluded that “such an assemblage is very reliable and can be counted upon to absorb a definite
amount of energy in each cycle for a prescribed displacement.” They further estimated that “the
number of repeated and reversed loadings which can be safely sustained appears to be in excess
of that which may be anticipated in actual service.”
Although the tests showed substantial inelastic capacity, descriptions of the ultimate failure
modes are interesting from a post-Northridge perspective. After repeated inelastic excursions,
“fracture was frequently in or near the welds, with several failures occurring in the groove welds
of the flanges to the column face….” In addition “sharp cornered web copes were a recurring
source for initiation of web cracks.”
Popov and Pinkney offered two important warnings. First, “the importance of careful
inspection during fabrication was brought out by the premature failure of two improperly welded
connections,” and second, “extrapolation to members with … cross sections [other than 8WF20]
must be done with caution.” These can be seen as prophetic now that Northridge-type fractures
have been associated with poor inspection (Paret, 1998; Kaufmann et al., 1997) and deep beam
sections (Roeder and Foutch, 1996; Bonowitz, 1999a), although similar results were later
achieved with W18 and W24 beams (Popov and Bertero, 1973).
Also interesting is the authors’ note that in specimens found to have been fabricated with
“poor workmanship” and inspected by ultrasonics before testing, “indications produced by the
unwelded contact surface were mistakenly interpreted as being due to the back-up bars.” Though
noted in 1969, the ambiguity and inconsistency of ultrasonic testing has now been demonstrated
by Northridge data (Paret, 1999).
3.4

Bertero, Popov, and Krawinkler (1972)

Through the early 1970s, there were no special requirements in the code for the strength of
panel zones. In theory, large panel zone deformations could reduce the gravity load capacity of
the column. Bertero, Popov, and Krawinkler tested two subassemblies, one with a panel zone
that was relatively weak compared to the beams, and one with a panel zone relatively strong. In
3-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

general, the authors concluded that “the energy absorption and dissipation capacity… exceeds
the required energy, even for cases of extreme earthquakes.”
In the latter test “the weak parts of these specimens were clearly the regions of plastic hinges
in the beams. This required high rotation capacities of the beams.” This is an important
statement about ductility demand. In the weak panel zone test, lateral torsional buckling of the
beams “prevented a complete stabilization of the hysteresis loops,” and testing did not reach
failure of either the beam flanges or the welds. (In practice, beams are typically restrained
against lateral torsional buckling by a floor slab or by other members.)
An interesting note by the authors was that excessive panel zone deformation led to
“kinking” of the column flanges which “precipitated failure of the beam flanges at the welds.”
The authors therefore recommended strong panel zones “designed for the real plastic capacity of
the beams.” Recent finite element studies have reached similar conclusions (El-Tawil et al.,
1999).
3.5

Popov and Stephen (1972), Popov and Bertero (1973)

In 1973, Popov and Bertero revisited the issue of beam plastic hinging as a desirable
mechanism for ductile frames. (Their paper is based on the same test program as Popov and
Stephen, 1972.) Here, specific mention of the welding procedures was made. “For all flanges
full penetration welds with backup bars were used. Beam webs were coped…. The flux-cored
wire E70T-4 was used throughout. All of the structural welds were sonic tested.” This is
notable considering the current findings that electrodes with no notch toughness requirements
(such as E70T-4) perform poorly in pre-Northridge connections (Bonowitz, 1999a).
Popov and Bertero, however, make only passing mention of the possibility of weld failures.
Of the seven tests, two “exhibited superior ductile behavior,” but “three of the hybrid
connections failed prior to the completion of the large hysteresis loops.” Failures were typically
in the welds. Still, several times the authors describe the “remarkable stability” of the joints. One
would assume this describes the joints that did not fail abruptly through the welds. The beam
sizes used in the tests ranged from W8x20 to W24x76 sections. The authors noted that there was
a “close correlation over so wide a range of beam sizes.” Later work would show, however, that
similar performance should not be expected over a wide range of member sizes (Roeder and
Foutch, 1996; Bonowitz, 1999a).
3.6

Popov, Amin, Louie, and Stephen (1985)

These tests were performed to “verify the design criteria for beam-column joints under
extreme seismic loading conditions for a 47-story building in San Francisco.” Although the test
specimens involved W18 beams, deeper than most specimens previously studied, the building
was to use 36-inch deep members.
The authors describe the welding procedure: “the back-up plates for the welds on the beam
flange-to-column flange connections were removed after the full-penetration flange welding was
completed and small cosmetic welds appeared to have been added and ground off on the
underside.” It is interesting that the authors considered the fillet welds to be “cosmetic” and not
3-7

 FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

placed to improve performance. While this may have been the case, it is noteworthy because of
the post-Northridge recommendations to remove backup bars and to grind and reweld the roots.
Of the eight tests, one failed apparently due to an obvious defect in the preparation. Of the
remaining seven, five developed some panel zone inelasticity, then failed abruptly in the welds
or in the heat affected zone of the beam. Only two specimens remained ductile through the end
of the test.
The authors set an acceptability criterion in terms of total plastic rotation (i.e. beam and panel
zone contributions combined), aiming for an “essential” capacity “of at least 2% times a
reasonable factor of safety.” Of the seven tests, three reached total plastic rotations in excess of
5%. The other four failed at an average of 1.7%. Yet, the authors concluded that “the objectives
of verifying the design criteria for the prototype were achieved by this experimental and
analytical program.” Interestingly, back in 1969, Popov and Pinkney had cautioned that it is
important when drawing conclusions about testing to have a “statistically valid number of
experiments, with as nearly identical as possible input parameters.”
3.7

Popov (1987)

In 1987, Popov reviewed “the state-of-the-art for the design of steel moment connections…
for regions of high seismic risk.” Looking back over his tests from the early 1970s, he called
attention to the “explosive flange failures” that ended many of the tests, and noted that they
occurred “only after a number of large cyclic load reversals.” Those specimens fractured at
plastic rotations of about .02 radians; post-Northridge standards consider this an unacceptable
rotation capacity. Despite his sanguine assessment of past tests, Popov would conclude his 1987
review by plainly acknowledging the unknown: “Due to the great uncertainty of the forces that a
structure may have to resist during an earthquake, complete reliance on the minimum code
provisions is hazardous.”
3.8

Popov and Tsai (1987); Tsai and Popov (1988)

In the first paper to offer a strongly worded warning about premature connection fracture,
Popov and Tsai presented a wide range of new specimens and evaluated the ductility of WSMFs
relative to code provisions. (The 1987 conference paper is a summary of the full research that
would be published in 1988.)
The authors noted that the ductility of the various tests was “erratic,” with only about half the
tests developing “satisfactory” inelastic performance. They concluded that “weld fractures at
connections are particularly dangerous.” They suggested that careful inspection and fabrication,
especially at weld access holes, could reduce the variability of performance.
Instead of discounting a few poorly fabricated specimens, Popov and Tsai also drew attention
for the first time to the real implications of welding quality control. Two of their specimens
failed, they noted, because the fabricator was unfamiliar with flux-core welding. Another FCAW
specimen “was fabricated with exceptional care [that] cannot necessarily be duplicated in the
field.” (Note that the 1987 paper is incorrect in describing these three specimens as having
SMAW welds; the welds were FCAW-SS.)
3-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

3.9

FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Popov, Tsai, and Engelhardt (1988)

This study compared the plastic rotation capacities of the 18 Tsai and Popov (1988)
specimens to theoretical demands. The researchers analyzed a hypothetical six-story building
subjected to the Parkfield, Mexico City, and El Centro ground motions. Parkfield and Mexico
City each imposed plastic rotation demands of about 1.5%. Of the 18 tests, only seven had
capacities exceeding this demand by at least 10%. The other eleven either just barely reached
1.5% or failed at under 1% rotation. Six of the eleven were considered to have been poorly
fabricated by welders unfamiliar with the specified welding process. The authors conclude that
“the experimental data for these connections shows that well fabricated connections are
adequate.”
The study also concluded that plastic rotation demands in the beams could be notably
reduced by making the panel zones more flexible. They cautioned, however, that excessive
inelastic panel zone shear deformation and kinking may lead to column instability, and to
damage to the beam-to-column groove welds. This latter concern is similar to earlier comments
by Bertero, Popov, and Krawinkler (1972).
3.10

Anderson and Linderman (1991)

This study is noteworthy because it explicitly acknowledges that WSMF connections will
develop cracks even when they perform as intended. Despite the record of marginal performance
through the 1980s, however, brittle fractures were not anticipated. Only a .02 radians total
(elastic plus inelastic) rotation demand was imposed. Three of the seven initial specimens barely
reached that rotation, and several developed full-width weld or beam flange fractures before the
onset of local flange buckling.
3.11

Schneider, Roeder, and Carpenter (1993)

The authors tested four weak column-strong beam joints using very light elements: W12x2630 columns and W12x16 to W14x26 beams. They concluded that the joints “sustained a major
earthquake and exhibited a tremendous amount of ductility. Special moment-resisting steel
frames are highly regarded for their seismic performance and the results from these four tests
justify this reputation.” This enthusiastic endorsement of the WSMF less than a year before the
Northridge earthquake reflects little concern even among academics regarding the likely
performance of WSMFs.
3.12

Engelhardt and Husain (1993)

Published one month before the Northridge earthquake, this critical paper represented at least
two years of study into the seismic performance of WSMFs. The authors tested eight specimens
with relatively large members: W12x136 and W18x60 columns, and W21x57 and W24x55
beams. Their primary goal was to evaluate supplemental shear tab welds. Self-shielded flux-core
welding was used because it “is frequently used in actual field welding for this connection
detail.” (E70T-7 electrodes were used. Like E70T-4, the T-7 electrode has no minimum specified
notch toughness.) Welds were made in a manner to simulate actual field conditions. Backup bars
remained in place.
3-9

 FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

A plastic rotation of at least 1.5% was sought for each of the tests “as a reasonable estimate
of beam plastic rotation demand in steel moment-resisting frames subject to severe earthquakes.”
Only one of the eight tests reached 1.5% rotation, and all failed in a “fracture at or near a beam
flange groove weld.” This failure occurred at the beams’ bottom flanges. The failures “generally
initiated at the edge of the beam flange and gradually spread across the width of the flange as the
loading progressed.” Once failure occurred, the beam was deflected in the opposite direction to
yield the top flange. In all of the tests, the top flange failure occurred not in the weld but in the
flange itself, outside the heat affected zone.
The test matrix considered various web connection details and Zf/Z ratios, but the authors
were forced to conclude that “variability in the performance of the beam flange welds appears to
have had a much greater influence on plastic rotation capacity than Zf/Z ratio or web-connection
detail.” The paper concludes with “concerns about the welded flange-bolted web detail for severe
seismic applications” and calls for “a careful review of design and detailing practices.”
In a wide-ranging February, 1993 presentation, Popov noted briefly that the work of
Engelhardt and Husain “raises the question of reliability of field flange welds” (Popov et al.,
1993). He then showed a number of possible reinforced details (with various cover plates, wing
plates, and ribs) “for situations where large rotations are anticipated.”
Suspecting that incomplete weld fusion undetected by ultrasonic testing might be causing the
premature fractures, Engelhardt and Husain also called for review of welding and quality control
issues. This suggestion, which appears to have been first made by the authors in a 1991 paper,
prompted some criticism by a prominent welding expert. “The assumption that the [flawed]
fabrication and inspection of the test specimens was typical of the state of the art in present day
structural steel construction is wrong and very much out of line” (Collin, 1992). The events of
January, 1994 would show that Engelhardt and Husain had identified problems that were indeed
common in current construction (Paret, 1999).
3.13

Roeder and Foutch (1996)

In the wake of Northridge, Roeder and Foutch compiled and analyzed results from various
test programs, including most of those summarized above. Using a consistently defined Flexural
Ductility Ratio, they found that the inelastic capacity of 91 comparable specimens was highly
scattered. Nevertheless, they identified a useful inverse relationship between beam depth and
expected ductility. The deeper the section, the lower the inelastic capacity. (Bonowitz, 1999a,
found the same relationship when the older results were removed and tests done after Northridge
were included.)
FEMA 273 (1997) has incorporated this relationship for use in evaluating existing structures.
The implications are profound. Despite the fact that testing from the previous two decades had
rarely used beam sections deeper than a W18, many of California’s “optimized” steel frames
built in the 1980s and early 1990s employed W30 and even W36 beam sections.

3-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

3.14

FEMA-355E
Chapter 3: Testing of
Steel Moment-Frame Connections

Connection Testing Since Northridge

As described later in this report, codes and standards changed after Northridge to require
designs based on comparable cyclic test results. In the six years between the earthquake and
publication of the SAC Guidelines, hundreds of full-scale beam-column subassemblies have
been tested both in academic studies and in qualification tests for specific building designs. The
SAC Joint Venture has compiled over 500 recent test results, many privately funded.
Some of the recent tests involved pre-Northridge details, either as benchmark specimens or
as initial specimens to be later repaired and reloaded. Database summaries from late 1998 and
mid-1999 confirmed that deeper beams have less capacity and that notch-tough weld metal
significantly improves performance of the pre-Northridge detail (Bonowitz, 1999a and 1999b).

3-11

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

4. CODES AND STANDARDS FOR STEEL MOMENT FRAMES
After the 1906 earthquake, San Francisco adopted a 30 pounds per square foot lateral design
load for new buildings. The new requirement was intended to account for both wind and
earthquake effects (SEAOC, 1968). This was probably the first quantified seismic code provision
in the world, even if it accounted only indirectly for a building’s key dynamic properties: mass
and stiffness. Following the 1925 Santa Barbara earthquake, engineers began to focus on the
complex interaction of parameters that affect building performance: structural system and
material, period of vibration, soil conditions, etc.
With each earthquake, building codes progress. The observed performance of real
buildings—especially poor performance—can have a profound impact on provisions for
structural materials and systems. Though changes are sometimes written and adopted slowly
even after earthquakes, they frequently take effect before thorough investigations are complete.
For steel moment frames, it was more the lack of earthquake damage data that propelled the
standards for their design. Until Northridge, WSMF buildings simply did not produce the
multiple and repeated failures that force building codes to change. As shown in the next section
of this report, that was as much due to their absence as anything else. But without notable
failures, seismic code provisions for steel frames developed incrementally, and almost always in
ways that would encourage and broaden their use. As a result, WSMF design practice was
shaped more by design and construction feasibility than by code limitations.
From a post-Northridge perspective, a review of code provisions and standards for steel
frames offers the following lessons:
•

WSMF code provisions have developed incrementally, based largely on specific academic
research. Since welding became feasible for building structures in the 1960s, provisions have
been adjusted to reflect the latest test results.

•

While code provisions have been based on research, they have not kept engineers from
extrapolating specific research results to untested conditions.

•

The lack of real data on the seismic response of WSMFs was perhaps misinterpreted by code
writers as evidence of “excellent” performance. This may have contributed to the code’s
preference for steel moment frame construction.

The preface to the 1927 Uniform Building Code thanked the individuals and organizations
who contributed to its development: building inspectors, contractors, and engineers, with
suggestions from “sales engineers for building materials” (Freeman, 1932). Since then, seismic
codes in the U.S. have been written largely by practicing engineers, academics, and building
officials who volunteer their time and expertise. And since then, code writers have been assisted
by vendors and industry representatives. Proponents of all structural materials, not just steel,
have contributed to code development efforts by sponsoring important research and by
participating on code-writing committees. As one would expect, these professionals, often

4-1

 FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

engineers, researchers, or contractors themselves, express preferences for their own products and
innovations, and support their positions with research results.
A study of the many influences on code provisions is beyond the scope of this report. The
building code is, after all, a public policy document. It suffices to note here only that technical,
political, and financial interests have sometimes been complementary, sometimes competitive.
Code provisions are a synthesis of these interests and frequently represent a series of unavoidable
compromises.
Tables 4-1 and 4-2 summarize some of the more important milestones in code development,
with emphasis on pre-Northridge steel moment frames. Since 1961, the UBC provisions listed in
Table 4-2 were based on the SEAOC Blue Book; Blue Book developments are described in the
text following the Tables.
Table 4-1
Date

Milestones in Code Development for Steel Moment Frames
Milestone in Code Development

1906

In response to 1906 earthquake, multistory buildings in San Francisco must be designed for 30 psf
lateral (wind and seismic) loads.

1927

First seismic provisions written into the UBC as an appendix. Seismic loads a function of mass and soil
profile only.

1933

In response to 1993 Long Beach earthquake, California passes the Field Act, regulating the design of
certain state buildings including schools. California passes the Riley Act, specifying base shear as a
function of both soil and building height.

1948,
1950s

UBC incorporates the K factor to differentiate between buildings and other structures. Base shear
becomes a function of period.

1959

SEAOC publishes its first Recommended Lateral Force Provisions (Blue Book). After 1961, UBC
adopts Blue Book recommendations directly into the Code. K factor refined to be a function of building
material and structural system. Section j favors steel moment-resisting frames.

1968

In response to perceived excellent performance of steel structures, Code defines special K factor for
ductile moment frames. For steel, defines properties of a ductile frame, including material
specifications, strength of girder-column connection, and an inspection program for complete
penetration girder welds.

1975

Blue Book further defines characteristics of ductile steel moment frames. Introduces panel zone and
continuity plate requirements. Recommends web welds for better performance of girder flange welds.

19851988

UBC requires that webs be welded if Zf /Z ratio is less than 0.7. Relaxes panel zone requirements to
permit yielding, thereby reducing girder stresses. Requires strong column-weak beam design in most
frames. The “prequalified” moment frame connection is included in the UBC; alternative details require
design for 125% of the girder flexural strength.

19891993

UBC moves toward strength design in most materials. Lateral force equations are changed to use an Rw
factor in place of K factor.

1994

In response to Northridge earthquake, ICBO enacts emergency code change requiring cyclic testing of
moment frame joint designs.

4-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 4-2
UBC

Uniform Building Code Provisions for Steel Moment Frame Buildings

Steel Frame Designation

1949,
1952,
1955,
1958

No special designation.
Appendix 2312b: F=CW, C
constant for all building and
bracing types

1961

2313b: Space Frame – Moment
Resisting. May or may not be
enclosed by rigid elements that
would prevent sidesway.
Table 23-F: K=.8 (frame resists
25%), or .67 (frame resists100%)
2313f: Story drift limits per
“accepted engineering practice.”
2313j: Buildings taller than 160’
must have complete MRSF.
Same as 1961, but moved to
Section 2314.
2314b: Space Frame – Ductile,
Moment-Resisting.
Table 23-H: K=.8 (DMRSF in
“dual bracing system”) or .67
(DMRSF alone)
2314f: Same as 1961.
2314j1: Buildings taller than 160’
or with K=.67 or .8 must have
DMRSF “of structural steel
(complying with Chapter 27) or
reinforced concrete (complying
with Section 2630 …).”

1964
1967

FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

Steel Frame Detail Provisions
Appendix 2312d: Plans must include floor load assumptions, “a
brief description of the bracing system used, the manner in which
the designer expects such system to act, and a clear statement of
any assumptions used” including assumed inflection points, and a
sample bent calculation.
2313j: “The [moment resisting space] frame shall be made of a
ductile or a ductile combination of materials. The necessary
ductility shall be considered to be provided by a steel frame with
moment-resistant connections or by other systems proven by tests
and studies to provide equivalent energy absorption.”

Same as 1961, but steel chapter is more detailed.
2314j1: DMRSF “may be enclosed by or adjoined by more rigid
elements which would tend to prevent the space frame from
resisting lateral forces where it can be shown that the action or
failure of the more rigid elements will not impair the vertical and
lateral load-resisting ability of the space frame.”
2314j2: “The necessary ductility for a ductile moment-resisting
space frame shall be provided by a frame of structural steel
conforming to ASTM A7, A36, or A441 with moment-resisting
connections.”
Chapter 27: Nothing system-specific or K-related.

4-3

 FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

Table 4-2
UBC

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Uniform Building Code Provisions for Steel Moment Frame Buildings
(continued)

Steel Frame Designation

1970,
1973

2314b: Same as 1967.
2314f: Same as 1961.
2314j1: Same as 1967, but steel
DMRSF to comply with section
2722.

1976,
1979,
1982,
1985

2312b: Ductile MomentResisting Space Frame, similar
to 1967, Section 2314b.
2312h: Story drift (using K =
1.0) limited to 0.005 times story
height.
2312j1: Ductility requirements,
similar to 1970, Section 2314j1.

Steel Frame Detail Provisions
2314j2: “The necessary ductility for a ductile moment-resisting
space frame shall be provided by a frame of structural steel with
moment-resisting connections (complying with Section 2722 …).”
2722: Steel Ductile Moment-Resisting Space Frames, Seismic
Zones No. 2 and No. 3 (Note: the most severe seismic zone at this
time was Zone 3.)
a: Welding to comply with UBC Std 27-6.
b: Defines joint and connection.
c: Defines materials.
d: Connections. “Each beam or girder moment connection to a
column shall be capable of developing in the beam the full plastic
capacity of the beam or girder.” Exception granted if “adequately
ductile joint displacement capacity is provided.”
Also, if Fu/Fy<1.5, no plastic hinges allowed where beam flange
area is reduced by bolt holes. (Note: A36 would not be subject to
this limitation, but A572Gr50 would.)
e: Local buckling (b/t limits)
f: Slenderness ratios
g: Nondestructive Weld Testing. “Tension butt welded connections
between primary members of the ductile moment-resisting space
frame shall be tested by nondestructive methods for compliance
with UBC Std No 27-6 and job specifications. A program for this
testing shall be established by the person responsible for structural
design.”
2312j1F, similar to 1970, 2314j2.
2722: Same as 1970, but slenderness provision removed, and
testing specified:
f: Nondestructive Testing. Subject to special inspection per 305.
“All complete penetration groove welds contained in joints and
splices shall be tested 100 percent either by ultrasonic testing or by
radiography.” For an individual welder, test rate may be dropped to
25% if defect rate after 40 welds is 5% or less.
Also, base metal thicker than 1.5 inches subject to throughthickness weld shrinkage strains shall be ultrasonically tested after
welding.

4-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 4-2
UBC

FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

Uniform Building Code Provisions for Steel Moment Frame Buildings
(continued)

Steel Frame Designation

Steel Frame Detail Provisions

1988

2312b: Special momentresisting space frame (SMRSF)
Table 23-O: Rw=12
2312e.8: Story drift limited to
story height times:
0.0033, for buildings up to 65 ft
tall
0.0025, for taller buildings.

1991

2331: Special moment-resisting
frame (SMRF)
Similar to 1988.

2722f: SMRSF Requirements
1: Scope
2: Girder to Column Connection
Required strength: the strength of the girder in flexure, or the
moment corresponding to development of the panel zone per
equation 22-1.
Connection strength: “The girder-to-column connection may be
considered to be adequate to develop the flexural strength of the
girder if … the flanges have full penetration butt welds to the
columns” and the web connection can resist gravity plus seismic
shear at the required flexural strength. Supplemental web welds
may be required. Alternative details must be designed for 125% of
the girder flexural strength.
3: Panel zone must resist gravity plus 1.85 times prescribed seismic
forces, but need not exceed 80% of the combined strength of the
girders.
4: Flange b/t must be less than 52/√Fy.
5: Continuity plates to resist flange force of 1.8btFy.
6: Strong column requirement.
2710g: SMRF Requirements, similar to 1988 2722f.

4.1

1906-1924

Prior to the 1906 San Francisco earthquake, no quantified seismic provisions are known to
have existed within building codes. Tall structures were required to meet wind criteria, so some
measure of lateral force resistance was provided by most codes. Lateral forces were typically
resisted by unreinforced masonry shear wall elements, or by steel girts and braces. After the
earthquake, San Francisco placed in its building code a provision that structures be designed to
resist earthquake forces as well as wind. The provisions called for the capability to resist 30 psf
of lateral pressure (SEAOC, 1968). The provision for earthquakes was only indirectly dependent
on building mass as a function of surface area. There were rules governing the design of
materials, but no provisions for steel frame design with specific reference to cyclical demands
imposed by earthquakes.
Between 1906 and 1925, engineers had begun to understand the effect of building mass on
seismically induced inertial forces. They also learned that soil properties affected demands. The
30 psf lateral pressure adopted in San Francisco after 1906 as a surrogate for seismic shear would
actually be reduced in the following years to as little as 15 psf by 1926 (Tobriner, 1984).
Meanwhile, a 1911 Italian code required design for lateral forces equal to one twelfth of the
building weight (Holmes, 1998).

4-5

 FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

4.2

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

1925-1932

The Santa Barbara earthquake of 1925 caused widespread damage to structures. Coming in
its wake, the 1927 Uniform Building Code was the first edition to include specific provisions for
earthquake resistant design. It endorsed some fundamental concepts that remain the basis for
provisions even today. They included:
•

Masses are assumed concentrated at the floors.

•

Only permanent dead and live loads are included in the seismic mass.

•

The design force at each level is proportional to the level’s mass.

•

Forces are taken orthogonal to the building’s primary axes.

•

Stiffness should be symmetric about the center of mass (to control torsion).

•

Different lateral forces should be used for different soil conditions.

•

Calculations provided by the engineer to the building official should include a summation of
the seismic masses, a description of the bracing system and its intended behavior, and a
calculation of the stresses on a typical building frame.

The code also specified allowable stresses for different materials subject to earthquake
forces. Masonry stresses were limited to 40 psi and reinforced concrete stresses to 0.04f’c. For
gravity load design, the allowable stresses in concrete beams with or without stirrups was 0.02f’ c
or 0.06f’ c , respectively. This suggests that no special allowable stress increase was permitted for
concrete structures subjected to seismic forces. For wood, however, a one third increase was
permitted. And most interestingly, combined stresses in steel could exceed working stress limits
by 50%. This provision clearly identifies steel as a material more suitable for earthquake
resistant design than either concrete or masonry. The perceived performance of steel in the 1906
and 1925 earthquakes had made an impact on the design and construction community.
4.3

1933-1958

The Long Beach earthquake of 1933 also changed seismic design in California. In the codes
that followed, masonry bearing wall buildings were to be designed for 0.10g, while concrete
shear wall and frame structures could be designed for between 0.02 and 0.05g. In 1937, taller
(over three stories) shear wall buildings of all types were to be designed for 0.06 to 0.10g, while
complete moment frame buildings could be designed for 0.02 to 0.06g, provided the frame could
resist 0.02g on its own (SEAOC, 1968). The ranges of design values were based on soil
conditions. Interestingly, some practices adopted in the Los Angeles area after Long Beach did
not become standard in San Francisco until about 1950 (Steinbrugge and Moran, 1954).
Code provisions developed in the 1940s began to incorporate the period of the building into
the base shear, recognizing that the fundamental periods of taller structures resulted in lower
amplification of ground motion. Los Angeles adopted base shear provisions in 1943 that varied
with building height. In 1948, San Francisco adopted provisions making design base shear
inversely proportional to building period, based on recommendations of the Joint Committee on
Lateral Forces, formed from local chapters of ASCE and SEAONC (Popov et al., 1993). This
4-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

committee also developed the “K” factor, which initially distinguished buildings from nonbuilding structures, but which would come to represent the system ductility factor for different
materials and framing systems (SEAOC, 1968).
4.4

1959-1965

In 1959, SEAOC issued its first Lateral Force Recommendations, also known as the Blue
Book. The 1959 Blue Book redefined the “K” factor as a function of building material and
structural system. Lower K values for ductile systems and materials recognized better inelastic
performance and energy absorption (Strand, 1984).
Also in the 1959 Blue Book, buildings over 160 ft tall were required to have a lateral system
that included a complete moment resisting space frame “made of a ductile material or a ductile
combination of materials. The necessary ductility shall be considered to be provided by a steel
frame with moment-resistant connections or by other systems proven by tests and studies to
provide equivalent energy absorption” (SEAOC, 1959). In 1961, the UBC adopted the Blue
Book provisions; the UBC has incorporated SEAOC’s Lateral Force Recommendations ever
since.
The new moment frame requirement in code section 2313(j), or “section j,” was immediately
and highly controversial (Layne et al., 1963; Kellam, 1966). The height trigger of 160 ft was held
over from earlier Los Angeles codes so as not to suddenly render obsolete all older buildings
without moment frames. Nevertheless, together with the explicit endorsement of steel, the height
trigger was perceived as an arbitrary limit on concrete (Layne et al., 1963).
Engineers who participated in the drafting of “section j” were seeking a tough, reliable
system for “major buildings” and clearly favored steel frames. One later noted that if not for
political and legal considerations, steel frames would simply have been mandated for tall
buildings (Layne et al., 1963). The Blue Book writers instead accepted any system that could
demonstrate ductility equivalent to that of a steel frame.
4.5

1966-1985

In response to the UBC’s “section j” endorsement of steel frames, specific provisions for
ductile concrete frames were completed by SEAOC in 1966 and adopted by the UBC in 1967.
But while the new concrete provisions were expected to provide sufficient ductility and energy
absorption, the SEAOC code writers did not consider this concrete system equivalent to steel
(Kellam, 1966). The 1968 Blue Book commentary (SEAOC, 1968) made clear that steel
remained the standard for seismic performance:
“Moment-resisting frames of ductile materials have shown particularly good earthquakeresistant characteristics…. The ability of various building materials to achieve desired
ductility is not equivalent, by any means. The property exhibited by moment-resisting
space frames of structural steel of ASTM A-7, A-36 and A-441 has long been accepted as
the desirable standard.”

4-7

 FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Though intended only to distinguish new concrete provisions from established steel practice,
this is, in retrospect, a bold statement. Regarding the new ductile concrete frame provisions, the
Blue Book commented that “it has not been possible to make a comprehensive evaluation of …
the performance of such structures in response to earthquakes.” But the same could have been
said for welded or even bolted steel moment frames. Only a handful of steel moment frame
buildings were investigated thoroughly after earthquakes prior to 1968, so the perception of good
earthquake resistance must have relied on research and testing. However, while testing had been
performed as early as the late 1950s, significant study of the cyclic inelastic behavior of steel
members began in 1959 (Bertero and Popov, 1965), the same year that “section j” was drafted.
The first cyclic connection tests began only in 1966 (Popov and Pinkney, 1969).
Thus, while ductile material behavior had “long been accepted as the desirable standard,” a
track record of actual building performance had not been established. Indeed, Popov and Bertero
(1970) cited damage to the Cordova Building in the 1964 Alaska earthquake as motivation for
later tests. Still, the Blue Book assigned the lowest (best) “K” factor to ductile moment-resisting
space frames of steel or ductile reinforced concrete. Requirements for ductile steel moment
frames included:
•

Steel of grade A-36, A-440, A-441, A572, or A588.

•

Moment connections capable of developing the full plastic capacity of the girder. An
exception was made if “adequately ductile joint displacement [was] provided.”

•

For high strength steels, plastic hinges away from bolt holes.

•

Testing of butt welded connections between the girder and column flanges. A testing
program was to be established by the engineer.

These 1968 provisions represented the first codified description of a WSMF. The exception
given in the second requirement is interesting because the necessary calculations, involving
inelastic response to unreduced seismic loads, would have been unusual for most engineering
offices of the time.
The last requirement shows an attempt to address quality control in the welded joints from
the very introduction of WSMFs. At the time, ultrasonic testing (UT) was relatively new in
building construction. Standards and specifications did not match those available for radiography
or magnetic particle testing, and the value of UT depended on the skill of individual technicians.
Still, the nature of UT was considered well suited to WSMF construction. Experts from
Bethlehem Steel noted in 1965 that “Moment connection welds, and especially the vertical
fusion areas of such welds, are more susceptible to definitive ultrasonic examination than to
inspection by other methods” (Couch and Olsson, 1965). The value of UT for inspection of
1980s era WSMFs would be questioned after the Northridge earthquake (Paret, 1999).
The 1976 UBC required all welds to be inspected until a rejection rate “consistently” under
5% was established, at which time the testing rate “might be reduced to 25%.” It has been
suggested that the relaxation of inspection may have contributed to the performance of buildings
in Northridge (Goltz and Weinberg, 1998). This concern may not be justified, as many buildings

4-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

exhibited damage in the lower floors, where 100% of the welds would have been tested during
construction.
In 1975, the SEAOC Blue Book recommended additional requirements for ductile steel
moment frames. With respect to connection welds, the commentary notes that residual stresses
in beam-column welds can result from the welding sequence. This may have been a reference to
damage to the ARCO tower discovered after the 1971 San Fernando earthquake (see the section
on past performance below). Several figures are included in the Blue Book commentary
showing welded moment frame and stiffener plate details. The use of welded webs is also
suggested to increase the ductility relative to a bolted web connection. Mention is made of
column panel zone and continuity plate requirements, which were intended to limit shear
deformation and column flange distortion.
Ironically, the need for reliable inelastic behavior at the beam end may have hastened the
transition to the welded pre-Northridge detail now considered inadequate. From 1963 until after
1978, Part 2 of the AISC Specification, which addresses plastic design, warned against punched
holes in the beam tension flange, perhaps discouraging the use of bolted connections. The 1973
Specification, citing Popov and Pinkney (1969), made a point of noting that full plastic capacities
could be achieved with bolted connections “instead of full penetration groove welds.” But by
1975, the standard connection in California joined the beam to the column by welding the beam
flange and bolting the beam web to a shear tab (SEAOC, 1975). A bolted flange connection was
simply more expensive (see Table 2-2).
FCAW single-bevel groove welds were prequalified by the AISC Specification for the first
time in the 1973 Seventh Edition.
4.6

1986-1988

UBC provisions for steel moment frames remained essentially unchanged from 1970 through
1985. Perhaps this was related to the attention given to concrete after the 1971 San Fernando
earthquake. Research performed in the 1970s and early 1980s eventually led to significant
modifications in the 1988 UBC. Krawinkler (1985) summarized this work and SEAOC’s
tentative provisions that were ultimately adopted by the UBC.
•

Beam-to-column connections. Welded beam web connections should be used to develop
the moment strength of the beam. Krawinkler noted that in beams with a Zf/Z ratio (flange to
beam plastic section modulus ratio) less than 0.7, a connection that develops the moment
capacity of the beam as required by the code will rely heavily on the moment strength
provided by the beam web. Under bending and shear loading, a bolted connection can slip,
leaving the flange weld to carry the moment by itself. This may result in flange failures
“within the weld, at the toe of the weld, or at the interface between the weld and the column
flange.” Krawinkler, citing Popov and Stephen (1972), noted that a welded web that
supplements or replaces the bolted connection can delay this failure until larger plastic
rotations have been achieved.

•

Panel zones. High panel zone strengths demanded by earlier codes often required the use of
doubler plates to thicken the panel zone. In part to avoid the substantial fabrication costs of
4-9

 FEMA-355E
Chapter 4: Codes and Standards
for Steel Moment Frames

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

attaching doubler and continuity plates, engineers in the 1980s began to design WSMFs with
larger column sections. These sections had flanges and webs of sufficient thickness to avoid
doubler and continuity plates.
SEAOC’s 1985 tentative provisions relaxed the panel zone requirements to permit some
yielding there. This would reduce the need for doubler plates and lessen the plastic
deformation requirements in beams. Kinking of the column flange at high panel zone
rotations remained a concern, however.
•

Strong column–weak beam. The 1988 UBC also formalized the strong column–weak beam
concept. The goal was to avoid frame configurations that would be subject to single story
mechanisms or collapse due to P-delta effects. This was primarily considered a problem
when columns had initially high axial loads. This condition would have been more
pronounced in single-bay moment frames which typically had relatively large overturning
forces.

Popov (1987), aware of coming 1988 provisions, remarked, “Due to the great uncertainty of
the forces that a structure may have to resist during an earthquake, complete reliance on the
minimum code provisions is hazardous.”
The 1988 UBC also finally included language for the “prequalified” WSMF connection that
had been standard practice in California since the mid-1970s (SEAOC, 1975). A connection was
considered adequate to develop the moment capacity of the beam if the beam flanges had full
penetration welds to the column and if the beam web connection was able to resist both its
gravity and seismic shear demands. In addition, if the Zf/Z ratio of the beam was less than 0.7,
the web had to be welded to the shear tab to provide additional moment strength.
4.7

1989-1993

After 1988, the Uniform Building Code and the SEAOC Lateral Force Recommendations did
not revise any detailed provisions for steel moment frames. However, the UBC’s general seismic
provisions have undergone substantial changes. The transition to ultimate stress design began or
was completed for most materials, and changes in the calculation of lateral-force demands were
introduced, using an “Rw” factor to replace the “K” factor. Issues relating to building
irregularities were fleshed out in more detail, considering the performance of buildings in the
1985 Mexico City and 1989 Loma Prieta earthquakes. The concept of strength design and more
realistic earthquake force and drift demands were incorporated into the code with the
introduction of the (3/8)Rw factor.
In 1992, AISC published its first set of seismic provisions. With respect to WSMFs
(designated as Special Moment Frames), the AISC provisions endorsed the same prescriptive
detail as the 1988 UBC. Later editions of the AISC provisions would be adopted by both NEHRP
and the International Building Code.
Code changes in response to the January 1994 Northridge earthquake are discussed in a
separate section below.

4-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

5. PERFORMANCE OF STEEL FRAME BUILDINGS IN PAST
EARTHQUAKES
The effects of large earthquakes on structures in the United States have been observed since
the early 1800s. Since around 1900, steel framed buildings have experienced heavy shaking in
almost every major event. By observing their performance, engineers are able to advance the
state of the art and practice in steel moment frame design. While buildings erected before the late
1960s generally did not employ welded moment connections, a review of steel performance in
prior events teaches us how the use of steel evolved into a common and popular material. During
the 1970s, welded steel moment frames began to quickly replace the all-bolted moment frame in
most regions of high seismicity. Lessons learned from the performance of steel frames since the
1970s are crucial to the advancement of the WSMF state of the art and practice.
One problem reviewing records of damage in past earthquakes is that the term “steel framed”
has been used to mean more than just moment resisting frames and certainly more than just
typical pre-Northridge WSMFs. For example, diagonally braced frames (Yanev et al., 1991),
infilled frames (Steinbrugge and Moran, 1954), intended or unintended dual systems (Berg and
Stratta, 1964; Yanev et al., 1991), steel gravity frames not designed to resist lateral load, and
prefabricated “rigid frame” warehouse structures have all been categorized as “steel frame
structures.”
Another problem is that careful postearthquake inspection of beam-column joints was rarely,
if ever, performed before Northridge. In several cases, WSMF damage from the 1989 Loma
Prieta earthquake was found only after the Northridge experience prompted some reinspections
of structures previously considered undamaged. Frames analyzed after the 1971 San Fernando
earthquake (see Table 5-3) were also classified as undamaged based on nominal inspections.
Except for one that was studied again after Northridge and found damaged, those San Fernando
case study buildings probably have not been looked at closely since.
But the major problem in trying to gauge the past performance of steel moment frames is the
simple lack of data. Preece (EERI, 1976) recognized this in a reconnaissance report after the
1976 Guatemala earthquake:
One high-rise structural steel building 22 stories (sic) in this City can hardly be
considered a test of structural steel performance, especially when a 21-story reinforced
concrete building next to it also came through unscathed.
Despite these difficulties, the SEAOC Blue Book noted through the 1970s that steel moment
frames “have shown particularly good earthquake-resistant characteristics,” and through 1990
that WSMFs “are believed to be a proven, reliably ductile structural system” (emphasis added).
A few successes, or rather the lack of any notorious failures, established a reputation that
spanned decades, even as details and construction techniques changed profoundly.
The previous section of this report showed how a 1968 Blue Book statement about steel’s
ductility could have been misinterpreted as a record of actual building performance. Post5-1

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

earthquake observations may have been misinterpreted or misapplied by later engineers as well.
Goel made this point, if somewhat obliquely, in 1968. Commenting on a lack of test data, he
described contemporary building codes—in general, not just the steel provisions—as having
produced “designs that have successfully withstood severe earthquakes in the past with little or
no damage at all.” Yet his citation was to a 1955 report on the 1952 Kern County earthquake,
which affected engineered structures built mostly in the 1920s and 1930s.
The AISI-sponsored study by Yanev et al. (1991) is noteworthy. The authors intentionally
chose modern era earthquakes affecting steel, concrete, and masonry structures in order to make
useful comparisons. Their basic conclusion:
Buildings of structural steel have performed excellently and better than any other type of
substantial construction in protecting life safety, limiting economic loss, and minimizing
business interruption due to earthquake-induced damage.
Yanev would emphasize the point at an AISC conference (Melnick, 1991):
[S]teel will always outperform concrete in an earthquake … If you want to go beyond
code without paying for it, go steel. What is the only building designed for Zone 2 that
can survive a Zone 4 earthquake? Steel.
Indeed, steel has outperformed other structural materials in earthquakes (see Table 5-2). But
in the Northridge earthquake at least, WSMFs did not live up to engineers’ high expectations.
Whether or not the WSMF remains the system of choice for seismic resistance, it is clear in
retrospect that enthusiastic pre-Northridge endorsements suffered from all three of the problems
noted above: a conflation of reports from various structural systems old and new, cursory postearthquake inspection, and generally sparse data. This is perhaps the singular lesson of this
report.
From a post-Northridge perspective, other principal lessons from a review of earthquake
records prior to Northridge include:
•

Until Loma Prieta in 1989, only a handful of modern WSMF buildings had ever been shaken
by a major earthquake. Fewer than a dozen WSMFs were closely inspected after the 1985
Mexico City, 1971 San Fernando, and 1964 Prince William Sound earthquakes combined.

•

WSMF buildings are still relatively rare, and WSMF damage is less obvious than concrete,
wood, masonry, or even steel braced frame damage. When damaging earthquakes occur, it is
therefore easy to overlook potentially damaged WSMF structures and to conclude, perhaps in
error, that as a class they offer excellent performance. That early post-Northridge reports
were prepared to make such an assessment (AISC, 1994) is instructive.

•

Engineers have consistently attributed structural failures to poor construction quality.

Table 5-1 summarizes the recorded performance of steel buildings of the WSMF era (post1960) prior to Northridge. Due to the volume of collected data, WSMF performance in the 1994
Northridge earthquake is discussed in a separate section. Table 5-2 briefly compares the broad

5-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

performance of different structure types in major North American earthquakes since 1960. An
exhaustive comparative study is beyond the scope of this report. The text that follows describes
effects of selected pre-Northridge earthquakes relevant to the design, testing, regulation, and
performance of steel frame structures. Appendix C includes descriptions of steel moment frame
performance in the 1995 Kobe (Japan) and 1999 Ji-Ji (Taiwan) earthquakes.
5.1

San Francisco, 1906

Buildings that experienced the 1906 San Francisco earthquake did not have welded steel
moment frames. San Francisco and Oakland did have, however, many multi-story steel, cast iron,
or wrought iron frame or skeleton buildings with riveted connections and masonry infill.
Reports compiled by the United States Geological Survey (USGS, 1907) described some of
the damage to steel buildings. Steel framed structures “braced” for wind resistance performed
better than those without bracing. (In this era, the term “bracing” referred to any means of
resisting lateral loads, including diagonal rods, portal frames with rigid moment connections, and
knee braces.) For example, of the Call Building it was written, “Had the building been as well
designed to resist fire as to resist earthquake, it is probable that the total damage would have
been very much less than it was.” But from the damage descriptions it is clear that most steel
frames were not as heavily braced as the Call Building.
Partial infill and masonry piers in steel frames were widely damaged by story racking.
Damage to frame components included buckled column plates associated with failure of terra
cotta or plaster fireproofing (e.g. Aronson Building, Bullock & Jones Building, Hotel Hamilton)
and some sheared rivets (e.g. Union Trust Building). Much of the damage to riveted connections
was attributed to “faulty construction” or “careless workmanship.” Diagonal rod bracing in some
steel structures (e.g. Call Building, Union Ferry Building) was yielded and buckled.
Earthquake or fire damage to masonry or plaster fireproofing left the steel frames vulnerable:
“In the San Francisco fire, for the first time, the collapse of protected steel frames, due to the
destruction of the fireproof covering at a comparatively early stage in the fire, was a matter of
common occurrence (USGS, 1907).” Interestingly, however, while some city blocks were
dynamited to create fire breaks, one writer noted in the USGS report how unsuccessful it was
and how difficult it would be to bring down a steel frame building with dynamite.

5-3

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Table 5-1

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era

Earthquake and
Magnitude
(References)
Prince William Sound,
Alaska, 1964
8.4, 6.7
(Berg and Stratta, 1964;
Yanev et al., 1991. See
also Table 5-2)
Venezuela, 1967

Performance of Steel Moment-Frame
Buildings

Overall Performance at Nearest Urban
Center (epicentral distance in km)

Multistory steel frames rare.

At 120 km: Several complete collapses of
multistory concrete and masonry
structures. Small rigid masonry structures
mostly undamaged due to low frequency
shaking. Much damage related to soil
failures. Most casualties due to tsunami.

At 120 km: Cordova building: At first
story, buckled steel columns and
damaged concrete core walls. Steel frame
connections had beam flanges bolted to
top and bottom clip angles.

At 50 km (Caracas): Many reinforced
concrete buildings with “major damage,”
including four complete collapses of 10 to
12-story buildings.

(Hanson and Degenkolb,
1969)

At 50 km (Caracas): “There were a few
multistoried steel buildings in Caracas—
none of which suffered significant
damage.” Only the Simon Bolivar Center
is described: mid 1950s 30-story “steel
frame.”

Tokachi-Oki, 1968

No steel moment frame data.

At 200 km: Hundreds of collapses, mostly
wood residences. Several concrete
buildings with severe damage or collapse.
In steel braced frames, some buckling and
fracture at splices.

No steel moment frame data.

At 25-50 km: Extensive and severe
damage to concrete, historic unreinforced
masonry, and adobe.

At 10-40 km: 30 steel moment frames
with no observed damage. Connection
types not given. All but 5 erected before
1967. At 25 km: One 16-story 1969
building described as undamaged (Yanev)
would later be significantly damaged by
1994 Northridge earthquake (Kariotis and
Eimani, 1995). At 40 km: cracked welds
in two WSMF highrises under
construction. At 3 km: “some working of
the connections” observed (Yanev).

At 10-40 km: Many concrete collapses,
including medical facilities. Tilt-up
damage and collapses. Some damage to
steel diagonal braces.

Only one steel building, frame and
connection type not described. “Signs of
yielding” in some ground floor columns.
Nonstructural damage included broken
exterior glass panels.

Most modern commercial buildings,
typically with soft stories and masonry
infill, “performed very poorly.” Some
masonry and concrete buildings “escaped
serious damage.”

6.5

7.9
(Yanev et al., 1991)
Peru, 1970
(EERI, 1970)
San Fernando,
California, 1971
6.6
(Steinbrugge et al., 1971;
Yanev et al., 1991. See
also Tables 5-2 and 5-3.)

Managua, Nicaragua,
1972
6.2
(Yanev et al., 1991)

5-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 5-1

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era (continued)

Earthquake and
Magnitude
(References)
Guatemala, 1976
7.5
(EERI, 1976)

Friuli, Italy, 1976

Performance of Steel Moment-Frame
Buildings

Overall Performance at Nearest Urban
Center (epicentral distance in km)

No steel moment frames. One 19-story
steel braced frame / dual system
described (150 km from 7.5 main shock,
40 km from 5.8 second shock): A36 steel,
E60XX full-penetration girder-column
welds, A490N bolted webs, Seventh
Edition of AISC specified.
Postearthquake visual inspection of some
welds by reconnaissance team revealed
“excellent” quality and, by inference, no
visible damage. No structural damage and
very light nonstructural damage (although
brand new, so no contents in place).
Adjacent 21-story concrete building also
undamaged.

At 50-150 km: Four collapses of infilled
concrete frames. Extensive damage and
collapse to adobe residential construction.

No steel frame data.

At 0-30 km: General building damage:
about 80% of private buildings heavily
damaged or destroyed.

No steel frame data.

General building damage: 35 mid-rise
concrete collapses in Bucharest (at 170
km), 32 of them pre-1940.

Likely several hundred steel structures,
few with obvious damage, and few
studied. Three steel moment frame
buildings described: one 1973 17-story
tower apparently undamaged, but no
postearthquake joint inspection. One 18story dual system with minor shear wall
cracks. One 4-story steel frame with
precast panel failure.

In Sendai (100 km, 0.25g to 0.40g pga):
“Good general performance of modern,
engineered buildings up to 20 stories
high.” But at least four complete collapses
of concrete structures. Some steel brace
buckling and fracture.

6.5, 6.0
(Stratta and Wyllie,
1979)
Romania, 1977
7.1
(Berg et al., 1980)
Miyagi-Ken-Oki, Japan,
1978
MS = 7.4
(EERI, 1978)

5-5

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Table 5-1

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era (continued)

Earthquake and
Magnitude
(References)

Performance of Steel Moment-Frame
Buildings

Overall Performance at Nearest Urban
Center (epicentral distance in km)

Oaxaca, Mexico, 1978,
and Guerrero, Mexico,
1979

No steel frame data. Sheared high
strength bolts in trussed portal frame at
steel mill building near Guerrero.

Oaxaca: Generally minor damage, no
collapses. Heavy damage to one 2-story
concrete frame.

MS = 7.8, 7.6

Guerrero: Widespread damage to
unreinforced brick, adobe. Isolated
concrete frame damage. Significant
damage to masonry infill.

(Forell and Nicoletti,
1980)

At 300-500 km: Felt strongly in Mexico
City despite distance. Pounding and
nonstructural damage to tall buildings.
Collapse of one 3-story concrete frame.
Montenegro, Yugoslavia,
1979

No steel frame data.

At 10-25 km: Severe damage to old
unreinforced stone masonry, minor
damage to 1950s brick and block masonry,
good performance of concrete bearing wall
and precast bearing wall buildings, poor
performance of concrete frames (many
infilled).

No steel frame data.

At 20-90 km: General building damage:
many collapses and near collapses of stone
masonry residences and infill concrete
frames.

No steel frame data.

At 0-20 km: 20% of El-Asnam buildings
collapsed, 60% severely damaged.
Collapses included many “modern”
multistory concrete structures.

No steel frame data.

At 20-70 km: General building damage:
severe damage to block masonry and infill
concrete frames.

MS = 6.6
(EERI, 1980)

Campania-Basilicata,
Italy, 1980
6.8
(Stratta et al., 1981)
El-Asnam, Algeria, 1980
MS = 7.3
(EERI, 1983)
Central Greece, 1981
6.7, 6.3
(Carydis et al., 1982)

5-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 5-1

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era (continued)

Earthquake and
Magnitude
(References)
Coalinga, California,
1983
6.7

Performance of Steel Moment-Frame
Buildings

Overall Performance at Nearest Urban
Center (epicentral distance in km)

Little steel frame data.

At 10-15 km: Severe damage to about half
of wood cripple-wall residences, and most
unreinforced masonry buildings. Very light
damage to wood-frame commercial,
concrete block, and cast-in-place concrete
buildings.

Two 1940s steel frame buildings with no
visible damage (Yanev).

(Tierney, 1985; Yanev et
al., 1991)
Borah Peak, Idaho, 1983

No steel frame data.

At 30 km (Mackay): slight to moderate
unreinforced masonry damage. At 0-60
km: mostly minor damage; worst damage
involved masonry parapet and veneer
failures.

One 1976 steel frame building described
(20 km, 0.04g recorded pga, 0.18g roof
acceleration): Nonstructural and contents
damage, but evacuated because of long
duration lightly damped response.
Retrofitted in 1991 with dampers. No
systematic joint inspection after
Northridge, but 1991 retrofit exposed
about 200 connections, and no obvious
visible damage was reported (Crosby,
1999).

At 10-30 km, no structural damage or light
structural damage to “engineered” steel
and concrete buildings. Some tilt-up
damage. Structural damage, including a
few collapses, to less than 10% of
residences.

The few steel industrial structures
performed well, but unbraced frames had
substantial nonstructural damage. At 64
km: Wide-flange sections with bolted
connections used as horizontal braces
suffered some web tearing and gusset
buckling in one industrial building.

At 100 km (Santiago): Heavy damage to
historic adobe and URM structures.
Hundreds of 5-25 story concrete buildings;
most appeared to perform well, but many
with significant damage, and several
collapses.

Few “modern” WSMFs in the region.
Some older moment frames with infill,
knee braces, or riveted connections
collapsed. Few post-1957 moment frames
damaged, but some weld fracture noted.
Collapse of 3 steel frame structures with
braced bays due to column overload
unrelated to frame action.

At 300-500 km: Hundreds of collapses,
thousands of buildings damaged. Unique
ground motion and soil conditions hit midrise buildings especially.

7.3
(Earthquake Spectra,
November 1985)
Morgan Hill, California,
1984
ML = 6.2
(Earthquake Spectra,
May 1985)

Chile, 1985
MS = 7.8
(Earthquake Spectra,
February 1986)

Mexico City, 1985
8.1
(Osteraas and
Krawinkler, 1989; Yanev
et al., 1991. See also
Table 5-2.)

5-7

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Table 5-1

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era (continued)

Earthquake and
Magnitude
(References)
Whittier, California,
1987
ML = 5.9

Performance of Steel Moment-Frame
Buildings

Overall Performance at Nearest Urban
Center (epicentral distance in km)

No modern steel frame data, but Los
Angeles area WSMFs certainly were
shaken by the earthquake. No structural
damage to 1920s steel frame with
masonry cladding.

At 0-30 km: Substantial damage to
masonry bearing wall buildings, with
lower damage rates among reinforced
buildings. Significant structural damage to
several modern concrete structures,
including parking garages and tilt-ups.

No steel frame data.

At 10-30 km: collapses and severe damage
to typical multistory stone masonry and
precast frame structures.

About 30 WSMFs inspected, nearly all
after Northridge damage found. Five with
connection damage. Pounding damage at
seismic joints between sections of ductile
steel frame complex.

At 60-100 km: Most fatalities from URM
failures and collapse of concrete highway
structure. Severe damage to isolated
concrete, wood frame, and tilt-up
structures.

Based on preliminary observations, and in
contrast to adjacent concrete structures,
“performance of steel-frame buildings
was excellent, consistent with
observations in other earthquakes.” Even
in areas of extensive liquefaction and
settlement, steel buildings were observed
to be undamaged.

At 40-60 km: Many concrete collapses in
Baguio. Central business district of
Dagupan “essentially destroyed” due in
part to soil spreading.

One two-story steel frame structure with
severe cracking discovered after
Northridge. Los Angeles area WSMFs
were shaken by the earthquake, but no
steel frame data reported.

Near epicenter: Isolated wood frame and
concrete block structure damage.

(Earthquake Spectra,
February 1988; H.J.
Degenkolb Associates.)
Spitak, Armenia, 1988
MS = 6.8
(Earthquake Spectra,
August 1989)
Loma Prieta, California,
1989
7.1
(Phipps, 1998; Yanev et
al., 1991. See also Tables
5-2 and 5-4.)
Luzon, Philippines, 1990
MS = 7.8
(Earthquake Spectra,
October 1991; EQE,
1990)

Landers and Big Bear,
1992
MS = 7.5, 6.6
(Phipps, 1998; Reynolds,
1993)

5-8

At 170 km: 0.04g maximum ground
acceleration recorded in Los Angeles.

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 5-1

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Earthquake Performance of Steel Moment-Frame Buildings in the
WSMF Era (continued)

Earthquake and
Magnitude
(References)
Hokkaido, Japan, 1993
MW = 7.8, 6.3
(Earthquake Spectra,
April 1995a)

Guam, 1993

Performance of Steel Moment-Frame
Buildings

Overall Performance at Nearest Urban
Center (epicentral distance in km)

At 50-80 km: 2 steel frames noted, both
undamaged, but neither similar to typical
California office building. At batch plant
hopper, first story steel frame bent
undamaged despite buckling of braced
frame structure above. Two-story steel
frame building survived devastating
tsunami with no structural damage.

General building damage: light to
nonexistent due to shaking, moderate to
total due to tsunami. Extensive
liquefaction.

No steel frame data.

Significant damage to non-ductile concrete
buildings. Dozens of buildings severely
damaged but no deaths.

MS = 8.1
(Earthquake Spectra,
April 1995b)

5-9

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Table 5-2

Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Damage by Structure Type in Selected North American Earthquakes of the
WSMF Era
Wood frame
residential

Unreinforced
Masonry

Precast Concrete

Concrete frame
or wall

Steel frames

Prince William Excellent
Sound, 1964
performance
(Wood, 1967).

Not a common
Typically slight to 1-5 story: generally
construction type moderate damage, no significant
in the area.
less than expected structural damage,
(Wood, 1967).
one partial
collapse. Taller:
considerable
structural damage.

Inconclusive.
Considerable structural
damage, but only a few
buildings had complete
steel frames.

San Fernando, Majority of
1971
buildings under
20% loss. Most
structural damage
in foundation
anchorage and
open fronts.

Moderate or
Multiple tilt-up
severe damage to collapses.
half of brick
buildings in
downtown San
Fernando.

Many collapses
generally caused by
poor ductility and
irregularities.

No significant structural
damage was observed
in general. Cracked
welds observed in two
buildings under
construction.

Mexico City,
1985

Not a common
building type in
the area.

Many URMs
Not a common
severely
construction type
damaged. Wall- in the area.
floor connections
and out-of-plane
failures.

7-15 story, frame
and infill structures
heavily damaged or
collapsed. Similar
low-rise structures
perform better
(Bertero and
Miranda, 1989).

Some pre-1950 steel
frames collapsed.
Collapse of isolated
braced frame buildings
due to column failure.
Little other moment
frame damage reported.

Loma Prieta,
1989

Severe damage to
wood structures
with open fronts
or tuck-under
parking,
especially on soft
soils.

Many URMs
severely
damaged or
collapsed. Wallfloor connections
and out-of-plane
failures.

Many examples of
connection
damage although
collapses were
uncommon.

Nonductile frames
and wall structures
damaged, including
fatal freeway
collapse. Newer
structures generally
performed well.

Five buildings known to
have slight to
significant weld
damage, some
discovered only upon
post-Northridge
inspection.

Northridge,
1994

Severe damage to
multi-story wood
structures with
tuck-under
parking.

Many URMs
damaged but
many collapses
avoided by
previous
retrofits.

Many examples of
connection
damage and
several significant
collapses.

Older structures
damaged, including
freeways. Newer
structures generally
performed well.
Deflection
compatibility
issues noted.

Widespread,
unexpected connection
damage, several
buildings declared
unsafe, at least one
irreparable. No
collapses.

Note: See Table 5-1 for additional information and references.

5-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Frank Soulé, dean of the University of California college of civil engineering, was given the last
word in the USGS report. About twenty years after the first steel frame buildings were erected,
and with only the San Francisco earthquake as an historical record, he described structural steel
as “no longer in the experimental stage as to resistance … to earthquake tremors.” While calling
diagonal bracing an “absolute necessity,” he nevertheless praised the San Francisco performance:
Undoubtedly many of the high steel buildings in San Francisco were designed without
reference to earthquakes, but they have nobly withstood their effects, and steel frames
have proved themselves entirely adapted to earthquake countries. [In the San Francisco
earthquake], they suffered comparatively little injury, … confined to the shearing of
rivets and connections … and to some buckling of braces.
Despite such an endorsement, it is hard to draw any lessons of particular usefulness from this
event with respect to the performance of modern WSMFs. Any buildings from 1906 that are still
in service face potential problems different from those of WSMFs. Indeed, several steel
structures with diagonal bracing or knee braces, with or without masonry infill, collapsed in the
1985 Mexico City earthquake (Osteraas and Krawinkler, 1989).
5.2

Kanto, Japan, 1923

The use of steel in Japan at the time of the Kanto earthquake was relatively new, with a
history of only about five years. Four large buildings had been completed and two were almost
complete. These buildings suffered little to no damage. The use of masonry infill was common.
Damage to the infill and facades was significant, but the frames performed well. Braced frames
also suffered relatively little damage. Two steel bridges, one on masonry piers and one made
entirely of steel, performed quite differently, with the former collapsing and the latter suffering
almost no harm (Hadley, cited by SAC, 1998). Japanese building codes adopted seismic design
coefficients after this earthquake (Tobriner, 1984).
5.3

Santa Barbara, 1925

As did the 1906 earthquake, the smaller event in Santa Barbara highlighted the good
performance of steel frame infill buildings relative to masonry and concrete structures.
Seventeen concrete and masonry buildings were destroyed or eventually demolished, but two
steel frames close to the epicenter were not severely damaged. The largest of the two was a post
office in an area severely hit by the earthquake. The other building was a church. (California
Institute of Steel Construction, cited by SAC, 1998).
The lack of damage to these buildings again led people to think of steel as a material well
suited to resisting earthquakes. The concept of incorporating flexibility into a building to help it
resist damage was becoming popular. This earthquake gave rise to the first seismic design
provisions in U.S. building codes, most of which were probably based on work done in Japan
(Tobriner, 1984; Strand, 1984).

5-11

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

5.4

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Long Beach, 1933

While few steel structures were affected by the 1933 Long Beach earthquake, the damage to
other buildings was again used to gauge the relative performance of steel versus masonry and
concrete. The sixteen story Villa Riviera building was a steel frame built in 1928. It suffered
virtually no damage, while several concrete and masonry structures nearby collapsed or were
heavily damaged (SAC, 1998).
Though not of great seismic magnitude, the Long Beach earthquake did extensive damage to
a densely populated area with many vulnerable buildings, including several schools (Coffman
and von Hake, 1973). As noted in section 4 of this report, the Long Beach earthquake was
followed by significant earthquake safety legislation (Strand, 1984; SEAOC, 1968).
5.5

Kern County, 1952

Following the 1952 Kern County earthquake, Karl Steinbrugge and Donald Moran (1954)
studied the “damagability” of different structural systems. The steel structures affected by this
earthquake were almost all infilled frames erected in the 1920s or 1930s in Los Angeles, over
100 km from the epicenter. Some of them suffered minor damage. The only nearby steel frame
building suffered some pounding damage unrelated to its frame connections.
Steinbrugge and Moran considered the affected steel frames to be in the second best category
of building performance. Ranking as best were small wood structures under 3,000 square feet.
Following steel structures in order of performance were: concrete frames and shear walls, large
wood frame buildings, steel frames with URM infill, concrete frames with URM infill, precast
concrete and other flexible diaphragm buildings, and finally URM and adobe structures.
5.6

Prince William Sound, Alaska, 1964

Most of the research into building performance in the 1964 Alaska earthquake focused on
concrete buildings and on geotechnical effects. Of the two dozen or so buildings discussed in
detail by Wood (1967) and Berg and Stratta (1964), only one, the Cordova building, had a steel
frame as its primary lateral force-resisting system.
Berg and Stratta identified one welded beam-to-column failure in the steel framed shop
building at the Alaska Highway Department Yard. The damage was in a welded angle clip
connecting the beam web to the column. The clip appears to have torn away from the flange of
the column. While certainly a fracture of the weld, it is not typical of the WSMF damage seen in
Northridge. The clip was attached to the column with fillet welds that appear to have failed in
prying. Elsewhere in the building, Berg and Stratta noted that “some columns yielded below the
beam connections.” The authors also point out that anchorage of the steel columns to base plates
and of base plates to the footings failed in many places. The failures were attributed to poor
quality welds between the columns and base plates, shearing of the anchor bolts, or spalling of
the footings. Tearing or fracture of base plates was not mentioned.
The Hill Building, an eight-story office building, was a steel and concrete frame with
concrete and CMU walls. The concrete and CMU were badly damaged in several locations, but
5-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Berg and Stratta noted that “there was apparently no damage to the steel frame. Several of the
beam-to-column connections were exposed for the purposes of inspection and found to be in
good condition.” Reports of sheared bolts in the building may have led them to inspect the
connections, which turned out to be simple bolted connections. The authors point out that the
frame was designed to carry gravity loads only.
The Cordova building, a six-story office building, suffered perhaps the worst steel frame
damage. The building had a full steel moment frame in one direction and “partial momentresisting beam-to-column connections [in the other].” Moment connections were made using
shop welding and field bolting with high-strength bolts. Damage was generally concentrated at
the first floor where a number of wide flange columns (typically 14WF30 or 14WF61) buckled.
The damage appears to be axially induced with the column flanges sometimes tearing away from
the column web directly below the beam-column connection. A damaged column is shown in
Figure 5-1. Popov and Bertero (1970) would later cite damage at the Cordova Building as an
indication that monotonic testing is inadequate for demonstrating seismic performance.

Figure 5-1

Damaged Moment-Frame Column, Prince William Sound Earthquake, 1964
Source: Berg and Stratta, 1964

Several other steel buildings were investigated, including a three-story psychiatric institute, a
six-story hospital, and a one-story university building. None were found to have any significant
structural damage, although it is unlikely that the joints were uncovered and closely inspected.
Whether or not other steel frame buildings in Anchorage had weld failures will probably
never be known at this point. This highlights the natural tendency of engineers to focus on the
more obvious building damage, especially in such a powerful event. Damage to concrete
5-13

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

structures and landslide effects were clearly dramatic and evident and kept engineers and
researchers busy. Since the use of welded steel moment frames was not widespread at the time,
especially so in Alaska where there were many more concrete buildings, it may not have
occurred to engineers to look behind lightly damaged furring and curtain wall systems for
evidence of damage to structural joints.
Berg and Stratta concluded in their study of the Alaska earthquake that “structures with steel
frames generally withstood the earthquake well. Steel frames which were damaged were
repairable with ease, speed and economy.” While probably speaking to concrete performance,
they also note that “connection details deserve special attention in earthquake zones. To take
advantage of the energy absorbing capacity of the structural members, one should design the
connections so that first failure would occur in a member rather than in the connection.”
5.7

San Fernando, 1971

To structural engineers, the San Fernando earthquake is infamous for exposing the seismic
hazards of non-ductile reinforced concrete frames and for highlighting the dangers of soft stories.
Several prominent failures led to changes in the building code. There was also some study of
steel structures’ performance. However, as with the Alaska earthquake seven years before,
apparently little attention was paid to buildings that did not exhibit obvious structural damage.
The Pacific Fire Rating Bureau quickly studied thirty completed steel buildings and two
under construction at the time of the earthquake (Steinbrugge et al., 1971). Some stairs, concrete
walls, and nonstructural elements were damaged, but no structural damage to the completed steel
frames was noted: “With respect to complete buildings, the authors know of no significant
structural damage to steel frame high-rise buildings as opposed to several cases known in
reinforced concrete construction.”
By 1973, more data had been collected and analyses completed, including case studies of
instrumented buildings. Table 5-3 lists the recorded peak ground acceleration (PGA) for these
buildings, along with a description of some of the observed damage. Noteworthy is the low
shaking intensity, except at Bunker Hill, where WSMFs would have qualified for inspection per
FEMA-267. Also, the likely inspection scope probably did not involve close scrutiny of
connections. As Jennings (1971) had noted earlier, “it should be emphasized that this earthquake
was too far away from downtown Los Angeles to be a test of the ultimate strength of [the tall
buildings there].”
Records from two of the buildings in Table 5-3 were later analyzed by Foutch et al. (1975).
They noted that the Kajima and Union Bank buildings, about a mile apart in downtown Los
Angeles, were both subjected to an unusual displacement pulse about ten seconds into the
shaking. Accelerogram records also indicated that after initial cycles, each building oscillated at
a natural period longer than had been measured by pre-earthquake ambient vibration tests. The
authors attributed the period shifts in both cases to “cracking and other types of degradation of
nonstructural elements” during the early strong shaking.

5-14

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 5-3

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Case Studies of Instrumented WSMF Buildings Affected by the 1971 San
Fernando Earthquake

Building

Recorded
PGA (g)

Observed Damage

0.29

“The owner of the building reported that no earthquake
damage to any structural elements was observed, and that
only minimal damage to nonstructural elements occurred
… such as cracking to drywalls … Four elevators were
temporarily out of service” (John A. Blume & Associates,
1973).

0.15

No observed structural damage. Minor nonstructural
damage: partitions, seismic joints (Gates, 1973b).

Bunker Hill Tower, 32 stories
800 West First Street, Los Angeles

KB Valley Center, 16 stories
15910 Ventura Boulevard
Kajima International Building, 15
stories

0.14

No observed structural damage. Nonstructural damage to
plaster partitions around elevator and stair cores (Gates,
1973a).

250 East First Street, L.A.
Union Bank Square, 42 stories
0.14
445 S. Figueroa Street, L.A.
1901 Avenue of the Stars,
19 stories, Century City
(moment frame in NW-SE direction
only)

0.12

Nonstructural damage only: superficial plaster cracking in
core walls and stair shafts, elevators out of service
temporarily. (Albert C. Martin & Associates, 1973)
“[No] major structural damage, and only minor
nonstructural damage.” (Hart, 1973)

(NW-SE)

Gates (1973b) performed a thorough analysis of the KB Valley Center, a 16-story WSMF
near Sherman Oaks with 42-inch deep plate girders. A peak ground acceleration of 0.15g was
recorded in the building basement. Gates described the damage as follows:
“There was no observed structural damage to the structural elements of the building as a
result of the San Fernando earthquake. Minor nonstructural damage occurred in
partitions, at seismic joints, and in mechanical equipment mounts.”
This building would later be the subject of a detailed post-Northridge case study (Kariotis
and Eimani, 1995. See also the Northridge section below). In the 1994 earthquake, for which
the PGA at the site is estimated as 0.38g (see Appendix B), the elevators were damaged and
permanent drift was measured in the top third of the building. About 20% of the connections in
the north-south frames were damaged. In seven places, all in the upper stories of the north-south
frames, cracks went through the column flange into the column web.
Another steel frame shaken by both earthquakes was a nine-story structure with about 460
moment frame connections at 18321 Ventura Boulevard. The building was reported as
5-15

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

undamaged after San Fernando (Jennings, 1971), but the scope of inspection is unknown. It is
unlikely that connections were carefully inspected. After Northridge, thirty-one connections
were inspected, and one was reported damaged (Los Angeles Building & Safety, 1998). The
other four towers listed in Table 5-3 were in downtown Los Angeles or Century City, two areas
exempted from mandatory post-Northridge inspection.
The two buildings listed by Steinbrugge et al. as under construction did suffer structural
damage, and the report by the authors gives some insight as to the types and causes of damage.
Their report, which remains the best-known account of the damage and which represents the
thinking of many engineers at the time, states:
“The twin 52-story office towers of the $175 million dollar Atlantic-Richfield [ARCO]
Plaza Towers in downtown Los Angeles were in the latter stages of construction when
the earthquake occurred. Apparently, a 25% increase in the number of cracks in the welds
in the two lower stories of both steel framed towers was found after the earthquake, with
this increase seemingly due to the earthquakes. Miniscule cracks in the welds connecting
heavy metal members occur during the normal welding process and these cracks are
normal to this work; routine ultrasonic testing is used to discover these cracks and allow
for repairs. It is premature to speculate very far into this particular case due to the lack of
time and detailed information, but the potential problem of earthquake induced weld
stress cracks in modern steel frame buildings is disquieting. Additionally, there is no
assurance that all welded steel frame buildings will be as adequately inspected as was the
Atlantic-Richfield towers. The cost of the repair of all welds, regardless of origin, has
been placed at $400,000.”
Notable are the comments that miniscule but rejectable cracks are common when welding
large sections (no reference was cited), that testing and repair was routine, and that the testing at
ARCO might have been better than standard practice. Post-Northridge inspections would later
find that weld inspection might not have been reliable at ARCO or any number of other WSMFs
(Goltz and Weinberg, 1998; Paret, 1999).
In hindsight, the most compelling statement made by Steinbrugge et al. was that “the
potential problem of earthquake induced weld stress cracks in modern steel frame buildings is
disquieting.” While true in general, the nature and cause of the damage at ARCO remains
debatable. A private study for one of the building’s tenants by the J.H. Wiggins Company
(1971) reported cracking in 29% of the second and third floor joints. But in these towers, the
second and third floor framing is part of a full story transfer truss, so joints at those levels are not
typical of the framing above or of WSMFs in general (Phipps, 1998).
In its summary, the Wiggins report described three types of cracks that occurred either alone
or in combinations: “(1) lamellar tearing within the thick column flanges, (2) brittle fracture
within the flanges of the girders and (3) brittle fracture within the webs of the girders.” Lamellar
tearing and pure base metal fractures away from the welds would suggest fracture mechanisms
different from those most commonly seen after Northridge. But later in the Wiggins report the
three damage types are described again: “(1) tearing within the column material … opposite
either the flange or web parts of the girder … (2) cracking in the weld within each flange of the
5-16

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

girder … (3) cracking in the weld of the web of the girder” (emphasis added). These latter
descriptions are consistent with Northridge patterns.
Three sketches in the Wiggins report schematically describe the crack types “within or
adjacent the welds.” Though imprecise and inconclusive, they do suggest fractures initiating
near the mid-length of the groove weld or at a weld access hole and running from there either up
the column flange or across the beam flange and up the beam web. While relatively rare, some
Northridge damage did involve fracture outside the heat-affected zone, typically where the beam
web is coped to form a weld access hole (see, for example, Uang et al., 1995). This location,
especially when not ground smooth, has been a fracture-sensitive point in past tests and in postNorthridge tests (Lee et al., 2000). It now appears that in the pre-Northridge connection, if
fracture at the weld root can be avoided, the weld access hole is the next weakest link.
The Wiggins report attributed the ARCO fractures to earthquake exacerbation or “triggering”
of “internal stresses generated during the original fabrication process.” The potential for
cracking or tearing due to weld cooling in restrained conditions is well-recognized (Daniels and
Collin, 1972; Putkey, 1993). Recent studies support the hypothesis that welding-induced
residual stresses can reduce plastic deformation capacity and promote brittle fracture of the type
seen after Northridge (Zhang and Dong, 2000). With respect to this particular building,
however, other experts discount the likelihood that pre-earthquake residual stresses would have
led to fractures under the relatively small additional effects of the earthquake (Tide, 2000). If the
fractures had initiated before the earthquake, however, they might have grown during the
shaking (Tide, 2000). Whether or not the earthquake made any contribution to the observed
damage, the atypical framing conditions at the second and third floors clearly played a role, as no
fractures were found anywhere else in the building (Phipps, 1998).
5.8

Mexico City, 1985

Osteraas and Krawinkler noted that “the 1985 earthquake was probably the first event in
which a significant number of steel buildings, including modern ones, were subjected to a severe
test.” A 1986 damage survey counted about 100 steel structures in Mexico City, including about
sixty built after Mexico’s benchmark 1957 earthquake (Martinez-Romero, 1986, cited in
Osteraas and Krawinkler, 1989). Nevertheless, none would be considered typical of preNorthridge WSMF practice in California. The 1985 performance of modern steel structural
systems in Mexico City, as reported by Osteraas and Krawinkler (1989), is summarized in Table
5-4.
Osteraas and Krawinkler cited poor construction quality in most of the damaged buildings
surveyed by Martinez-Romero, and drew the general conclusion that “in most cases, the damage
in the post-1957 [steel] structures was minor to moderate.” They studied three steel frame
structures in detail.
The first, 77 Amsterdam Street, was an 11-story structure built around 1970. Its beams and
columns were built up from channels and plates to form box columns and I-beams. The beamto-column connections consisted of a cover plate fillet welded to the beam flange and full
penetration welded to the flange plate of the box column. The beam web was attached to the
5-17

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

column plate with a bolted and partially welded shear tab. Failures in this joint typically
occurred in the vertical fillet welds connecting the box column flange plates to the column web
channels. The full penetration weld from the beam flange plate to the column did not fail. While
interesting, the observed failure mode does not give significant insight into the pre-Northridge
WSMF problem. Osteraas and Krawinkler computed that the connection was barely able to
resist gravity loads considering the types and eccentricity of the welds.
Table 5-4

Performance of Modern Steel Structural Systems, 1985 Mexico City
Earthquake

Structural system

General structural performance

Remarks

Moment resistant
frame

41 surveyed, all at least 12 stories.
1 with severe damage
1 with repairable damage
3 with minor damage

Typical MRF has box columns, rolled
beams up to W18 or truss girders.
Damage “concentrated at welded beam-tocolumn connections or in truss girders.”

Moment resistant
frame with braced
bays
(similar to UBC Dual
System)

17 surveyed
2 total collapses, 1 partial collapse
4 with structural damage

Almost all reported damage was at the
Pino Suarez Complex.

Steel frames with
concrete shear walls

21 surveyed
1 with significant damage
3 with minor damage

Most steel damage to truss girders.
Source: Osteraas and Krawinkler, 1989.

As at 77 Amsterdam, the failures were not typical of those seen in the Northridge earthquake.
The connections of the girders to the columns were in most cases weak, and the combined
stresses on the flanges of the box columns were large. Generally, axial overstress of the columns
due to large brace forces is considered the most likely cause of failure. Indeed, following the
Mexico City earthquake, U.S. seismic provisions were changed to prohibit hinges in columns of
braced frames (SEAOC, 1988).
The second study addressed the Pino Suarez Complex. This group of five structures suffered
some of the most dramatic damage in the earthquake. A 21-story building collapsed onto a 14story building, and two other 21-story buildings sustained “severe structural damage,” with one
of them close to collapse.
The Pino Suarez buildings used a combination of steel moment frames and braced frames.
Box columns were formed from four welded plates. Girders were built up sections consisting of
plates welded to angle sections forming a flange, with diagonal angle webs. The flange plates
were welded to horizontal shear tabs, which were welded to the columns. Figure 5-2 shows a
damaged column. Damage was typically in the box columns, which buckled and were probably
the ultimate cause of the collapse and near collapse of the 21-story structures.

5-18

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Figure 5-2

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Damage to Steel Frame Column, Mexico City Earthquake, 1985
Source: EERI Annotated Slide Collection

The third building was Torre Latino Americana, a 44-story structure built in 1956 with large
built-up wide-flange columns and I-beams. Moment connections were made with all riveted Tflange and web tabs as shown in Figure 5-3. No structural damage was noted in the building,
which according to Osteraas and Krawinkler came “as no surprise [considering the building was]
a well designed long-period structure.” This tower also experienced earthquakes in 1957 and
1962, apparently without any damage obvious or remarkable enough to have been noted in a
1962 presentation on its instrumentation (Zeevaert, 1962).
5.9

Loma Prieta, 1989

The Loma Prieta earthquake caused connection damage in several steel buildings in the San
Francisco Bay Area. Some damage to architectural finishes was observed immediately following
the earthquake, but in all but one case investigation of beam-column connection failures was not
initiated until after the Northridge earthquake more than four years later. It is estimated that
inspections have been performed on about thirty buildings. Most were made during pre-purchase
investigations. Some were required by refinancing, and a few others were initiated at the request
of concerned building owners. Most of the investigations relied on ultrasonic testing in addition
to visual inspection (Phipps, 1998).
Of the buildings inspected, five were found to have damage. The three most heavily damaged
buildings are all located on soft soil where ground accelerations exceeded 0.20g. Each of the five
buildings is at least 35 miles from the epicenter of the earthquake (Phipps, 1998).
5-19

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Figure 5-3

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Typical Undamaged Joint in Torre Latino Americana, Mexico City
Earthquake, 1985
Source: Osteraas and Krawinkler, 1989

The SAC Joint Venture brought this damage to the attention of Bay Area building officials
and engineers with a special notice (SAC Steel Project, September 1996). The damage to the five
buildings is tabulated and described in Table 5-5. Figure 5-4 maps the buildings’ approximate
locations.
Table 5-5

Damage to WSMF Buildings in the 1989 Loma Prieta Earthquake
Connection damage

Building description

Distance from
epicenter (mi)

(% of joints inspected)
Longitudinal

Building 1
6-story, 200,000 sf, 1989

Approximate
repair costs

Transverse

37

<10%

50%

$2,500,000
(FEMA-267)

57

15%

None

$630,000

Building 2
12-story, 234,000 sf, box
columns, 1978
Building 3
20-story, 400,000 sf
Building 4
20-story
Building 5
14-story, under construction

38

5% total

$300,000

53

One connection

Not available

50-60

28 connections total

Not available
Source, unless noted: Phipps, 1998.

5-20

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Figure 5-4

FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Location of WSMF Buildings with Known Connection Damage, Loma Prieta
Earthquake, 1989

The following building descriptions are taken from Phipps (1998).
Building 1 is located on bay mud along the edge of San Francisco Bay. Constructed in 1989,
it was nearly complete at the time of the earthquake. Peak ground accelerations of 0.29g (E-W)
and 0.26g (N-S) were recorded approximately three miles from the site during the earthquake.
Immediately following the earthquake, some minor cracking of architectural finishes was
observed in the building and some movement of the glazing within window frames was found.
The steel frame was not investigated until 1996. Initially, about sixteen connections were
visually inspected. Twelve had readily observable damage. The investigation was expanded,
and ultimately a total of 107 damaged connections were identified. Damage included girder
bottom flange fractures, column divots, panel zone fractures, and girder top flange fractures.
Building 2 is also located on the edge of San Francisco Bay. A peak ground acceleration of
0.26g was recorded approximately 0.6 miles from the site. Minor damage to architectural
finishes was observed and repaired following the Loma Prieta earthquake. In 1996,
approximately 210 connections were inspected, and 41 were found with damage. Repairs were
made using procedures consistent with the recommendations of FEMA-267.
5-21

 FEMA-355E
Chapter 5: Performance of Steel Frame Buildings
in Past Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Building 3 is located on soft soil, again along the edge of the Bay. The nearest strong motion
data, recorded five miles from the site, gave peak ground accelerations of 0.26g (N-S) and 0.29g
(E-W). In 1998, the steel frame was investigated as part of a pre-purchase due-diligence survey
following procedures of FEMA-267. One hundred and fifty connections were visually and
ultrasonically inspected. Damage was found in eight connections at the third, fourth, and fifth
floors.
The Building 4 site experienced an estimated PGA of 0.18g, recorded about one mile away.
An inspection of the building immediately following the earthquake revealed damage to one
WSMF connection in a stairwell. The observed damage was reported to be similar to the
damage found after Northridge. After the Northridge earthquake, a limited investigation of the
frames was undertaken, and no additional damage was found.
Building 5 was under construction at the time of the Loma Prieta earthquake. It is located in
downtown Oakland and uses a dual system of concrete shear walls and WSMFs. The nearest
free-field record was located less than a mile from the building, and the recorded peak ground
acceleration was 0.18g. During the earthquake, about half of the freestanding columns fell over,
and almost all of the perimeter beams at the fifteenth and sixteenth floors fell off their bolted
seats. From the seventh to the tenth floor, several of the bottom flange welds in the N-S WSMF
cracked. A total of twenty-eight cracked welds was discovered by visual inspection or NDT. It
was reported that the vast majority of the cracked welds were made by one welder, who had been
fired prior to the earthquake, and that most of the welds had not been inspected. Detailed
investigations conducted on the damaged steel connections revealed cracking of the same types
found in the Northridge earthquake. Samples taken from the building showed a lack of fusion at
the root pass, particularly in the E-W moment frames, which had been inspected prior to the
earthquake.
5.10

Landers and Big Bear, 1992

The following information is taken entirely from Anderson and Bertero (1997).
The 1992 Landers earthquake was the second largest in California in the 1900s, measuring
7.3 on the Richter scale. It was closely followed by a magnitude 6.8 event about twenty miles
away in Big Bear. Damage, however, was not as significant as in either the Loma Prieta or
Northridge earthquakes because the epicenter was far from a major metropolitan area.
Nonetheless, damage estimates exceeded $100 million.
A two-story steel frame building was damaged in the city of Big Bear Lake, although damage
to the steel frame was not recognized immediately. Only after Northridge was the building
studied in more detail. It was built in 1986 with several lines of moment frames in each
direction. Cracks were found in many of the moment frame connections. They were severe
enough that the owner decided to retrofit the building with new braced frames, rather than restore
the frame to its original strength and stiffness. Near field effects and directivity may have
contributed to the observed damage.

5-22

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

6. PERFORMANCE OF WSMFs IN THE 1994 NORTHRIDGE
EARTHQUAKE
The 1994 Northridge earthquake was the event that triggered the nationwide study of WSMF
seismic performance. Confidence in the performance of steel frame buildings seemed to increase
incrementally with each major earthquake from 1906 to 1992. As described above, however,
observations from past events and from research prior to 1994 suggest that the problems
observed after Northridge were not new and should not have been wholly unexpected. Still, the
quantity and severity of the 1994 damage was disturbing. “The Northridge earthquake of
January 17, 1994, has fundamentally shaken engineers’ confidence in the seismic performance
and safety of WSMF buildings” (Mahin et al., 1996).
By mid-1994, the engineering and research community had outlined a program of data
collection to determine the extent of the problem. In the years since, dozens (if not hundreds) of
articles, technical papers, and research reports large and small have been published on the topic
of steel frame connections alone. Many, including even some studies sponsored by SAC, were
based on preliminary data and are already obsolete. Others were prepared concurrently and are
therefore incomplete, unable to reference contemporary findings, whether supportive or contrary.
This is to be expected. In a few more years, perhaps, we will be removed enough from the event
to recognize its lasting lessons. For now, this section summarizes briefly the latest data
available, including damage counts. Readers are urged to consult the original sources and
authors.
Without question, pre-Northridge WSMFs must now be considered more vulnerable than
they were thought to be. But much of the research since the event paints a less dramatic picture
than the one that emerged in 1994. Some important Northridge lessons that can be drawn at
present are:
•

Over the population of WSMF buildings, actual earthquake damage was far less extensive
than suspected in mid-1994. Nearly half of all inspected buildings had no connection
damage at all.

•

Despite a low overall damage rate, some WSMF buildings were left in damage states that
must be considered hazardous. Those with very high fracture rates were left vulnerable to
future earthquakes. Those with fractured column webs or severely damaged shear
connections may have posed local collapse hazards in aftershocks.

•

The quality of pre-Northridge welding and inspection was poor in enough buildings that it
cannot be considered an anomaly. Even if nonconforming practices did not directly cause
damage, their widespread presence indicates how engineers, contractors, and inspectors had
not been in complete control of WSMF design and construction.

•

For typical WSMF buildings, the presence of connection damage is not predictable from
broad building attributes known in advance (height, frame configuration, age, etc.).

•

Structural connection damage cannot be ruled out in the absence of nonstructural
“indicators,” such as elevator or partition damage. Damage does correlate mildly with broad
6-1

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

ground motion parameters, but not closely enough to raise the threshold for postearthquake
inspection. Damage is predictable from analysis, but only in a probabilistic sense; the
relationship is strong enough to aid postearthquake inspection in large buildings.
6.1

Early Findings and Engineering Response

Welding contractors working on buildings still under construction were the first to discover
the damaged WSMF connections (Gates and Morden, 1995; Buildings 9070 and 9054 in
Appendices A and B). “Damage was first identified by examining the distance between the
back-up bars and columns in buildings that were visibly damaged. Others reported that in some
buildings, elevators were not functioning properly, and in the process of examining them, the
damage to welded joints was discovered” (Goltz and Weinberg, 1998). SAC case studies by
Green and Hajjar et al. (see below) have described the process of finding unexpected fractures.
Typically, there were no obvious signs of structural distress. An April 1994 article described
some weld damage discovered “during routine tenant improvement work” (Modern Steel
Construction, 1994).
Some early published reports underestimated the damage, capping it at “perhaps as many as a
dozen” structures, and noting that it had mostly been repaired within a few months of the
earthquake (Modern Steel Construction, 1994; AISC, 1994). These figures were low; published
estimates that followed were almost certainly high. Gates and Morden (1995) have described
how the damage count rose steadily through 1994. Most interesting and instructive, however, is
how the count continued to rise through estimates, speculation, and misunderstanding. First,
nearly all inspections through 1994 counted weld flaws as damage. As described below, these
“W1” flaws are now considered pre-existing conditions. But at the time, they accounted for
more than half of all the “damage” found. Several buildings were considered extensively
damaged even though only W1 flaws had been found.
Second, a tentative list of WSMF buildings identified by the City and a list of buildings
scheduled for inspection were probably both misunderstood at one time or another as lists of
damaged structures. Although the two leading testing firms in Los Angeles estimated in early
1995 that they had already inspected about 200 steel frame buildings between them (Gates and
Morden, 1995), that number is almost certainly incorrect. At that time, inspections in Los
Angeles had not yet been mandated, and even with contributions from twenty-five engineering
firms, the SAC survey had identified fewer than 100 inspected buildings, many of which were
undamaged (Bonowitz and Youssef, 1995). Nevertheless, articles and presentations too
numerous to mention, even some by highly knowledgeable authors, managed to repeat the phrase
“over 100 damaged buildings” as established fact. An article in the October 1996 newsletter of a
Northern California engineers association even claimed two hundred damaged WSMFs.
The inflated damage counts, as well as the mistaken impression among some engineers that
the worst damage patterns were typical, certainly changed some minds about steel frames. A
year after the earthquake, despite no steel frame collapses or casualties, some engineers familiar
with the issues had come to regard WSMFs as less likely to provide Life Safety performance
than well-designed concrete shear wall buildings (Gates and Morden, 1995).

6-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Now, the figure of a hundred verified damaged WSMF buildings is probably correct. By May
1998, the City’s records indicated only about 230 inspected buildings, 90 of which reported no
damage at all (Los Angeles Department of Building & Safety, 1998). Many more reported only
one or two “damaged” connections from their first phase of inspection; undoubtedly, many of
these had weld flaws only. Inflated figures are still frequently cited. Even SAC reports have
recently cited over 200 Northridge-damaged steel frames, a figure that is unsupported (Goltz and
Weinberg, 1998).
If the damage numbers were exaggerated, the findings of inadequate fabrication and welding
probably were not. Postearthquake joint inspections revealed widespread conditions of poor fitup, improper joint preparation, undersized weld access holes, unacceptable weld weaving and
bead thickness, use of weld dams, and other nonconforming practices. Some inspectors felt that
the poor construction quality was especially prevalent in low-rise buildings (Gates and Morden,
1995).
6.2

New Regulation

In response to the damage (real and imagined) and unexpected questions about in-place weld
quality, the City of Los Angeles organized a Steel Frame Building Task Force of local structural
engineers and Building & Safety Staff. Membership quickly grew to include researchers,
contractors, steel and welding industry representatives, and the Building Owners & Managers
Association (Gates and Morden, 1995). Task Force members shared, confidentially, the data
available to them, and developed tentative procedures for inspection, evaluation, and repair.
Many of these are reflected in SAC Advisory 3 (SAC 95-01). Research had also begun.
Principal among these efforts were full scale tests of connection specimens that matched the
conditions where some of the first fractures were found (Engelhardt and Sabol, 1994) and a
systematic and centralized data collection effort (Youssef et al., 1995).
Meanwhile, the City had ten subcommittees and Task Forces looking into the performance of
other structural materials and systems. By the end of 1994, the Department of Building and
Safety would enforce emergency measures regarding wood frame construction, reinforced
concrete structures, and tilt-ups, as well as WSMFs (Deppe, 1994). As a result of Northridge,
the City would ultimately adopt new code provisions for voluntary earthquake hazard reduction
in hillside structures, wood cripple walls, and infilled concrete frames (ICBO, 1999).
The many WSMF-related technical guidelines, directives, and code interpretations to arise
from the early efforts of the Los Angeles Task Force and others included the following. These
were eventually compiled into or superseded by SAC Advisory 3 (SAC 95-01) in February,
1995, the SAC Interim Guidelines (FEMA 267) in August, 1995, and subsequent building codes
and standards.
•

City of Los Angeles, March 18, 1994: The first post-Northridge requirements for welding in
repair and new construction, calling for reinforcing fillet welds after removal of backing bars,
and specifying “small diameter wire electrode” for new full penetration welds.

•

City of Los Angeles, May 11, 1994 (revision): Specific welding and procedure requirements
for repair and new construction.
6-3

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

•

County of Los Angeles, July 25, 1994: “Emergency Regulations” for repair, and suspension
of the prequalified connection for new construction.

•

City of Los Angeles, August 1, 1994: “Effective immediately, the use of Section 2710(g)1B
of the Uniform Building Code for the design of girder-to-column connections is suspended.”

•

Building Standards (1994): A one-page article describing “the emergency code change action
taken by the ICBO Board of Directors on September 14, 1994.” The code change essentially
replaces the prequalified WSMF connection with a requirement for designs supported by
cyclic test results.

•

City of Los Angeles, December 27, 1994: Interdepartmental Correspondence setting forth
City building department requirements for welding in repairs and new construction.

•

DSA, March 17, 1995: Interpretation by the California Division of the State Architect,
commenting on the recent deletion from the state building code of the prequalified WSMF
connection. This 34-page document includes some of the earliest post-Northridge cyclic
testing requirements and acceptance criteria.

•

AWS (1995): Recommendations of an AWS Task Group for changes to the D1.1 Structural
Welding Code regarding WSMF weld design, fabrication, etc.

After the Interim Guidelines came a number of model building codes and standards that
addressed aspects of WSMF design with a post-Northridge perspective and in light of early postNorthridge research. Principal among these were FEMA-267A, a supplement and update of the
Interim Guidelines, and the AISC Seismic Provisions (1997). The AISC provisions included an
appendix dedicated to cyclic testing of beam-column connections.
The Reliability/Redundancy Factor, ρ (rho), in current building codes was motivated in part
by early observations of Northridge WSMF damage. The ρ factor represents a substantial
change in the codified seismic design philosophy for moment-resisting frames. It was developed
and introduced concurrently in the 1996 SEAOC Blue Book, the 1997 UBC (ICBO, 1997), the
1997 NEHRP provisions (FEMA 302 and 303), and FEMA-267A. The commentaries to those
documents describe its motivation and its intended effect. For a typical low-rise WSMF office
building with a floor plate of about 25,000 square feet, the UBC ρ requirements would
effectively require frames providing 12 to 16 WSMF connections in each principal direction at
each floor (for example, three or four two-bay frames, or six one-bay frames).
While ρ was motivated by general trends in the design of all types of structures, the 1996
Blue Book commentary refers specifically to “findings by the SAC joint venture (sic).” Indeed,
some of the first WSMF damage found after the earthquake was in a five-story building in
construction with one-bay frames (building 9070 in Appendices A and B). However, as
described below, further work with Northridge damage data has not found a useful correlation
between damage and structural redundancy.
6.3

Social, Economic, and Political Effects

Changes in WSMF engineering practice were accompanied by financial, legal, and political
effects. Perhaps the first of these was a change in working relationships between engineers,
6-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

contractors, and inspectors. Market demands put some immediate pressure on steel contractors,
stalled designs of new steel frames, and spurred development of new proprietary details. The
City ordinance mandating inspection of some WSMFs offered useful lessons on politics and the
legislative process. Ultimately, financial losses led to lawsuits.
6.3.1

Changes in Practice

Goltz and Weinberg (1998) investigated the ramifications of sudden WSMF damage on the
market for qualified welders, steel fabricators, and inspectors. They concluded that the regional
market was able to withstand unexpected pressures, in part because Los Angeles’ mandatory
inspection ordinance (discussed below) was limited in scope and was implemented in phases.
At the same time, inspectors, welders, and engineers came to realize how little they knew
about each other’s work. Engineers, for example, did not know why E70T-4 electrodes were
routinely used or to what effect backing bars were left in place. Welders did not know the
specific requirements for the processes and electrodes they were using. Neither did the inspectors
who checked their work. Engineers could not gauge the significance or the validity of ultrasonic
test reports. None of the three groups was sufficiently familiar with the Welding Procedure
Specifications they were each supposed to have approved. Indeed, production of a WPS had
been recognized as essential but often overlooked (Putkey, 1993). Much of this changed, or was
expected to, after Northridge. (For more discussion on the nature of early findings and changing
relationships between various parties, see Gates and Morden, 1995; Goltz and Weinberg, 1998;
and SAC 95-01.)
Uncertainties in the wake of the earthquake changed the relationships between engineers and
building owners as well. The technical questions—Are the frames damaged? Are the damaged
frames safe?—had no certain answers and quickly led to questions of liability, insurance, due
diligence, construction scheduling, and budgets. Engineers could not perform limited intrusive
investigations, recommend remedial measures, or even complete in-progress designs with any
assurance that local building departments would not later require something different. Except for
repairs to the most severely damaged frames, the result was an industry-wide “wait and see”
policy. While engineers and owners waited, insurers were apparently not bothered; Goltz and
Weinberg report a “disappointing level of detachment” among the insurance professionals they
surveyed.
New requirements for connection qualification tests affected new WSMF design throughout
California for several years after the earthquake. With no standards or consensus acceptance
criteria, building departments were reluctant to approve designs. Fast track projects could not
afford the time it took to design and execute a series of tests. And if any of the tests were to fail,
the cost of redesign and retesting could kill a project. Developers and engineers turned away
from steel frames to other “proven” systems. By late 1996, a body of successful tests had
developed, and some engineers began to cite the available results in support of new designs.
Unfortunately many of those tests had intentionally chosen large member sizes to match the 1994
Engelhardt and Sabol tests. As a result, a disproportionate number of early tests involved heavy
W36 beams and jumbo columns. New designs relying on those tests were thus constrained to the
tested sizes. More important, the eventual requirement to provide two or three matching tests
6-5

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

ignored questions of reliability, just as engineers and researchers had done in the years before
Northridge (SEAONC, 1998; Bonowitz, 1999b).
Technical solutions had interesting nontechnical aspects as well. A few innovative engineers
designed, tested, and began to market alternative connection details for both retrofit and new
construction. To offset their costs, they sought patents for their designs. Two proprietary
designs that would later undergo peer reviews and receive approvals from Los Angeles County
and other permitting bodies were the “slotted web” detail (SSDA, 1996) and the SidePlate
system (LACOTAP, 1997). Trade Arbed, a Luxembourg steel manufacturer, had patented a
beam with shaved flanges for seismic applications in the 1970s but had never enforced the
patent. The “dog bone” or “reduced beam section” (RBS) designs that would emerge later were
similar to Arbed’s concept. Taiwanese researchers had also developed a proprietary RBS design.
But the idea of a patented steel connection, to be designed by a sole-source contractor, was
relatively new to California engineers (even though proprietary technologies for isolation and
energy dissipation had been on the market there for years).
While the proprietary designs were in development, the rest of the steel construction and
engineering world was following the recommendations of SAC, a federally funded joint venture.
Because of its limited public funding, SAC did not support or participate in these proprietary
efforts. However, SAC documents did acknowledge and refer to them.
6.3.2

Legislation and Public Policy

Building code changes and interpretations motivated by the Northridge steel fractures are
discussed above. Nontechnical public policy actions that arose from WSMF issues included the
following.
California Assembly Bill 3772 was signed into law after the 1995-96 legislative session.
According to an engineering association newsletter, the law allows the California Building
Standards Commission “to adopt emergency regulations outside the regular cycle for adopting
building codes. The need for this measure was precipitated by the failures in steel moment frame
structures after the Northridge earthquake” (SEAOC Plan Review, 1996).
Through the spring of 1994, new construction was proceeding with the old detail even in Los
Angeles, as building departments had no authority and insufficient data to support a moratorium.
On July 21, the L.A. City Council passed emergency ordinance 169949, granting “blanket
authority” to the Department of Building & Safety (EERI, 1996). By August, the City and
County had proscribed the prequalified pre-Northridge connection. In September, it was
removed from the Uniform Building Code (Building Standards, 1994).
In Los Angeles, however, it also remained to find the damage and mitigate the hazards. The
damage count was still rising, and many WSMF owners were understandably reluctant to have
their buildings inspected. Their buildings were functional, and their nonstructural damage had
been repaired. Finding damage would leave them in financial limbo until consensus repair
standards could be developed (Smith, 1995). Building officials saw a need for mandatory
inspection.

6-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

On February 22, 1995, the Los Angeles City Council passed ordinance number 170406
mandating connection inspections and repairs in some WSMF buildings. The story of its
passage has been told by Gates and Morden (1995) and by an EERI White Paper (1996), from
which the following chronology is taken:
•

January 1994: The Department of Building & Safety (B&S) establishes task force groups to
study performance of various building types. With the expectation of little damage, the steel
frame task force is assigned to the busy Chief of the Building Bureau, Richard Holguin. Each
task force head reported to Councilman Hal Bernson.

•

Spring 1994: The Building Owners and Managers Association (BOMA) requests a seat on
the steel task force group. BOMA advocates case-by-case inspections and repairs, as opposed
to broad prescriptive procedures.

•

June-July 1994: B&S consults with Bernson and drafts an ordinance mandating inspection
and repair of all WSMFs in the city. Efforts stall when early tests suggest that simple repairs
may be costly and no better than the previous condition. High repair costs are expected for
residential buildings (which have more expensive finishes), and condominium owners
contest the proposed repair ordinance.

•

June 1994: For political viability, B&S cuts the geographic scope of the proposed inspection
ordinance to areas where serious damage had already been found, and estimates that the
number of potentially affected buildings would drop from 1000 to 300.

•

July 1994: Under pressure from owners groups, B&S drops residential buildings—only about
30 out of 300—from the proposed ordinance.

•

October 1994: B&S modifies the proposed ordinance, lengthening the time to comply. A
minimum inspection scope is also removed to protect the City from potential liability in the
event that no damage would be found.

•

Autumn 1994: B&S lobbies individual Council members. Bernson now feels that the
ordinance is too weak and not prescriptive enough, and he too must be convinced to schedule
a vote.

•

February 1995: With the 1995 Kobe earthquake as an anniversary reminder, the Council
passes the ordinance by a 12-0 vote.

Ordinance 170406 has been incorporated into the City of Los Angeles building code as
section 8908. Damage data collected as a result of Ordinance 170406 is discussed below and
included in Appendices A and B. The final ordinance covered a geographic area that excluded
some parts of the city with high concentrations of WSMFs. In particular, the highrise buildings
downtown, some of which had been analyzed after the San Fernando earthquake, were excluded.
The included areas were chosen as those where significant damage had already been found in
WSMFs or other structure types. Although the shaking was lighter downtown than on the west
side or in the San Fernando valley, it is possible that no damage was found in downtown
structures because so little inspection had been done there early in 1994.

6-7

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

The Los Angeles County building code (Chapter 94) would later adopt a similar inspection
and repair requirement. The county defined a geographical area that included Universal City and
areas adjacent to Santa Clarita.
There are dozens of small, incorporated jurisdictions in and around the City of Los Angeles.
Most have only a small number of WSMF buildings, if any. As of April 2000, some of the
jurisdictions with or near known WSMF damage had taken the following steps:
•

Burbank in 1998 adopted section 7-140 into its Municipal Code, requiring inspection per
FEMA and SAC Guidelines and repair to the pre-earthquake condition, upon notice by the
building official (City of Burbank, 1999). A total of ten notices were sent out, and about half
of the notified buildings reported some significant damage (Sloan, 2000).

•

Santa Monica adopted chapter 8.76 of its Municipal Code Article 8 in June, 1999. The
provision requires inspection and demonstration of conformance to the latest FEMA and
SAC Guidelines (City of Santa Monica, 1999). The time allowed for repairs is given as a
function of the number of occupants. As of April 2000, the requirements had only been
applied to WSMFs seeking permits for other work. The Building Official expects proactive
notification of WSMF building owners to begin in July 2000 (Mendizabal, 2000).

•

Glendale has not mandated any inspections. All WSMF connections in City buildings were
inspected since the earthquake, however, and no damage was found. The Glendale Building
Official estimates that his jurisdiction has more and taller WSMFs than Burbank (Tom,
2000).

•

Santa Clarita, which has fewer than ten WSMFs, has not mandated inspections, but the
Building Department did send letters advising inspections and did issue several permits to
repair weld damage (Bear, 1999).

•

San Fernando (Mendoza, 1999), Beverly Hills (Moon, 1999), and Simi Valley (McDonald,
1999) have not required inspections.

Outside of Southern California, the Northridge damage prompted investigations of some
WSMF buildings that had been subject to strong ground motion in the 1989 Loma Prieta and the
1992 Landers and Big Bear earthquakes (see the section above on past earthquakes). While no
inspections have been mandated, the fact of observed damage has been brought to the attention
of Bay Area building officials and structural engineers (SAC Steel Project, September 1996).
Once the Los Angeles ordinance went into effect, the costs of repairs became a real issue.
B&S staff speculated in 1995 that they might have to identify and develop financing options in
order for the program to succeed (EERI, 1996). Indeed, in 1997, the City Council approved an
“unusual” $200 million bond issue specifically to fund long-term loans for mandated WSMF
repairs (ENR, January 27, 1997). Los Angeles had been through mandatory seismic hazard
reduction before, when it addressed its thousands of unreinforced masonry (URM) buildings.
Interestingly, URM retrofit costs fell as contractors gained experience; with WSMFs, whose
repairs were mandated before any standards had been proven, costs rose (EERI, 1996).

6-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

6.3.3

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Legal Implications

Financial losses to some WSMF building owners led to litigation involving design engineers,
contractors, fabricators, erectors, electrode manufacturers, and insurance companies. Most cases
concerned individual buildings, but one alleged $1 billion in damages to WSMF structures as a
class. This class action was later voluntarily withdrawn by the owners who had initiated it. As of
December 1999, no court had rendered a decision on any legal claim, but several lawsuits have
been settled. About eight lawsuits remain pending.
The following details from the first of the individual lawsuits and the aborted class action
(ENR, January 20, 1997 through February 10, 1997; September 1, 1997; February 23, 1998)
offer some indication of the non-engineering response of the courts and various stakeholders to
the unanticipated WSMF damage.
In April 1995, during repair of a five-story WSMF in Santa Monica, workers discovered
previously undetected damage. The structural engineers assessed the newly found cracks and
called for evacuation of the building’s tenants. (The building is number 9017 in Appendix A.) In
January 1996, the owners filed suit against the original general contractor, steel fabricator,
inspector, and structural engineer, seeking $10 million. What followed:
•

May 1996: The Santa Monica plaintiffs amend their suit to include Lincoln Electric, the
nation’s leading maker of welding materials, citing the prevalent E70T-4 electrode as a factor
in the damage. A jury trial is scheduled for May 1997.

•

January 1997: Lincoln is sued in a $1 billion class action. By September, Lincoln would be
named in seven other individual building suits.

•

August 1997: Lincoln settles with the Santa Monica plaintiffs for $6 million. The suit against
the original defendants remains, with a trial scheduled for January 1998. Lincoln attempts to
recover its loss from the other defendants, but later abandons the effort.

•

February 1998: The remaining Santa Monica defendants settle for a combined $5.5 million.

•

February 1998: Lincoln remains a defendant in eight lawsuits. The class action is eliminated,
perhaps, ENR speculates, because identical causation could not be shown.

The March 1997 issue of California Construction Law featured a series of articles that
debated the charges against Lincoln (Castro, 1997; Jenks and Ritts, 1997). They offer a
decidedly non-technical perspective on research, design, and the meaning of structural
performance. Comparing pre-Northridge connections to faulty automobile airbags, Attorney Joel
Castro presents a variety of “theories of recovery” premised on the claim that Lincoln’s
ubiquitous E70T-4 electrode was defective. In particular, he argues that mere repair, like tape
over a punctured airbag, is inadequate compensation. He quotes (without citation) from
Lincoln’s marketing materials: “Where buildings must be designed to withstand seismic
disturbances, Innershield is the architect’s choice.” (Innershield is the name of a Lincoln product
line that includes their E70T-4 electrode.) The brittle weld metal and Lincoln’s support of it
were “substantial factors” in the damage, he argues, so Lincoln bears responsibility.

6-9

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Lincoln’s attorneys respond that too many factors were at work for the weld metal alone to
be held responsible. Besides, they argue, E70T-4 specifications never included a notchtoughness requirement. Taking up the airbag analogy, they characterize the pre-Northridge
connection as an accident waiting to happen:
Imagine a person driving down an icy mountain road on a foggy winter night. He is
speeding, his tires are bald, his brakes are worn, he is not wearing a seat belt. When he
bought the car he chose to buy an ordinary air bag, not the heavy-duty one that would
have cost an extra $1,000. He slides into a guard rail at a speed guaranteed to produce
injury whatever the air bag chosen. He is injured—and he sues the airbag manufacturer.
The lawsuit has no chance of success.
6.4

Damage Data

How bad was the damage? This has been a principal question since the first fractures were
discovered in the spring of 1994. Unfortunately, the verifiable answer has changed as the scope
of the problem became known, as inspections went from voluntary to mandatory, as certain
damage types were discounted, and as analyses and case studies attempted to describe damage in
the context of structural performance. The speculative answer was even more difficult to
quantify, as photographs of the most severe damage circulated and as lists of buildings scheduled
for inspection were mistaken for lists of buildings damaged.
In mid-1999, SAC researchers compiled and cross-checked a master list of over 200 WSMF
buildings inspected after the Northridge earthquake (Maison and Bonowitz, 2000). The source
lists, described in Table 6-1, varied in their size, completeness, and intended use. Appendices A
and B give the master list and a building by building damage summary. Tables B.2 and B.3
summarize the damage data from the master list.
Table 6-1

Source Lists of WSMF Buildings Affected by the 1994 Northridge
Earthquake

Reference

Sponsor

Scope

Notes

Los Angeles
Department of
Building and
Safety, 1998

City of Los
Angeles

City of L.A., about
220 buildings in
various stages of
inspection

Departmental record tracking mandatory inspection and
repair. Commercial buildings only. Specified areas within
the city only. Hard copy data available from LAB&S in
May 1999 was current only through May 4, 1998.

Youssef et al.,
1995

NIST

All available data,
51 buildings

The first systematic post-Northridge data collection effort,
from August through November 1994. Voluntary
inspections only, with various inspection scopes.

Bonowitz and
Youssef, 1995

FEMA,
SAC

All available data,
79 buildings

Continuation, expansion, and completion of the NIST
effort, ending in March 1995. Voluntary inspections only,
with various inspection scopes. Damage reported for each
set of connections in a “floor-frame.”

Durkin, 1995

FEMA,
SAC

Random survey of
150 buildings
within 0.2g contour

Intended to characterize the local WSMF population and
the response of building owners by late 1994.

6-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 6-1

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Source Lists of WSMF Buildings Affected by the 1994 Northridge
Earthquake (continued)

Reference

Sponsor

Scope

Notes

Dames and
Moore, 1998

FEMA,
SAC

49 selected
buildings, most
with nearly
complete
inspection

Buildings in West L.A. and southern San Fernando Valley
selected based on available damage and construction data.
Data compiled for individual connections with specific
damage types noted. Some buildings overlap with
Bonowitz and Youssef, but Dames and Moore data likely
to be more complete and updated.

Paret, 1999

FEMA,
SAC

35 selected
buildings, most
with nearly
complete
inspection

Regular buildings in West L.A. and southern San
Fernando Valley selected for study of W1 causes and
effects. Data reported on a building level from review of
postearthquake inspection reports.

Durkin, 1999

FEMA,
SAC

100 randomly
selected buildings

Intended to characterize the local WSMF population, the
response of building owners, and repair costs and
approaches by late 1998.

Maison and
Bonowitz, 2000

FEMA,
SAC

Compilation of all
of the above

Compiled and cross-checked for loss estimation study.
Damage collected on a building level, with site specific
ground motion data added. See Appendices A and B.

The master list summarized here and given in Appendix A is believed to be representative of
the greater Los Angeles WSMF population. For purposes of regional impact studies and loss
estimation, Seligson and Eguchi (1999) used Assessor’s records to estimate the number of
WSMF buildings in Los Angeles County. They made some assumptions about age and lateral
systems and concluded that the steel buildings covered by the City of Los Angeles ordinance are
representative of the complete class of steel frame buildings identified from the Assessor’s
records. Therefore, since the master list is made largely from buildings covered by the L.A.
ordinance, the damage data may, for predictive purposes, be reasonably extrapolated to the wider
population of moment frames. Nevertheless, there are some significant differences to keep in
mind:
•

The Los Angeles ordinance mandated inspection of commercial buildings only. The collected
data might not be representative of other occupancies. For example, according to Seligson
and Eguchi, steel-framed residential structures have total floor areas much smaller than
typical office buildings. Some hospital buildings may also be missing.

•

The data summarized here (and listed in Appendix A) include about two dozen inspected
buildings from jurisdictions other than the City of Los Angeles, such as Santa Monica and
Santa Clarita. These buildings were inspected voluntarily, sometimes because there was
substantial nonstructural damage. They are more likely to have been damaged than an
average or random WSMF in the same area. On the other hand, because jurisdictions outside
the City of Los Angeles might not have mandated inspections, there may be damaged
buildings in those areas that are missing from the Appendix A database.

•

Though statistically representative, the Los Angeles ordinance data does exclude some parts
of Los Angeles with high concentrations of WSMF buildings, notably downtown, the eastern
6-11

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

half of the mid-Wilshire district, and the area around LAX airport. WSMFs in these areas
generally were not inspected after the earthquake.
•

By excluding the downtown area, the collected data probably does not offer a good data
sample for buildings taller than about twenty stories.

In Tables B-2 and B-3, the 155 buildings with at least 16 inspected WSMF connections are
counted. Small buildings with fewer than 32 connections are included if 50% of their
connections were inspected. This eliminates the few buildings whose inspection was so nominal
as to be deemed inconclusive. (Note that these screening criteria differ from those used in the
text and figures of Appendix B.) The damage rate is counted as the number of damaged
connections divided by the number of connections inspected. Each connection consists of two
beam flange welds and a beam web connection. The connection is considered damaged if there
is fracture at or near either the top or bottom weld, whether the cracking is in the beam, the
column, or the weldment. Weld flaws, labeled W1 or W5 in SAC documents, are not counted as
damage. These flaws, some of which are acceptable by AWS standards, are now widely
believed to have pre-dated the earthquake (Bonowitz and Youssef, 1995; Paret, 1999). Typical
W1 flaws are planar discontinuities at the weld-column interface, frequently up to ¼” in height
and as long as half of the weld length, sometimes longer.
6.4.1

W1 Flaws

The structural significance of W1 flaws is not yet fully resolved. The principal question is
whether original flaws increase the likelihood of earthquake damage. Readers are referred to
Paret (1999) and Kaufmann et al. (1997) for more on the detection and interpretation of W1
flaws. Some basic findings on the subject from post-Northridge data collection and analysis
include:
•

Until W1 flaws were shown to have predated the earthquake, they accounted for about two
thirds of all the so-called damage (Paret, 1999).

•

Within the greater L.A. population, one can now expect to find W1 flaws in about 15% of
existing WSMF connections (Bonowitz and Youssef, 1995; Paret, 1999). The original
occurrence rate was probably higher, since some original flaws certainly grew into fullfledged fractures in the earthquake (Paret, 1999; Kaufmann et al., 1997).

•

Ultrasonic testing (UT) is not well suited to finding W1 flaws due to technical limitations and
unreliable application (Paret, 1999). UT has been used for field inspection of WSMF welds
since the late 1960s, but UT findings have always been highly dependent on the skill of the
operator (Couch and Olsson, 1965; Preece, 1981; Preece and Collin, 1991). Nevertheless,
engineers and contractors appear to have relied on the technique almost exclusively, and this
may have invited abuse (Goltz and Weinberg, 1998). Under pressure during construction, UT
technicians may have been predisposed to read flaws as the reflection of the backing bar gap.
After the earthquake, technicians may have felt similar pressure to find damage.

6.4.2

Damage Data

Table 6-2 summarizes the damage data given in Appendix B. In Table 6-2, the building’s
damage rate, DR, is equal to the number of connections found damaged divided by the number
6-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

of connections inspected. Each connection represents the joining of one beam to one column. A
typical connection comprises two beam flange welds and a bolted or welded beam web. A
connection is considered damaged if any part of it is damaged.
The most typical damage involved fractures at or near beam flange-to-column flange welds.
In the worst cases, fracture extended through the column flange into the column panel zone. In
other critical cases, the beam flange detached from the column completely (or nearly so), and
damage to the beam shear connection—either the bolts or the shear tab—followed. These last
two damage classes, shear and panel zone, were especially rare; Table 6-2 gives the number of
inspected buildings in each height range with even one incident of shear or panel zone damage.
A few observations on the summarized data:
•

About 40% of buildings overall and in each height range had no connection damage at all.
Among the 1-story buildings, 11 of 13 were undamaged.

•

Overall and in each height range, the median damage rate is around 5%, and the worst
damage is around 50%. Even with the potential impact of clustered damage, it is unlikely that
more than four or five of these 155 buildings would have been classified as hazardous by the
2000 SAC Guidelines. (The Guidelines require a determination of damage severity at each
inspected connection and of the expected damage rate within each critical group of
connections. A hazard is recognized if any group has an expected capacity loss exceeding
50%. A more complete study of the damage with respect to “tagging” criteria is warranted.)

•

About a quarter of all damaged buildings had at least one damaged shear connection. The
mechanism of failure is such that shear damage never happens without flange fracture. In a
building with shear damage, typically less than 5% of the building’s connections and less
than a third of its damaged connections are affected, although in the hardest hit buildings the
numbers are higher. In many cases, typical shear damage does not affect the connection’s
gravity capacity, since most of the bolts are needed not to carry expected gravity loads but to
develop the beam’s flexural strength. The type of shear connection is not reported in much of
the collected data. However, buildings erected before 1975 are likely to have fully-welded
beam webs, and those built after 1988 are likely to have high strength bolts and supplemental
shear tab welds. Most of the surveyed buildings have shear connections with bolts but
without supplemental welds (Bonowitz and Youssef, 1995).

•

Shear connection damage in non-moment frame connections was observed in some heavily
damaged buildings (including the Borax building described below). Astaneh and Liu (1999)
have studied the deformation capacity of pre-Northridge single plate shear connections.

•

About a third of all damaged buildings had at least one damaged column panel zone. Panel
zone fracture only happens if the column flange fractures first. In a building with panel zone
damage, up to 50% of the damaged connections may involve some fracture into the panel
zone. In the worst cases, panel zone fractures severed the column over nearly its entire depth.
Most of the panel zone damage was less severe, typically involving a fracture that extended
just barely into the column web. Anderson, Johnston, and Partridge (1995) have studied the
residual capacity of severely damaged columns with panel zone fractures, using specimens
taken from their case study building, described below.
6-13

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

•

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

The data summarized here and presented in Appendix B have not been analyzed with respect
to the location of damage within a building. Previous studies with a preliminary data set
suggested that damage does tend to cluster, that 3- to 4-story buildings tend to have more
damage at lower floors, and that buildings taller than 18 stories had especially light damage,
if any, in their lowest eight floors (Bonowitz and Youssef, 1995).

Despite relatively low damage to the WSMF population as a whole, it is important to note
that there was serious damage to a wide variety of steel frame buildings. Among the most
heavily damaged were:
•

Building 9069 (see Appendices A and B), a one-story frame with column flange and panel
zone damage to about half of its twenty connections.

•

Building 9068, the Borax building described below, a four-story building with a 75% damage
rate and 21 panel zone fractures in its 112 connections.

•

Building 9017, St. John’s Medical Plaza (the subject of the lawsuit described above), a fivestory building with 50% damage and ten panel zone fractures in 96 connections.

•

Building 9008, a ten-story frame with 26% of its connections damaged, including 23 out of
688 with panel zone fractures.

Table 6-2

Number of WSMF Buildings with Various Northridge Earthquake Damage
Rates
1 story

2-4 story

5-12 story

13+ story

All

All buildings

13

69

47

26

155

No damage

11

26

16

12

65

0 < DR ≤.05

0

7

6

5

18

.051 < DR ≤ .10

0

10

8

1

19

.11 < DR ≤ .20

0

12

11

6

29

.21 < DR ≤ .50

2

13

4

2

21

DR > .50

0

1

2

0

3

Shear damage

0

9

10

4

23

Panel zone damage

1

16

8

4

29

6.4.3

Using the Damage Data

Collected damage data is most useful if it can address two questions: Where did damage
occur in the last earthquake, and where will damage occur in the next one?
6-14

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Appendix B shows the general scatter of damage rates with respect to demand. Additional
plots are included in Maison and Bonowitz (2000). The observed scatter is, of course, due in
large part to variability in design, construction, and material quality. Two other major
contributors to the scatter are the large number of buildings with no damage and substantial
uncertainty in demand estimates.
Preliminary statistical analyses of the Appendix B data were performed in mid-1999 for
purposes of developing loss estimation models (Maison and Bonowitz, 2000). The analyses first
considered whether building damage rates are significantly related to the following global
demand estimates: peak ground acceleration, peak ground velocity, spectral acceleration, spectral
displacement, spectral displacement divided by building height (a surrogate for drift), and
Modified Mercali Intensity. Derivation of the various building-specific demand estimates is
described in Appendix B and its references.
Given the demand estimates and known damage rates, one-sided t-tests determined whether
the mean demands of buildings in various damage ranges were significantly distinct from each
other. Chi-squared tests determined whether higher demands were associated with higher
damage rates to a degree greater than would be expected by chance.
Of the various demand estimates, peak ground acceleration (PGA) and spectral displacement
were found to correlate best with the observed damage. However, these damage-demand
relationships are clear in a probabilistic sense only. That is, there is no identifiable ground
shaking level above which WSMFs are damaged and below which they are not. Nevertheless,
the statistical link between damage and demand supports the notion of a demand-based
postearthquake inspection trigger. It also requires that studies of potential damage predictors
must use data with statistically equivalent demands.
Within a narrow demand range, are certain building configurations more prone to damage?
Analysis of the Appendix B data found no strong correlation at all between damage and building
height. While a previous analysis with preliminary data found a strong relationship in low-rise
buildings between damage and floor area per connection (Bonowitz, 1998), the 1999 analysis of
the updated and more complete data did not support that pattern.
Because the Appendix B data was compiled principally to study loss estimation, it was not
broken down by location in the building, by floor or frame, by member size, etc. Therefore, it
supports only very limited correlation studies. Analysis of SAC Phase 1 data has suggested
some useful relationships (Bonowitz, 1998), but failure of the newer, larger data set to support
one of them (area per connection, as noted above) suggests that the following should still be
considered preliminary:
•

More damage occurred in the lowest floors of 3- and 4- story buildings.

•

Connections with supplemental shear tab welds (as opposed to bolted or fully welded beam
webs) appear to have been more prone to damage. This may be an indirect predictor, as
supplemental shear tab welds were required only from 1988 to 1994 and only for the lightest
wide-flange sections of a given depth.

6-15

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

•

There is mixed evidence that 1- and 2-bay frames were more prone to damage than multi-bay
frames.

•

Far more damage was observed at beam bottom flanges than at top flanges. Composite
behavior is at most only partly responsible for the discrepancy. The relative ease of welding
and inspecting top flanges may explain some of the difference. Top flange fractures happen
at a greater rate in lab testing than the postearthquake field observations would predict. It is
also likely that much top flange damage was not found in buildings because, compared with
easily accessible bottom flanges, top flanges were less frequently and less completely
inspected.

•

After an earthquake, the general presence or absence of nonstructural damage or of nonWSMF structural damage is not a useful indicator of WSMF connection fractures. All of the
highly damaged WSMFs had both nonstructural and other structural damage, but the data
was not robust enough to be statistically meaningful (Bonowitz and Youssef, 1995).

If damage cannot be reliably found or predicted by obvious building attributes, can
postearthquake analysis help? A number of case studies have suggested that analysis can help
locate the areas of a building most likely to be damaged, but again, the damage-demand
relationship is probabilistic. Statistical analyses of case study data by Uang et al. and Naeim et
al. in SAC 95-04 have shown that beam ends with higher computed elastic demand-capacity
ratios are clearly more likely to have been damaged. Other case studies have shown the same
probabilistic relationship (Bonowitz, 1998). On the other hand, there are cases of twin frames
and twin buildings similarly shaken but quite differently damaged (for example, Paret and
Sasaki, 1995). The reliability and usefulness of analysis is expected to be greatest in buildings
with several hundred connections, where the data is robust and where postearthquake inspection
is most in need of direction.
6.5

Case Studies

Several case studies of specific WSMF buildings have been completed, many with
sophisticated computer modeling. Some actually included testing of damaged joints removed
from the buildings. These studies and the research prompted by them played a substantial role in
the development of FEMA-267.
Table 6-3 describes the published case studies known to date. The earliest of these were
published in SAC documents 95-04 and 95-07. Researchers are encouraged to consult the
original reports for details. The 95-04 studies have been summarized and analyzed separately
(Deierlein, 1995; Bonowitz and Youssef, 1995; Bonowitz, 1998). Extended reports for some of
them have since been published as university research reports, and many shortened versions have
been published in journals and conference proceedings. Some of the listed case study buildings
have also been the subjects of parameter studies and experimental analyses by others (for
example, Maison and Kasai, 1997; Song and Ellingwood, 1999).
In addition to case studies of actual buildings, SAC also designed a matrix of generic
“model” pre-Northridge WSMF buildings for intensive parameter studies by SAC (Krawinkler,
2000; Cornell and Luco, 1999) and others (Maison and Bonowitz, 1999).
6-16

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table 6-3

Building or
Reference

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Case Studies of WSMF Buildings Affected by the 1994 Northridge
Earthquake
Building ID
(Appendix A)

Description

Recorded or
estimated PGA
[g]

Connection Damage

SAC 95-04
Krawinkler et al.

9096

Northridge, 1993,
2 stories

0.40

no damage

Krawinkler et al.

9095

Northridge, 1993,
4 stories

0.40

13% damage rate, shear
damage

Engelhardt et al.

9075

Santa Monica, 1988,
6 stories

0.64

59% damage rate

Hart et al.

9059

Woodland Hills, 1993,
5 story hospital

0.4 to 0.6

no damage

Hart et al.

9060

Woodland Hills, 1993,
5 story hospital

0.4 to 0.6

10% damage rate, shear
and panel zone damage

Naeim et al.

9023

West Los Angeles, 1982,
11 stories

0.26 to 0.41

14% damage rate, shear
and panel zone damage

Uang et al.

9107

Canoga Park, 1975,
13 stories

0.41

10% damage rate, shear
damage

Kariotis and Eimani

9088

Encino, 1969,
16 stories

0.38

16% damage rate, shear
and panel zone damage

Paret and Sasaki

9121

Canoga Park, 1987,
17 stories

0.41

9% damage rate

Paret and Sasaki

9122

Canoga Park, 1987,
17 stories

0.41

12% damage rate

Santa Clarita City Hall
(Green)

9098

Santa Clarita, 1986,
3 stories

0.59

unknown damage rate,
shear damage

Borax Corporate
Headquarters (Hajjar
et al.)

9068

Valencia, 1993,
4 stories

0.6

75% damage rate, shear
and panel zone damage

Anderson et al., 1995

9114

Santa Clarita, 1991,
2 stories

0.6

50% damage rate, panel
zone damage

SAC 95-07

6-17

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Table 6-3

Building or
Reference

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Case Studies of WSMF Buildings Affected by the 1994 Northridge
Earthquake (continued)
Building ID
(Appendix A)

Description

Recorded or
estimated PGA
[g]

Connection Damage

NISTIR 5944
Kaufmann et al.: A

9084

Simi Valley, 1980,
6 stories

0.3

19% damage rate, panel
zone damage

Kaufmann et al.: B

9022

West Los Angeles, 1984,
4 stories

0.2

14% damage rate

Kaufmann et al.: C

9021

Sherman Oaks, 1983,
4 stories

0.4

14% damage rate, shear
and panel zone damage

Kaufmann et al.: E

9023

West Los Angeles, 1982,
11 stories

0.2

14% damage rate, shear
and panel zone damage

Kaufmann et al.: F

9020

Sherman Oaks, 1985,
4 stories

0.4

33% damage rate, shear
and panel zone damage

Naeim et al., 1999

none

Encino, 20 stories

0.41

4% damage rate, panel
zone damage

Naeim et al., 1999

none

Tarzana, 10 stories

0.47

2% damage rate

Naeim et al., 1999

none

North Hollywood,
8 stories

0.30

no damage

Naeim et al., 1999

9148

Sherman Oaks,
16 stories

0.45

2% damage rate

Islam et al.: A

9017

Santa Monica, 1987,
5 stories

0.6

50% damage rate, shear
and panel zone damage

Islam et al.: B

9028? 9044?
9045?

West Los Angeles, 1981,
11 stories

0.3

26% damage rate, shear
and panel zone damage

CSMIP

Others

Bertero et al., 1994

9050

Sherman Oaks,
6 stories

not available

60% damage rate

Observations from the SAC 95-04 case studies provided guidance for the development of the
FEMA-267 Interim Guidelines. Aggregate conclusions from those studies included:

6-18

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

•

Analytical procedures can find general locations within buildings that are more likely to be
damaged.

•

Higher mode effects seem to have been the cause of concentrated damage in upper stories.

•

Ground motions generated by the Northridge earthquake did not generate large inelastic joint
rotation demands.

Following are brief descriptions of the SAC 95-04 and 95-07 case studies. Readers are
encouraged to consult the references for details. Damage types refer to those defined in FEMA267.
6.5.1

Krawinkler et al.

Adjacent two and four-story buildings with perimeter moment frames were evaluated and
inspected. Both were designed by the same structural engineer under the provisions of the 1988
UBC. According to construction reports, 100% of the first 40 welds were ultrasonically
inspected, and 25% thereafter. Both buildings had base shear capacities on the order of 3.5 to 4
times UBC demands. The four-story building was found to have damage to 14 bottom flange
and two top flange connections out of approximately 120 total (91 inspected). Damage was
confirmed with magnetic particle and ultrasonic testing. Typical failures included pullout of
column flange material above the toe of the weld and cracks through the weld throat. The twostory building had no visible damage.
Researchers analyzed these buildings to determine whether the damage could have been
predicted. Results were mostly inconclusive. High elastic demand-to-capacity ratios and
interstory drift generally were fair predictors of damage, as was excessive inelastic deformation
of the panel zones. However, while both the two-story and four-story buildings had relatively
high inelastic deformation demands, only the taller one was damaged.
6.5.2

Engelhardt et al.

This six-story building in Santa Monica consisted of one- and two-bay frames with relatively
large members (W24 to W33 beams and W14x176 to W14x193 columns at the base). The
column sections did require doubler plates. The typical connection used welded flanges and
bolted webs.
After the earthquake, discontinuities were found in 92 of 120 welded joints. The
discontinuities were of various widths, but all stayed within the girder flange welds. About a
third were classified as W1. Some of the W2, W3, or W4 fractures could have propagated from
original W1 flaws.
6.5.3

Hart et al.

Two adjacent six-story hospital buildings were evaluated, one with observed weld fractures
and one without. The two buildings were constructed at the same time, and each had both
perimeter and interior frames. The damaged building was found to have 134 damaged

6-19

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

connections, all but 19 of which were classified as type W1 or W2. Most of the remaining 19
had column divots. One also had a panel zone crack.
The buildings were studied extensively to try to correlate the observed damage with
modeling parameters. The type W incidences appeared to be spread randomly throughout the
building since they could not be accurately predicted by analysis. But the study also concluded
that the locations of clear earthquake damage, such as type C fractures, were not well predicted.
6.5.4

Naeim et al.

This 11-story building in West Los Angeles was designed to the provisions of the 1979 UBC.
913 of 920 total connections were inspected, with damage observed in 258. The building
exhibited few outward signs of obvious nonstructural or structural damage, although a one-inch
permanent drift was measured.
Damage to the moment connections was more varied than in the buildings discussed above.
While the most common type was the W1 flaw (41% of all incidences), column flange damage
types C3 and C2 represented 25% and 18% of all damage incidences respectively. The
remaining fractures occurred about equally in the panel zones and as type C5 tearing of the
column flange. The damage patterns were well distributed throughout the building.
Researchers concluded after analysis that the overall correlation between observed and
predicted damage was tenuous. Elastic demand-capacity ratios did predict damage better than
other analysis parameters.
6.5.5

Uang et al.

This 13-story structure was located approximately three miles from the epicenter of the
Northridge earthquake. The building has perimeter frames and was probably designed to the
requirements of the 1973 UBC. Notably, the code design criteria at this time did not include
specific provisions for panel zone strength. An analysis of the building following the Northridge
earthquake indicated that substantial panel zone shear yielding should have been expected and
may have contributed a significant amount of energy dissipation.
Damage in this building indicated a directionality of the shaking, as two parallel perimeter
frames sustained significantly more damage than the two frames in the other principal direction.
The researchers concluded that while analysis could not locate specific damage, elastic demandcapacity ratios could identify subsets of connections that were significantly more likely to have
been damaged.
6.5.6

Kariotis and Eimani

This building has an aspect ratio of about 2.5:1. In the longitudinal direction, the lateral
system consists of two perimeter six-bay moment frames. In the transverse (north-south)
direction, there are two heavy three-bay moment frames with W14x370 to W14x420 base
columns and 42-inch plate girders. The building was designed to the 1969 L.A. City code,
whose base shear requirements were similar to those in the 1970 UBC. After the 1971 San

6-20

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Fernando earthquake, base shear requirements were modified. If designed to the 1976 UBC, this
building would have been designed for almost twice the base shear.
Damage was concentrated in the transverse frames. Widespread W1 and W2 incidences
could not be consistently predicted by analysis. Nineteen type C fractures in the transverse
direction were through the column flange, with some extending into the panel zone.
This building had also been analyzed after the 1971 San Fernando earthquake. See Table 5-3
(KB Valley Center) for additional information.
6.5.7

Paret and Sasaki

This 17-story building is an excellent example of trends toward less structural redundancy
and very large moment frame sections. The building has two two-bay moment frames in each
direction, with columns ranging from W14x311 to W14x730 and beams from W30x99 to
W36x300. The weld fractures all occurred in the beam bottom flanges in one direction. Over
80% of the fractures occurred above the 9th floor of the building.
6.5.8

Santa Clarita City Hall (Green)

Santa Clarita City Hall was among the first buildings to be found with Northridge earthquake
WSMF damage. The building initially appeared to have experienced only non-structural
damage, including partition and tile cracking, fallen ceiling panels, and overturned cabinets.
After ceiling tiles and fireproofing were removed, inspectors found sheared bolts at several beam
web-to-column connections. Bottom beam flange welds had also fractured. Repair of the
damaged joints took ninety days and included a seven-day per week schedule with overtime. The
total cost of the repairs was about $2,000,000.
6.5.9

Anderson, Johnston, and Partridge

This two-story structurally irregular building was near Santa Clarita City Hall, but its damage
was more obvious. The building had a noticeable permanent drift of about 2% in the first story.
All moment connections at the second floor were damaged. The majority of the damage
appeared to have started at the bottom beam flange welds with cracks propagating through the
column flange and in many cases into the panel zone.
The owner decided to demolish the building. Before demolition, the researchers were able to
remove some of the WSMF joints for testing. Based on the tests, the researchers concluded that:
•

Even in its badly damaged state, the beam-column joint had substantial ability to resist
additional load cycles. While the researchers believe that the damaged building might have
withstood moderate aftershocks, they could not conclude that it would have survived another
major earthquake.

•

The original strength and stiffness of the joints could be restored if the right repair methods
were used. It was typically not sufficient to simply reweld the cracked bottom flange weld.

6-21

 FEMA-355E
Chapter 6: Performance of WSMFs
in the 1994 Northridge Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

6.5.10 Borax Corporate Headquarters (Hajjar et al.)

This building was one of a two-building complex. No structural damage was immediately
suspected from an initial postearthquake inspection. However, one of the two buildings
experienced obvious permanent interstory drift, prompting inspection of the WSMF connections.
When connection damage was found, inspection continued to the other structure. Damage was
found in about 75% of all moment connections, but the second and third floor connections saw
100% and 93% damage respectively. Damage occurred only at girder bottom flanges. Cracks
observed in top flange welds were determined to have occurred prior to the earthquake. The
most common fracture types involved cracked welds or divots torn from the column face. In at
least 23 locations, the cracks extended into the panel zones.
The researchers cited inadequate weld fusion as the critical factor in the damage. They found
lack of fusion in three main locations: at the backing bar, at weld ends where weld dams were
used, and at the weld access hole by the beam bottom flange where the welder must stop and
restart the weld.

6-22

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

APPENDIX A. WSMF DATA FROM THE NORTHRIDGE EARTHQUAKE
Table A-1
Bldg ID

Number

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID

Street

City

Zip

Lat.

Long.

Sty

LA

Bonowitz

Durkin

Dames

21700

Oxnard St

Woodland Hills

91367

34.18

-118.60

22

156

9002

21650

Oxnard St

Woodland Hills

91367

34.18

-118.60

25

16

9003

15260

Ventura Blvd

Sherman Oaks

91403

34.15

-118.47

22

65

9004

14144

Sherman Oaks

91423

34.15

-118.44

3

198

9005

21041

Woodland Hills

91367

34.17

-118.66

3

27

9006

16650

Los Angeles

91406

34.20

-118.50

2

206

13

9007

21051

Woodland Hills

91367

34.18

-118.59

2

80

15

9008

9200

Ventura Blvd
Warner Center
Ln
Sherman Way
Warner Center
Ln
Oakdale

Chatsworth

91311

34.24

-118.56

11

4

9009

21550

Oxnard St

Woodland Hills

91367

34.18

-118.60

11

59

9010

21800

Oxnard St

Woodland Hills

91367

34.18

-118.60

11

124

95

20

9011

5950

Canoga Ave

Woodland Hills

91367

34.18

-118.60

6

152

5

21

9012

5850

Canoga Ave

Woodland Hills

91367

34.18

-118.60

6

159

48

23

9013

3301

Barham Blvd

Los Angeles

90068

34.13

-118.34

4

151

9014

15821

Encino

91436

34.16

-118.48

7

149

9015

5990

Van Nuys

91411

33.98

-118.39

6

52

9016

15315

Ventura Blvd
N Sepulveda
Blvd
Magnolia Blvd

Sherman Oaks

91403

34.16

-118.47

4

19

SOA

9017

1301

20th Street

Santa Monica

90404

34.03

-118.48

5

na

BJ19

9018

20700

Ventura Blvd

Woodland Hills

91364

34.17

-118.58

3

20

MNH02?

9019

11900

Olympic Blvd

Los Angeles

90064

34.03

-118.45

8

106

9020

15165

Ventura Blvd

Sherman Oaks

91403

34.15

-118.46

4

6

JAM7484

49

6 Kaufmann/NIST

9021

15060

Ventura Blvd

Sherman Oaks

91403

34.15

-118.46

4

5

JAM7482

50

Kaufmann/NIST

A-1

5

68

9

43

10

Case Study

9001

ESI8

78

Paret

65

11
1

BJ05

28

12
206

16
18

26
69

28

149

27
19

39
43

60

Islam/NIST

42
45

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

Long.

Sty

LA

Bonowitz

Durkin

Dames

9022

12300

Wilshire Blvd

Los Angeles

90025

34.04

-118.47

4

71

JAM7485

51

9023

11150

Olympic Blvd

Los Angeles

90025

34.04

-118.44

11

14

JAM7480

52

9024

5602

DeSoto Ave

Woodland Hills

91367

34.17

-118.59

5 162?

9025

21860

Burbank Blvd

Woodland Hills

91367

34.17

-118.60

3

10

9026

19809

Prairie

Chatsworth

91311

34.24

-118.56

2

182

9027

21600

Oxnard St

Woodland Hills

91367

34.18

-118.60

20

172

9028

11444

Olympic Blvd

Los Angeles

90026

34.05

-118.24

11

86

9029

16830

Ventura Blvd

Encino

91436

34.16

-118.50

6

31

9030

15301

Ventura Blvd

Sherman Oaks

91367

34.15

-118.46

5

210

9031

20000

Prairie

Chatsworth

91311

34.24

-118.57

2

79

9032

5900

Canoga Ave

Woodland Hills

91367

34.18

-118.60

4

153

9033

11300

34.04

-118.44

9

89

29

5200

91601

34.17

-118.37

8

90

30

9035

15350

Los Angeles
North
Hollywood
Van Nuys

90064

9034

91406

34.20

-118.47

4

216

31

9036

6800

Canoga Park

91303

34.19

-118.90

5

51

8?

32

9037

15400

Olympic Blvd
Lankershim
Blvd
Sherman Way
Owensmouth
Ave
Sherman Way

Van Nuys

91406

34.20

-118.47

5

217

8?

33

9038

16030

Ventura Blvd

Encino

91436

34.16

-118.48

6

39

88

38

9039

19634

Ventura Blvd

Tarzana

91356

34.17

-118.56

3

171

74?

40

9040

9045

Corbin Ave

Northridge

91324

34.23

-118.56

3

208

31

41

9041

12121

Wilshire Blvd

Los Angeles

90025

34.04

-118.47

14

63

Simi Valley

93065

34.30

-118.74

6

na
70

none
9043

12424

Wilshire Blvd

Los Angeles

90025

34.04

-118.47

13

9044

11355

Olympic Blvd

Los Angeles

90064

34.04

-118.44

10 67.2

A-2

Paret

Case Study
Kaufmann/NIST
Naeim/SAC;
Kaufmann/NIST

58
84
BJ06

17
25

49

24
59

NYA550

Islam/NIST?

62
22

14
19

4

MNH04

NYA577

22

153

51

44
47

JAM7486

48
36?35?

53

Islam/NIST?

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID

Number

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

Long.

Sty

LA

Bonowitz

Durkin

Case Study

11355

Olympic Blvd

Los Angeles

90064

34.04

-118.44

10 67.1

9046

11835

Olympic Blvd

Los Angeles

90064

34.03

-118.45

12

8

9047

1950

Sawtelle

Los Angeles

90025

34.04

-118.44

3

11

AC1

9048

701

N Brand

Glendale

91203

34.16

-118.26

8

na

AC2

9049

27125

Sierra Hwy

Santa Clarita

91351

34.41

-118.46

2

na

AV1

9050

--

Riverside Dr

Sherman Oaks

6 115?

BAK

9051

2796

Simi Valley

93065

34.29

-118.74

2

none

--

Sherman Oaks

91403

34.16

-118.43

2

9053

1919

Santa Monica

90404

34.03

-118.48

4

na

BJ01

9054

na

Sycamore Dr
Riverside Dr &
Woodman
Santa Monica
Blvd
na

34.14

-118.35

3

na

BJ02E

9055

4730

Woodman Ave

Sherman Oaks

91423

34.16

-118.43

4

225

BJ04

9056

5601

DeSoto Ave

Woodland Hills

91365

34.17

-118.59

3

162

BJ07

58?

9057

5601

DeSoto Ave

Woodland Hills

91365

34.17

-118.59

3

162

BJ08

58?

none

5601

DeSoto Ave

Woodland Hills

91365

34.17

-118.59

5

162

BJ09

58?

9059

5601

DeSoto Ave

Woodland Hills

91365

34.17

-118.59

5

162

BJ10

58?

Hart/SAC 95-04

9060

5601

DeSoto Ave

Woodland Hills

91365

34.17

-118.59

5

162

BJ11

58?

Hart/SAC 95-04

9061

321

N Canon

Beverly Hills

90210

34.07

-118.40

3

na

BJ14

9062

401

Santa Monica

90401

34.02

-118.50

13

na

BJ16

9063

5550

Woodland Hills

91367

34.17

-118.61

3

37

BJ18

9064

6041

Wilshire Blvd
Topanga Cyn
Blvd.
Cadillac Ave

Los Angeles

90034

34.04

-118.37

4

9065

550

Wilshire Blvd

90401

34.02

-118.50

4

na

BLC1

9066

12020

Chandler

91607

34.17

-118.40

4

?

CAB

9067

101

Continental

Santa Monica
North
Hollywood
El Segundo

90245

33.92

-118.39

15

na

DM1

Universal City

A-3

57

Paret

9045

na

35?36?

Dames

Islam/NIST?

63

Bertero

BBRS1
BC1

BJ20

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

Long.

Sty

LA

Bonowitz

9068

26877

Tourney

Valencia

92385

34.41

-118.57

4

na

EQE1

9069

26877

Tourney

Valencia

92385

34.41

-118.57

1

na

EQE2

9070

1200

90049

34.08

-118.48

5

94

ESI1

9071

--

90095

34.07

-118.45

6

?

ESI10

none

808

Getty Center Dr Los Angeles
Circle Drive
South &
Westwood
Westwood/LeCo
nte
Wilshire Blvd
Santa Monica

90401

34.02

-118.49

5

na

ESI2

none

3601

W Olive Ave

Burbank

91505

34.15

-118.34

8

na

ESI3

9074

10580

Wilshire Blvd

Los Angeles

90024

34.06

-118.43

27

?

ESI4

9075

1250

4th St

Santa Monica

90401

34.02

-118.50

6

na

ESI5

9076

13949

91423

34.15

-118.44

3

64

ESI7

9077

--

90095

6

?

ESI9

9078

11000

Ventura Blvd
Sherman Oaks
UCLA Campus
nr Westwood/
Westwood
LeConte
Wilshire Blvd
Los Angeles

90024

34.06

-118.45

17

?

FE1

9079

11755

Wilshire Blvd

Los Angeles

90025

34.05

-118.46

24

69

JAM7479

9080

16000

Ventura Blvd

Encino

91436

34.16

-118.48

12

81

JAM7487

9081

16133

Ventura Blvd

Encino

91436

34.16

-118.48

13

73

JAM7488

9082

16027

Ventura Blvd

Encino

91436

34.16

-118.48

6

33

JAM7489

none

8949

Wilshire Blvd

Beverly Hills

90211

34.07

-118.39

7

na

JAM7586

9084

3041

Simi Valley

93065

34.28

-118.74

6

na

JAM7642

9085

15451

Mission Hills

91345

34.27

-118.47

4

101

JM1

9086

15451

Mission Hills

91345

34.27

-118.47

4

102

JM2

9087

6200

Cochran
San Fernando
Mission Blvd
San Fernando
Mission Blvd
Canoga Ave

Woodland Hills

91367

34.18

-118.60

4

221

KAR2

A-4

Durkin

Dames

Paret

Case Study
Hajjar/SAC 95-07

Engelhardt/SAC
95-04
38

66?

Kaufmann/NIST

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID

Number

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

Long.

34.16

-118.48

9088

15910

Ventura Blvd

Encino

91436

9089

--

--

Santa Clarita

9090

--

--

9091

--

9092

7033

9093

18000+/-

9094

Sty

LA

Bonowitz

16

21

KAR3

91355

2

na

KPFF1A

Santa Clarita

91355

1

na

KPFF1B

Santa Clarita

91355

2

na

KPFF1C

Canoga Park

91303

34.20

-118.60

3

179

L&M

Northridge

91330

34.24

-118.53

4

?

LCIB

18111

-Owensmouth
Ave
Plummer &
Etiwanda
Nordhoff

Northridge

91330

34.24

-118.53

2

?

LCICH

9095

18111

Nordhoff

Northridge

91330

34.24

-118.53

4

?

LCIEA1

9096

18111

Nordhoff

Northridge

91330

34.24

-118.53

2

?

LCIEA2

9097

--

Plummer

Northridge

91330

3

?

LCIED

9098

23920

Valencia

Valencia

91355

3

na

MG1

34.41

-118.56

none

3

Dames

Paret

Case Study
Kariotis/SAC 9504

Krawinkler/SAC
95-04
Krawinkler/SAC
95-04
Green/SAC 95-07

MNH02

9100

12233

Olympic Blvd

Los Angeles

90064

34.03

-118.46

3 82.1 MNH03AB

9101

12233

Olympic Blvd

Los Angeles

90064

34.03

-118.46

3 82.2

9102

12233

Olympic Blvd

Los Angeles

90064

34.03

-118.46

3 82.3

MNH03F

9103

12233

Olympic Blvd

Los Angeles

90064

34.03

-118.46

3 82.4

MNH03G

9104

12233

Olympic Blvd

Los Angeles

90064

34.03

-118.46

3 82.5

MNH03H

9105

11845

Olympic Blvd

Los Angeles

90064

34.03

-118.45

13

8?

NYA501

9106

21900

Burbank Blvd

Woodland Hills

91367

34.17

-118.61

3

45

NYA539

9107

21555

Woodland Hills

91367

34.18

-118.60

13

147

NYA544

9108

10100

Oxnard St
Santa Monica
Blvd

Los Angeles

90067

34.06

-118.42

28

?

NYA591

A-5

Durkin

MNH03CDE

65?

147 Uang/SAC 95-04

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

Long.

Sty

LA

Bonowitz

90067

34.06

-118.41

20

?

NYA592

90405

34.02

-118.46

2

na

NYA629

9109

1888

9110

2701

Century Park
Los Angeles
East
Ocean Park Blvd Santa Monica

9111

501

Colorado

Santa Monica

90401

34.01

-118.49

3

na

NYA630

9112

611

N Brand

Glendale

91203

34.16

-118.26

14

na

NYA631

none

303

Burbank

34.18

-118.31

10

na

NYA653

9114

--

Santa Clarita

34.41

-118.55

2

na

RCRJ

9115

3903

Glenoaks
American &
Valencia
W Olive Ave

na

SGH1

9116

--

9117

111

9118

Burbank

91505

34.15

-118.34

6

Los Angeles

90036

34.07

-118.36

4

Burbank

91505

34.15

-118.34

4

--

Fairfax & 3rd
N Hollywood
Way
Valley & Soto

Boyle Heights

34.06

-118.20

5

WHLHSC

9119

--

Valley & Soto

Boyle Heights

34.06

-118.20

4

WHLHSE

9120

--

--

Glendale

9121

6320

Canoga Ave

Woodland Hills

91367

34.19

9122

6320

Woodland Hills

91367

9123

11111

9124

4605

9125

6345

Canoga Ave
Santa Monica
Blvd
Lankershim
Blvd
Balboa Blvd

Durkin

Dames

Paret

Case Study

Anderson/SAC
95-07

SOM1
na

WEA

20

na

WHLOF

-118.60

18

?

WJE1

Paret/SAC 95-04

34.19

-118.60

18

?

WJE2

Paret/SAC 95-04

34.05

-118.44

21

57

2

34.15

-118.37

8

193

3

34.19

-118.50

3

56

6

9126

34.18

-118.59

1

7

9127

34.19

-118.46

9128

34.17

-118.59

99?
51?2
5
17?
5

34.19

-118.50

3

9129

6345

Balboa Blvd

91316

91316

A-6

56

8
9
10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID

Number

9130

21011

9131

14130

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street
Warner Center
Ln
Riverside Dr

City
Woodland Hills

Zip

Lat.

Long.

91367

34.18

-118.59

1

110

11

34.16

-118.44

3

189

12

34.20

-118.63

6

34.18

-118.59

1

34.17

-118.59

1

34.15

-118.37

6

34.24

-118.57

6

9132
9133

20950

Warner Center
Ln

9134
9135

4640

Lankershim
Blvd

9136

Sty

LA

Bonowitz

Durkin

98

22

16
17

34.18

-118.60

1

1

18

9138

12001

34.14

-118.39

6

186

20

9139

11175

34.05

-118.45

9

41

21

9140

1964

Ventura Place
Santa Monica
Blvd
Westwood Blvd

34.05

-118.43

4

83

23

9141

34.17

-118.59

3

48?

24

9142

34.20

-118.50

3

25

9143

34.15

-118.44

3

26

9144

34.06

-118.45

17

27

Bundy Dr

Los Angeles

90025

9147
9148

15303

Ventura Blvd

9149
9152
9153

1663

Sawtelle

Sherman Oaks

91403

34.03

-118.46

9

34.16

-118.41

5

34.15

-118.47

15

34.07

126?

29

168

30
32

42

33

-118.47

1 163?

34

34.05

-118.45

3

37

34.20

-118.50

3

A-7

110

15

Oxnard St

1940

Case Study

14

21530

9146

Paret

13

9137

none

Dames

194

39

42 Naeim/CSMIP

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

9154

11150

9155

15319

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Santa Monica
Blvd
Chatsworth St

9156

Lat.

Long.

Sty

LA

Bonowitz

Durkin

34.05

-118.45

16

3

41

34.26

-118.57

3

72

42

34.17

-118.61

22

34.05

-118.44

3

203

45

34.26

-118.50

3

190

46

34.19

-118.50

3

56

47

11040

9158

10605

Santa Monica
Blvd
Balboa Blvd

9159

6345

Balboa Blvd

9160

21150

Dumetz Rd

34.16

-118.59

2

88

50

9161

14500

34.22

-118.47

6

62

51

9162

11911

Roscoe Blvd
San Vicente
Blvd

34.05

-118.47

3

145

52

34.18

-118.59

34.17

91316

-118.61

26?1
09?
3 107

54

34.07

-118.46

6

150

55

34.17

-118.60

3

9

57

34.17

-118.58

3

34.19

-118.61

5

34.17

-118.58

3

34.17

-118.61

3

104

62

34.14

-118.36

3

161

63

34.21

-118.48

3

212

64

none

34.18

-118.60

12 147?

65

none

34.05

-118.46

25

66

9163
9164

22120

9165

500

9166

21820

Clarendon St
S Sepulveda
Blvd
Burbank Blvd

Woodland Hills

91367

91367

9167
9168

6325

Topanga Cyn
Blvd

9169
9170

22020

9171

4050

9172

7855

Clarendon St
Lankershim
Blvd
Haskell Ave

Los Angeles

91406

A-8

Paret

44

9157

Granada Hills

Dames

1

53
107

58
114

59
61

69?

212

Case Study

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Lat.

Long.

34.05

-118.46

3

34.20

-118.44

3

34.15

-118.47

3

160

71

34.17

-118.59

1

95

72

9179

34.26

-118.50

3

73

none

34.17

-118.56

4 171?

74

none

34.27

-118.44

4

75

34.14

-118.40

2

174

76

34.19

-118.47

5

119

77

34.18

-118.59

1

97?

79

9175

Number

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

11925

Street

City

Zip

Wilshire Blvd

9176
9177

15490

Ventura Blvd

9178

20951

Burbank Blvd

9182

12345

9183

6355

91367

Ventura Blvd
Topanga Cyn
Blvd

9184

Sty

LA
132

Bonowitz

Durkin

70

6345

Balboa Blvd

91316

34.19

-118.50

3

56

80

9186

21800

Burbank Blvd

91367

34.17

-118.60

3

146

81

9187

18801

Ventura Blvd

91356

34.17

-118.54

3

60

82

34.17

-118.59

1

35.05

-118.43

4

23

85

34.05

-118.44

5

61

86

34.06

-118.46

17

34.16

-118.50

3

34.14

9188
9189

10780

9190

1640

Santa Monica
Blvd
S Sepulveda
Blvd

9191
9192

16861

9193

3838

Ventura Blvd
Lankershim
Blvd

9194
9195
none

11100

Santa Monica
Blvd

83

87
219

89

-118.36

24 166.2

90

34.05

-118.44

3 180?

91

34.05

-118.44

16

34.27

-118.47

4

A-9

Paret

67

9185

Tarzana

Dames

58

92
93

60

Case Study

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

9197

11766

9198

3838

9199

11500

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Wilshire Blvd
Lankershim
Blvd
Olympic Blvd

9200
9201

Lat.

Long.

34.05

-118.46

34.14

Sty
17

LA

94

-118.36

4 166.3

96

34.04

-118.44

6

125

97

34.18

-118.59

1

99?

98

34.05

-118.43

3

71?

99

34.18

-118.47

5

142

100

3

2

3

7

9202

5900

none

11550

none

15503

Ventura Blvd

none

11400

Olympic Blvd

Los Angeles

90064

16 12.1

none

11400

Olympic Blvd

Los Angeles

90064

4 12.2

none

18370

91356

7

13

none

5921

1

15

none

6301

12

17

none

6300

Burbank Blvd
Owensmouth
Ave
Owensmouth
Ave
Canoga Ave

17

18

none

10877

22

25

none

20955

1

26

none

10920

Wilshire Blvd
Warner Center
Ln
Wilshire Blvd

20

28

none

11080

Olympic Blvd

4

29

none

10960

Wilshire Blvd

24

32

none

1460

Westwood Blvd

3

34

none

10866

14

35

none

7345

Wilshire Blvd
Medical Center
Dr

6

40

Woodland Hills

Woodland Hills

91411

91367

91367

Durkin

178

N Sepulveda
Blvd
Indian Hills Rd

Van Nuys

Bonowitz

34.18

34.18

-118.60

-118.59

A-10

Dames

Paret

142

17

53?

26

Case Study

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID

Number

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

none

9055

none

1801

none

10900

Reseda Blvd
Century Park
East
Wilshire Blvd

none

5000

none

City

Zip

Lat.

Long.

Sty

LA

1

43

25

46

16

49

Van Nuys Blvd

?

50

13248

Roscoe Blvd

3

53

none

13400

Riverside Dr

3

55

none

6060

3

66

none

11999

4

68

none

21054

Sepulveda Blvd
San Vicente
Blvd
Sherman Way

3

74

none

16461

Sherman Way

3

75

none

12400

Wilshire Blvd

15

76

none

1849

7

78

none

20935

1

85

none

22144

3

87

none

10585

3

92

none

10635

3

93

none

20971

1

96

none

20970

none

20920

none

20931

none

640

Sawtelle
Warner Center
Ln
Clarendon St
Santa Monica
Blvd
Santa Monica
Blvd
Burbank Blvd
Warner Center
Ln
Warner Center
Ln
Burbank Blvd
S Sepulveda
Blvd

Los Angeles

90024

34.06

-118.44

91367

Bonowitz

Durkin

Dames

Paret

61

Woodland Hills

91367

34.18

-118.59

1

97

79?

97

Woodland Hills

91367

34.17

-118.59

1

99

7?98?

99

1

100

3

103

91367

A-11

Case Study

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

none

5805

Sepulveda Blvd

none

6815

none

21031

none

1440

none

13425

none

11633

none

13245

Noble Ave.
Warner Center
Ln
S Sepulveda
Blvd
Ventura Blvd
San Vicente
Blvd
Riverside Dr

none

12100

none

11990

none

3838

none

10800

Wilshire Blvd
San Vicente
Blvd
Lankershim
Blvd
Wilshire Blvd

none

14140

Ventura Blvd

none

10840

Wilshire Blvd

none

16530

none

9375

none

20350

none

6400

none

10990

none

7301

West Hills

91307

34.20

none

15760

Ventura Blvd
San Fernando
Rd
Ventura Blvd
Laurel Canyon
Blvd
Wilshire Blvd
Medical Center
Dr
Ventura Blvd

Sherman Oaks

91436

none

16500

Ventura Blvd

Sherman Oaks

91436

Long.

Sty

LA

8

105

Van Nuys

91405

34.19

-118.46

5

108

Woodland Hills

91367

34.18

-118.59

1

109

3

111

3

112

3

113

6

115

20

116

3

117

36

120

25

121

3

122

13

123

6

127

6

129

Sherman Oaks

Sherman Oaks
Encino

Woodland Hills
North
Hollywood

91423

91423
91436

34.16

34.15
34.16

-118.42

-118.44
-118.49

Bonowitz

Durkin

Dames

Paret

Case Study

108
53?

109

BAK?

122
46

127

91364

34.17

-118.58

2

130

130

91606

34.19

-118.40

6

131

131

18

133

-118.63

5

134

134

34.16

-118.48

20

135

135 Naeim/CSMIP?

34.16

-118.49

4

136

136

A-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID

Number

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

none

4312

Woodman Ave

none

16603

none

11726

none

10474

none

6350

none

11846

none

11677

none

12925

none

5311

none

16221

Ventura Blvd
San Vicente
Blvd
Santa Monica
Blvd
Laurel Canyon
Blvd
Ventura Blvd
San Vicente
Blvd
Riverside Dr
Topanga Cyn
Blvd.
Mulholland Dr

none

1101

Gayley Ave

none

14546

none

15500

none

11601

Hamlin St
Van Nuys
S Stephen Wise
Dr
Wilshire Blvd

none

10936

Wilshire Blvd

none

16600

Sherman Way

none

21731

Ventura Blvd

none

13412

Ventura Blvd

none

22141

none

Sherman Oaks

North
Hollywood

Zip

91436

91606

Lat.

34.16

34.19

Long.

-118.49

-118.40

Sty

LA

3

137

5

141

6

143

3

144

4

148

3

155

3

157

Bonowitz

Durkin

Dames

Paret

141

148

Sherman Oaks

91423

34.16

-118.41

4

158

158

Woodland Hills

91364

34.17

-118.61

1

164

164

1

165

3

167

3

169

2

170

24

175

22

176

2

177

3

181

3

183

Ventura Blvd

3

184

21300

Victory Blvd

12

185

none

10515

Balboa Blvd

Granada Hills

3

187

none

16542

Ventura Blvd

Encino

5

188

Van Nuys
Sherman Oaks

91411

91406
91423

91436

34.19

34.20
34.15

34.16

-118.45

-118.49
-118.42

-118.49

A-13

169

177
56

60

188

Case Study

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Table A-1
Bldg ID

Number

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

none

3575

Cahuenga Blvd

none

12626

none

City

Zip

Lat.

Long.

LA

6

191

Riverside Dr

5

192

6445

Sepulveda Blvd

3

195

none

3330

Cahuenga Blvd

5

196

none

18321

Ventura Blvd

9

197

none

11570

Indian Hills Rd

1

199

none

17404

Ventura Blvd

3

200

none

1828

Sawtelle

3

201

none

15301

3

209

none

7230

6

211

none

1990

Ventura Blvd
Medical Center
Canoga Park
Dr
Westwood Blvd

3

213

none

22110

Roscoe Blvd

3

214

none

16501

Ventura Blvd

6

215

none

20750

Ventura Blvd

4

222

none

17547

Ventura Blvd

3

223

none

22025

Ventura Blvd

3

226

none

15650

Devonshire St

3

227

none

10800

W Pico Blvd

4

229

none

11859

Wilshire Blvd

6

230

none

15531

San Fernando Mission Blvd

3

231

none

22801

3

233

none

4705

5

234

none

1100

Ventura Blvd
Laurel Canyon
Blvd
Glendon Ave

20

242

none

90068

Sty

90068

Woodland Hills

91307

91364

34.20

34.17

-118.63

-118.62

Encino

20 135?

A-14

Bonowitz

Durkin

Dames

Paret

Case Study

211

blank

Naeim/CSMIP

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table A-1
Bldg ID
none
none

Number

FEMA-355E
Appendix A: WSMF Data from the Northridge Earthquake

Master List of Northridge WSMF Databases Sorted by SAC Building ID and L.A. Building ID (continued)
Street

City

Zip

Lat.

Long.

North
Hollywood
Tarzana

A-15

Sty

LA

Bonowitz

Durkin

Dames

Paret

Case Study

8

Naeim/CSMIP

10

Naeim/CSMIP

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

APPENDIX B. NORTHRIDGE EARTHQUAKE WSMF BUILDING
DAMAGE
Bruce F. Maison
B.1

Introduction

This appendix presents the results from a database of WSMF buildings in the immediate
region affected by the 1994 Northridge earthquake. The purpose is to summarize the actual past
performance of WSMF buildings based on interpretation of actual damage inspection reports.
The building database is described and several statistics summarized.
B.2

Connection Component Damage

Typical components of pre-Northridge moment connections are depicted in Figure B-1.
Building surveys revealed different numbers of damage incidents associated with each
component, and Figure B-2 illustrates the distribution found in 29 Northridge damaged buildings
(Dames and Moore, 1998). A connection can suffer multiple incidents so that the number of
damage incidents is greater than the number of damaged connections. Damage to girder groove
welds and column flanges are the predominant damaged components.
Column

Shear Tab
Continuity Plate
Supplemental
Weld

Panel Zone
Doubler Plate

Girder
Flange Groove Weld
(top & bottom flange)

Figure B-1

Typical Components of Pre-Northridge Moment Connections

FEMA 267 (1995) further differentiates damage types into specific categories and assigns
damage indices to each: groove weld (W1 to W5), column (C1 to C7), girder (G1 to G8), panel
zone (P1 to P9), and shear tab (S1 to S6). The indices (Table 4-3a in FEMA 267) represent SAC
judgmental estimates of the relative damage severity, and the most severe have indices of 8 or
greater, where 10 represents a total loss. These typically involve fracture initiating at the flange
groove weld root with the crack propagating into the weld or column flange. The most frequent
types are fracture through the weld metal thickness (W2) and full or partial flange crack in the
column HAZ (C4).

B-1

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

From sample of 29 buildings having 3425 reported damage incidents
1800

1778

Consists of types W1 and W5 indications
which are not considered earthquake
damage

Number of Incidents

1600
1400
1200

Damage Index < 8
1000
800

Damage Index ≥ 8
716
573

600
400
164

200

7

4

27

84

57

15

0
Groove Weld

Column Flange

Girder

Panel Zone

Shear Tab

Connection Component

Figure B-2

Distribution of Component Damage

Note that fully one-half of the total (1778 out of 3425 reported incidents) are groove weld
types having damage indices < 8. These are weld root (W1) indications and non-rejectable UT
detectable indications (W5) which were classified as possible damage by FEMA 267, but are
now considered by SAC as pre-earthquake existing defects (SAC 90% Draft, 1999, Paret and
Attalla, 1998). This reclassification sharply reduces the amount of connection damage attributed
to the Northridge earthquake.
This appendix uses the latter SAC interpretation, and does not count W1 and W5 incidents as
damage. A connection is considered damaged if it has damage at the bottom flange, top flange
or both locations. No differentiation is made here between the severity and numbers of the
differing damage types affecting a connection. The SAC definition of a connection is also used,
i.e., one connection is the attachment of one girder to one column.
B.3

Sources of Building Survey Data

Prior SAC studies involved the compilation of results from postearthquake building damage
surveys, and these data were used here. This data set represents virtually all known buildings
with available damage survey results in the area affected by the Northridge earthquake. The
three main sources are as follows.

B-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

SAC Phase 2 Subtask 3.1.1
This effort was the collection of data on buildings inspected under the Los Angeles
Inspection Ordinance (Michael F. Durkin and Associates, 1999). Following the 1994 Northridge
earthquake and the finding of fractured WSMF connections, the City of Los Angeles passed
Ordinance 170406 on February 22, 1995 requiring inspection and repair of certain buildings
(those within particular geographical area affected by Northridge with exemptions given to
residential buildings). The building owners were required to submit the inspection report to the
City and these were the subject of the Subtask.
Data on 242 buildings were obtained directly from the submittals, and more detailed
information on a sample of 100 buildings was obtained via contact with building owners. In
general, the data compilation from City records did not specifically identify how may
connections had W1 type damage and repairs. Since these are now not considered as damage,
many of the buildings could not be used. The 100 building sample had segregation of damage
into non-W1 and W1 damage types by use of SAC Phase 2 Subtask 3.1.3 results (Wiss, Janney,
Elstner Associates, 1999). These 100 buildings were used here. The data was on a buildingwide basis having no information on particular damage types or the amount of damage in each
building principal direction. The coordinates were slightly altered to maintain anonymity of the
building.
SAC Phase 2 Subtask 3.1.2.
This was the detailed data collection for selected WSMF buildings (Dames and Moore,
1998). Data on 49 buildings was compiled from project work performed by six engineering firms
who agreed to participate on the Subtask. A subset of 29 buildings had very detailed damage
information including the specific FEMA 267 damage types and locations within the building
frame.
SAC Phase 1 Task 2.
This task was the initial SAC data collection, assessment and interpretation of damage caused
by the Northridge event (SAC 95-06, 1995). Data on 89 buildings were compiled from
canvassing local engineering firms and testing labs. The data compilation was fairly detailed but
in a format that necessitated some review of the original survey reports to provide information
suitable for use here.
B.4

Seismic Demands

Seismic demands at each building site were estimated by assigning actual recordings of the
Northridge earthquake from a nearby recording station (Somerville, 1999). The vast majority of
buildings had recording instruments within 3 km. The seismic demand quantities were: Modified
Mercalli Intensity, peak ground acceleration, peak ground velocity, spectral acceleration (5%
damping) at the building fundamental, 0.03 sec, 1 sec, 2 sec, and 3 sec periods. The spectral
acceleration at the building fundamental period was computed by interpolation of the spectral
values using an estimated building period. Due to earthquake directivity effects, the demands

B-3

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

varied in the NS and EW directions, and the geometric mean (square root of the product) of the
two directions was used to characterize the demand at the building site.
B.5

Building Database

The three sources of building survey data discussed above were compiled into a database.
Each source had different formats, level of detail, data deficiencies, and errors. Hence, the
compilation took considerable effort including: elimination of duplicate buildings, resolution of
conflicting data from different sources, collection of missing and updated data, determination of
longitude and latitude coordinates, removal of W1 and W5 types from damage counts, and data
entry/formatting into a single product.
Judgment was used to resolve discrepancies and augment missing data where reasonable.
Some buildings did not have sufficient or appropriate data for use here, and were simply
excluded. Note that under this effort, only a few buildings had their data checked against the
original inspection reports prepared by the responsible engineer, and hence the quality of data is
largely dependent on that provided by the original sources. It is likely that errors exist in the
database presented here but it is believed that any errors are of such a nature as not to affect
overall trends/conclusions based on database analysis.
Definitions of field data contained in the database are listed in Table B-1. Tables B-2 and B3 are large tables, and are moved to the end of this appendix. They show a portion of the
building database, sorted by different fields.
Table B-1
Column Label
ID
Lat
Long
Conn
Insp
Bot
Top
B&T
Shr
PZ
Total

Area
Sty
MMI
Pga
Pgv

B.6

Definitions Used in Northridge Database (Partial List)
Description
Building identifier number.
Latitude.
Longitude.
Number of moment connections in building.
Number of moment connections inspected.
Number of connections having damage at bottom flange only.
Number of connections having damage at top flange only.
Number of connections having damage at both top and bottom flange.
Number of connections having damage to shear connectors.
Number of connections having damage to column panel zone.
Total number of damaged connections, Tdam = Bdmg + Tdmg + B&T. Note that
some building surveys reported total number of damaged connections only, and for
these cases, Bdmg, Tdmg, B&T are listed as 0.
Building area (sf)
Number of stories in building.
Modified Mercalli Intensity per Wald et al. (1999)
Peak ground acceleration (g).
Peak ground velocity (in/sec).

Summary Statistics

The database contained 185 WSMF buildings, but 18 were screened out because of either
low connection inspection rates or site locations outside the vicinity of the sample region.

B-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Buildings having low inspection rates were excluded because the resulting connection damage
rates were considered not representative for the entire building. The inspection rate cut-off was
taken as 5%. Several isolated buildings were located some distance away from the others (e.g.,
Ventura County, downtown L.A.) and these were excluded in order to make the remaining
sample more representative of the building population in the region most affected by Northridge.
About two-thirds of the screened database were L.A. Ordinance building.
Figure B-3 shows the spatial distribution of the screened 167 buildings (spectral acceleration
contours taken from Northridge, 1996). The geographic area extends from Santa Clarita in the
north to Santa Monica in the south, and from Woodland Hills in the west to Burbank in the east.
The area is roughly 600 sq. miles. The building distribution generally corresponds to the built
environment pursuant to population density. Most buildings were located south of the area that
experienced the most intense ground shaking. Clusters of buildings are found in several specific
areas: Woodland Hills/Canoga Park (southwest of epicenter), Santa Monica/West L.A. (south of
epicenter), and Sherman Oaks/Burbank (Ventura Blvd corridor east of highway I-405). Damaged
buildings (those having at least one damaged connection) were present throughout the region
represented by the database sample (Figure B-4).
The distribution of building heights and areas are shown in Figures B-5 and B-6. The median
height and area are respectively 4 stories and 70,000 sf. These are likely to be representative of
the buildings covered by the sample region, but the degree to which they match the true
population was not studied. Buildings having six stories or less constitute about 80% of the total
area (Figure B-7).
Figure B-8 shows that most buildings experienced peak ground accelerations over a fairly
limited range from 0.3g to 0.4g. The median PGA was 0.36g. Such PGAs were not extraordinary
versus those implicit in building design codes (e.g., 1994 UBC seismic zone 4 has Z = 0.4 which
implies 0.4g PGA shaking).
Figure B-9 shows the distribution of connection inspection rates. Inspection rate is defined as
the number of inspected connections divided by the total number of moment connections in the
building, expressed as a percentage. The rates have a bimodal distribution reflecting the idea that
during the building survey process, once a percentage of connections are inspected and no
damage found, then the survey is terminated. Finding damage triggers more complete
inspections. Hence, the distribution has peaks at both lower and higher rates. Damage types W1
were considered as damage when many of the inspections were performed on these buildings.
Figure B-10 shows the distribution of connection damage rates. Damage rate is defined as the
number of damaged connections discovered divided by the number of moment connections in
the inspection sample, expressed as a percentage. A striking observation is that only about onehalf (53%) of the buildings suffered damage, and of these only about 1 in 3 (27%) had rates
greater than 20%. The median and 90th percentile rates are 1.7% and 31%, respectively. This
suggests that initial post-Northridge impressions about the amount and severity of building
damage were overstated in large part due to consideration of types W1 as damage. The damage
picture changes significantly when these are excluded. Paret and Attalla (1998) have previously
noted this as well.
B-5

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Figure B-3

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Spatial Distribution of Screened Buildings

B-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Figure B-4

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Spatial Distribution of Damaged Buildings

B-7

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

90

Number of Buildings

80

78

70
Number of Buildings Within Story Range
(N = 167, Median = 4, 90th = 15)

60
52

50
40
30
20
6

7

8

8

8

7 to 9

10 to 12

13 to 15

16 to 18

> 18

10
0
<4

4 to 6

Number of Stories

Figure B-5

Distribution of Building Heights

120
103

Number of Buildings

100
Number of Buildings Within Area Range (N =
161, Median = 70,000 sf, 90th = 330,000 sf)

80

60

40

19

20

22
10
4

2

1

500 to 600

> 600

0
< 100

100 to 200

200 to 300

300 to 400

400 to 500

Building Area (sf x 1000)

Figure B-6

Distribution of Building Areas

B-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

7.00
5.98

Sum of building areas within
story height range (N = 161)

5.00
4.18

4.00

3.00

2.00

2.48

2.39

13 to 15

16 to 18

2.56

1.71
0.99

1.00

0.00
<4

4 to 6

7 to 9

10 to 12

> 18

Number of Stories

Figure B-7

Distribution of Total Areas

100
87

90

Number of Buildings

Building Areas (MSF)

6.00

80
Number of Buildings Within PGA Range
(N = 167, Median = 0.36g, 90th = 0.57g)

70
60
50
40
33

30
22

20
12

10

8
3

2

0.6 to 0.7

> 0.7

0
< 0.2

0.2 to 0.3

0.3 to 0.4

0.4 to 0.5

0.5 to 0.6

Peak Ground Acceleration (g)

Figure B-8

Distribution of Peak Ground Accelerations

B-9

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

70
60

Number of Buildings

60

Number of Buildings Within Inspection Range
(N = 167, Median = 42%, 90th = 100%)

53

50

40
31

30

20
13
10

10

0
5 to 20

20 to 40

40 to 60

60 to 80

80 to 100

Percentage of Connections Inspected

Figure B-9

Distribution of Connection Inspection Rates

90

Numbers of Buildings

80

78

70
Number of Buildings Within Damage Range
(N = 167, Median = 1.7%, 90th = 31%)

60
50
38

40
30

27

20
9

10

6

5

3

1

0
No
Damage

> 0 & < 10

10 to 20

20 to 30

30 to 40

40 to 50

50 to 60

> 60

Percentage of Connections Damaged

Figure B-10 Distribution of Connection Damage Rates

B-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Percentage Damaged Connections

Figure B-11 shows a plot of damage rates versus PGA. There is large scatter but a trend of
higher damage rates with increasing PGA is apparent. By inspection, at about 0.6g the median
rate is about 50% whereas below about 0.45g there are so many buildings with no damage
(cluster of points about 0% damage rate) that it causes the median rate to be very small in this
range. In any event, the correlation between damage rate and PGA is weak. Further statistical
analysis of the database was performed to create a methodology for rapid loss estimation. The
results of this work can be found elsewhere (Maison and Bonowitz, 1999).

80
Building Connection
Damage Rate (N = 167)

70
60
50
40
30
20
10
0
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Peak Ground Acceleration (g)

Figure B-11 Damage Rates Versus PGA
B.7

References

Dames and Moore, 1998, Survey of Damaged Steel Moment Frame Buildings, report for SAC
Phase 2, Task 3.1.2, version 1.00.
FEMA 267, 1995, Interim Guidelines: Evaluation, Repair, Modification and Design of Welded
Steel moment Frame Structures, Federal Emergency Management Agency, Washington, D.C.
Goel, R.K., and Chopra, A.K., 1997, Vibration Properties of Buildings Determined From
Recorded Earthquake Motions, Univ. of Calif. Earthquake Eng. Research Center Report No.
UCB/EERC-97/14.
Maison, B.F., and Bonowitz, D., 1999, Rapid Loss Estimation Methodology For Steel Moment
Frame Buildings, SAC Task 3.2.1, Report to SAC.
Michael F. Durkin and Associates, 1999, Collection of Data on Buildings Inspected Under the
Los Angeles Inspection Ordinance, database for SAC Task 3.1.1.
Northridge Earthquake Reconnaissance Report, 1996, Earthquake Spectra, Supplement to Vol.
II.

B-11

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Paret, T.F., and Attalla, M.R., 1998, “Changing Perceptions of the Extent of Damage to Welded
Steel Moment Frames in the Northridge Earthquake,” SEAOC 1998 Convention.
SAC 95-06, 1995, Survey and Assessment of Damage to Buildings Affected by the Northridge
Earthquake of January 17, 1994, SAC Joint Venture Technical Report SAC 95-06.
SAC 90%, 1999, Seismic Evaluation & Upgrade Criteria for Existing Welded Steel MomentResisting Frame Structures, Report No. SAC-2000-02, working draft at 90% level.
Somerville, P., 1999, Ground Motion Estimates for SAC Loss Estimation Task 3.2, report to
SAC.
Wald, D.J., Quitoriano, V., Heaton, and T.H., Kanamori, H., 1999, “Relationships between Peak
Ground Acceleration, Peak Ground Velocity, and Modified Mercalli Intensity in California”,
Earthquake Spectra, Vol. 15, No. 3.
Wiss, Janney, Elstner Associates, Inc., 1999, Evaluation of Inspection Reliability, Clarification
of the Origins of W1a and W1b and Distribution of W1 and non-W1 Conditions, report for
SAC Task 3.1.3.

B-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-2
ID
9090
9149
9126
9184
9188
9137
9134
9200
9130
9133
9163
9178
9069
9182
9114
9094
9089
9051
9096
9031
9049
9160
9006
9091
9007
9026
9110
9152
9167
9201
9131
9092
9103

Lat
34.412
34.066
34.175
34.175
34.174
34.179
34.174
34.175
34.175
34.175
34.175
34.174
34.410
34.143
34.412
34.236
34.412
34.286
34.236
34.239
34.410
34.158
34.201
34.412
34.175
34.239
34.019
34.046
34.167
34.047
34.157
34.198
34.032

Long
-118.554
-118.469
-118.589
-118.589
-118.591
-118.599
-118.591
-118.589
-118.592
-118.589
-118.590
-118.587
-118.574
-118.402
-118.554
-118.528
-118.554
-118.744
-118.528
-118.568
-118.459
-118.592
-118.495
-118.554
-118.590
-118.564
-118.457
-118.448
-118.584
-118.434
-118.441
-118.602
-118.456

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted By Height (As of 11/99)
Sty
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3

Area
7500
15000
16300
17700
19130
20000
20710
21500
22400
23400
23500
24670
27000
11000
15400
17930
19400
22600
31800
34600
36000
48000
49000
57600
62800
98444
112000
nr
nr
nr
nr
11100
13500

Conn
30
12
8
8
8
12
8
8
8
12
8
10
20
10
16
72
40
16
68
50
68
5
64
480
100
84
172
28
76
96
480
24
72

Insp
18
8
8
8
6
12
8
8
8
9
6
10
20
5
16
27
14
6
68
50
59
5
64
77
97
82
15
13
8
23
33
12
32

Bot
1
0
0
0
0
0
0
0
0
0
0
0
10
0
8
1
4
0
0
0
1
0
4
0
7
22
0
0
0
0
0
0
1

Top
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

B-13

B&T
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
3
0
0
0
0
0
0
0

Total
4
0
0
0
0
0
0
0
0
0
0
0
10
0
8
1
4
0
0
0
1
0
5
0
7
25
0
0
0
0
0
0
1

DR
0.22
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.50
0.00
0.50
0.04
0.29
0.00
0.00
0.00
0.02
0.00
0.08
0.00
0.07
0.30
0.00
0.00
0.00
0.00
0.00
0.00
0.03

Shr
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0

PZ
0
0
0
0
0
0
0
0
0
0
0
0
9
0
4
0
0
0
0
0
0
0
0
0
5
5
0
0
0
0
0
0
0

MMI
9.0
7.6
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
9.1
7.8
9.0
8.9
9.0
7.7
8.9
8.9
8.6
7.9
8.2
9.0
8.3
8.9
8.0
8.2
8.1
8.0
8.1
8.5
8.1

Pga
0.586
0.202
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.586
0.244
0.586
0.403
0.586
0.328
0.403
0.403
0.425
0.362
0.384
0.586
0.362
0.403
0.395
0.357
0.362
0.444
0.244
0.353
0.444

Pgv
33.1
10.4
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
33.1
14.5
33.1
18.3
33.1
12.9
18.3
18.3
16.7
19.1
15.2
33.1
19.1
18.3
11.7
10.8
19.1
19.4
14.5
16.7
19.4

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-2
ID
9143
9102
9040
9194
9169
9104
9157
9155
9111
9039
9175
9170
9164
9100
9192
9153
9179
9177
9076
9097
9056
9057
9129
9159
9125
9185
9022
9101
9004
9187
9047
9005

Lat
34.149
34.032
34.234
34.048
34.169
34.032
34.048
34.265
34.015
34.173
34.047
34.170
34.170
34.032
34.159
34.201
34.261
34.155
34.149
34.243
34.174
34.174
34.187
34.187
34.187
34.187
34.042
34.032
34.150
34.171
34.042
34.171

Long
-118.437
-118.456
-118.562
-118.442
-118.576
-118.456
-118.443
-118.573
-118.491
-118.561
-118.465
-118.606
-118.606
-118.456
-118.501
-118.495
-118.502
-118.472
-118.437
-118.532
-118.587
-118.587
-118.502
-118.502
-118.502
-118.502
-118.469
-118.456
-118.441
-118.543
-118.444
-118.657

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted By Height (As of 11/99) (continued)
Sty
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3

Area
16000
16800
17700
20000
20400
21000
22000
22000
23100
25000
25000
25600
30000
33600
35000
40000
42000
43000
45000
45900
48000
48000
50000
50000
50000
50000
50240
51000
52000
53000
54000
59700

Conn
34
102
46
36
54
80
24
114
48
22
46
62
50
152
54
64
90
96
36
172
84
84
57
57
69
75
100
312
114
57
146
36

Insp
20
44
9
5
10
32
8
17
6
7
10
16
13
76
16
64
19
20
26
168
16
79
15
13
20
14
100
154
114
9
38
36

Bot
0
4
0
0
0
0
0
0
0
0
0
0
0
11
0
0
0
0
5
26
0
1
0
0
0
0
13
12
17
0
16
7

Top
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

B-14

B&T
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
3
0
0
0
0
0
0
1
0
0
0
1
0

Total
5
4
0
0
0
0
0
0
0
0
0
0
0
13
0
4
6
1
5
29
0
1
1
2
2
0
14
12
17
0
17
7

DR
0.25
0.09
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.17
0.00
0.06
0.32
0.05
0.19
0.17
0.00
0.01
0.07
0.15
0.10
0.00
0.14
0.08
0.15
0.00
0.45
0.19

Shr
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0

PZ
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
9
0
0
0

MMI
8.2
8.1
8.9
8.1
8.2
8.1
8.1
9.0
7.9
8.2
8.2
8.1
8.1
8.1
7.4
8.2
9.4
8.0
8.2
9.1
8.3
8.3
8.1
8.1
8.1
8.1
8.2
8.1
8.2
8.1
8.2
8.3

Pga
0.244
0.444
0.384
0.357
0.362
0.444
0.357
0.384
0.571
0.417
0.243
0.362
0.362
0.444
0.328
0.384
0.841
0.381
0.244
0.403
0.362
0.362
0.384
0.384
0.384
0.384
0.357
0.444
0.244
0.418
0.444
0.362

Pgv
14.5
19.4
15.2
10.8
19.1
19.4
10.8
15.2
12.7
12.9
9.7
19.1
19.1
19.4
12.9
15.2
27.6
16.5
14.5
18.3
19.1
19.1
15.2
15.2
15.2
15.2
10.8
19.4
14.5
13.0
19.4
19.1

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-2
ID
9186
9063
9158
9172
9142
9141
9098
9018
9106
9162
9061
9166
9025
9054
9171
9198
9013
9095
9119
9055
9032
9065
9053
9140
9021
9020
9086
9085
9066
9117
9116
9064
9036

Lat
34.173
34.171
34.262
34.213
34.201
34.174
34.412
34.167
34.172
34.053
34.069
34.172
34.172
34.139
34.143
34.142
34.131
34.236
34.064
34.156
34.177
34.020
34.029
34.047
34.153
34.153
34.272
34.272
34.168
34.152
34.072
34.038
34.194

Long
-118.603
-118.606
-118.503
-118.475
-118.495
-118.587
-118.557
-118.584
-118.605
-118.470
-118.400
-118.604
-118.603
-118.353
-118.361
-118.361
-118.344
-118.528
-118.197
-118.431
-118.597
-118.497
-118.480
-118.435
-118.462
-118.462
-118.469
-118.469
-118.395
-118.340
-118.362
-118.375
-118.900

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted By Height (As of 11/99) (continued)
Sty
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

Area
60000
63000
67000
68000
70000
75000
81600
82000
84000
100000
108000
120000
120000
261000
865000
nr
26200
27880
36800
42400
46375
52000
54200
58000
61000
63640
64000
64000
68000
72000
73600
80000
80500

Conn
80
68
28
162
64
114
nr
108
132
28
202
140
216
135
220
240
136
96
240
74
112
78
132
298
88
40
76
76
170
48
96
228
86

Insp
14
68
9
16
11
22
48
65
33
4
24
124
106
120
39
21
133
96
30
73
112
49
110
43
88
40
72
70
25
48
17
21
41

Bot
0
9
0
0
0
0
0
11
0
0
5
0
30
33
0
0
0
11
0
5
0
0
9
0
9
10
4
12
1
10
0
3
0

Top
0
0
0
0
0
0
0
0
0
0
1
0
0
2
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0

B-15

B&T
0
0
0
0
0
0
0
0
0
0
0
0
0
10
0
0
0
1
0
0
0
0
0
0
2
3
0
0
1
1
0
0
0

Total
0
9
0
0
0
0
0
11
0
0
6
3
30
45
0
0
0
12
0
5
0
0
10
1
12
13
4
12
2
11
0
3
0

DR
0.00
0.13
0.00
0.00
0.00
0.00
0.00
0.17
0.00
0.00
0.25
0.02
0.28
0.38
0.00
0.00
0.00
0.13
0.00
0.07
0.00
0.00
0.09
0.02
0.14
0.33
0.06
0.17
0.08
0.23
0.00
0.14
0.00

Shr
0
0
0
0
0
0
0
0
0
0
0
0
0
6
0
0
0
3
0
0
0
0
5
0
1
6
0
1
0
0
0
0
0

PZ
0
0
0
0
0
0
0
6
0
0
0
0
6
8
0
0
0
0
0
0
0
0
0
0
2
1
2
4
0
8
0
0
0

MMI
8.2
8.2
9.4
8.1
8.2
8.3
9.0
8.1
8.2
8.1
7.5
8.2
8.2
7.2
7.2
7.2
7.2
8.9
7.1
8.1
8.3
7.9
8.1
8.0
8.1
8.1
9.9
9.9
7.7
7.4
7.6
7.6
6.5

Pga
0.362
0.362
0.841
0.439
0.384
0.362
0.586
0.362
0.362
0.243
0.238
0.362
0.362
0.187
0.187
0.187
0.187
0.403
0.336
0.244
0.362
0.571
0.642
0.444
0.414
0.414
0.686
0.686
0.244
0.244
0.185
0.428
0.237

Pgv
19.1
19.1
27.6
13.7
15.2
19.1
33.1
19.1
19.1
9.7
9.1
19.1
19.1
8.0
8.0
8.0
8.0
18.3
6.9
14.5
19.1
12.7
21.1
19.4
14.5
14.5
47.5
47.5
14.5
14.5
7.4
14.1
8.0

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-2
ID
9068
9189
9016
9087
9093
9035
9147
9128
9118
9070
9202
9168
9183
9127
9017
9030
9060
9190
9037
9059
9024
9165
9132
9138
9135
9161
9075
9012
9011
9050
9071
9082
9038

Lat
34.410
34.051
34.165
34.183
34.243
34.201
34.158
34.174
34.064
34.077
34.177
34.185
34.186
34.195
34.013
34.154
34.174
34.049
34.202
34.174
34.168
34.066
34.202
34.144
34.155
34.221
34.019
34.176
34.178
34.158
34.067
34.156
34.156

Long
-118.574
-118.434
-118.466
-118.597
-118.532
-118.466
-118.408
-118.592
-118.197
-118.475
-118.466
-118.606
-118.466
-118.462
-118.493
-118.465
-118.587
-118.445
-118.468
-118.587
-118.605
-118.460
-118.630
-118.393
-118.369
-118.466
-118.498
-118.597
-118.597
-118.422
-118.445
-118.482
-118.482

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted By Height (As of 11/99) (continued)
Sty
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
6
6
6
6
6
6
6
6
6
6
6
6

Area
86000
93502
94000
106800
124200
150000
50000
58500
68000
70800
77000
79000
79000
92000
102400
126020
130000
134600
151000
250000
376000
45000
50000
61200
70000
75000
90000
93514
93514
120000
120000
126000
150386

Conn
112
120
304
136
240
208
70
100
480
100
210
210
220
144
96
1014
288
462
268
372
960
156
108
168
180
190
100
192
192
228
415
378
204

Insp
112
18
154
102
240
37
8
97
32
100
16
57
24
21
96
96
284
73
58
35
621
23
20
56
42
35
100
192
192
72
33
8
32

Bot
84
0
49
32
59
0
0
0
0
13
0
0
0
0
17
0
23
0
0
0
72
0
0
0
0
0
51
59
55
43
0
0
0

Top
0
0
0
0
3
0
0
0
0
0
0
0
0
0
16
0
3
0
0
0
2
0
0
0
0
0
1
0
0
0
0
0
0

B-16

B&T
0
0
0
0
16
0
0
0
0
2
0
0
0
0
15
0
2
0
0
0
7
0
0
0
0
0
7
1
7
0
0
0
0

Total
84
0
49
32
78
0
0
7
0
15
0
1
1
0
48
14
28
2
0
0
81
0
0
0
4
2
59
60
62
43
0
0
0

DR
0.75
0.00
0.32
0.31
0.33
0.00
0.00
0.07
0.00
0.15
0.00
0.02
0.04
0.00
0.50
0.15
0.10
0.03
0.00
0.00
0.13
0.00
0.00
0.00
0.10
0.06
0.59
0.31
0.32
0.60
0.00
0.00
0.00

Shr
41
0
0
0
0
0
0
0
0
4
0
0
0
0
16
0
1
0
0
0
3
0
0
0
0
0
0
6
1
0
0
0
0

PZ
21
0
0
18
4
0
0
0
0
0
0
0
0
0
10
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

MMI
9.1
7.9
7.9
8.4
9.0
7.9
7.8
8.3
7.1
7.3
7.9
8.4
7.8
7.8
7.8
8.1
8.3
8.2
7.9
8.3
8.1
7.5
8.2
7.6
7.5
8.1
7.9
8.3
8.3
8.0
7.4
7.7
7.7

Pga
0.586
0.357
0.333
0.362
0.403
0.32
0.244
0.362
0.336
0.171
0.32
0.362
0.32
0.32
0.572
0.333
0.362
0.357
0.32
0.362
0.362
0.171
0.416
0.244
0.244
0.439
0.571
0.362
0.362
0.244
0.363
0.381
0.381

Pgv
33.1
10.8
17.3
19.1
18.3
12.7
14.5
19.1
6.9
8.8
12.7
19.1
12.7
12.7
12.7
17.3
19.1
10.8
12.7
19.1
19.1
8.8
19.7
14.5
14.5
13.7
12.7
19.1
19.1
14.5
8.6
16.5
16.5

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-2
ID
9015
9029
9084
9077
9136
9199
9115
9014
9124
9048
9019
9034
9139
9146
9033
9045
9044
9008
9009
9010
9028
9023
9080
9062
9043
9081
9046
9107
9105
9112
9041
9148
9067

Lat
33.976
34.159
34.279
34.067
34.240
34.036
34.151
34.156
34.154
34.157
34.033
34.165
34.047
34.034
34.038
34.040
34.040
34.239
34.179
34.179
34.053
34.040
34.156
34.020
34.041
34.157
34.033
34.179
34.033
34.156
34.044
34.154
34.916

Long
-118.393
-118.499
-118.737
-118.445
-118.570
-118.444
-118.341
-118.477
-118.368
-118.255
-118.450
-118.374
-118.446
-118.456
-118.440
-118.438
-118.438
-118.564
-118.605
-118.605
-118.242
-118.438
-118.482
-118.498
-118.470
-118.485
-118.450
-118.600
-118.451
-118.255
-118.467
-118.444
-118.391

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted By Height (As of 11/99) (continued)
Sty
6
6
6
6
6
6
6
6
8
8
8
8
9
9
9
10
10
10
11
11
11
11
12
13
13
13
13
13
13
14
14
15
15

Area
172000
223000
239400
240000
250000
250000
267600
567000
116000
152000
230000
232000
70000
164311
180000
183600
222400
275000
216640
216640
230560
409300
186000
221000
244000
247000
330000
332800
486720
262440
344400
270000
315000

Conn
180
252
568
360
50
700
312
964
92
224
188
208
440
320
278
440
503
688
342
342
528
920
392
1042
378
572
823
520
764
1344
420
576
720

Insp
137
207
560
36
50
87
312
189
17
8
163
66
44
73
40
369
232
626
342
342
521
917
94
299
113
63
750
507
107
19
60
133
13

Bot
15
0
104
0
0
0
33
14
0
0
17
0
0
0
0
0
0
156
25
45
53
123
1
10
0
0
0
46
34
0
0
0
0

Top
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
5
6
4
0
5
0
0
0
0
0
4
2
0
0
0
0

B-17

B&T
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
2
2
1
2
3
0
0
0
0
0
1
12
0
0
0
0

Total
16
0
104
0
0
2
33
15
0
0
17
4
0
1
0
32
38
163
33
50
55
131
1
10
0
0
130
51
48
0
1
3
0

DR
0.12
0.00
0.19
0.00
0.00
0.02
0.11
0.08
0.00
0.00
0.10
0.06
0.00
0.01
0.00
0.09
0.16
0.26
0.10
0.15
0.11
0.14
0.01
0.03
0.00
0.00
0.17
0.10
0.45
0.00
0.02
0.02
0.00

Shr
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
9
3
0
3
0
0
0
0
0
6
1
0
0
0
0

PZ
0
0
36
0
0
0
0
0
0
0
0
0
0
0
0
0
0
23
6
4
7
19
0
0
0
0
0
0
2
0
0
0
0

MMI
7.4
7.4
7.7
7.4
8.9
8.1
7.4
7.8
7.4
7.0
8.1
7.5
8.2
8.2
8.1
8.0
8.0
8.9
8.3
8.3
6.7
8.0
7.8
7.9
8.2
7.7
8.1
8.3
8.1
7.1
8.2
8.2
6.9

Pga
0.366
0.328
0.328
0.363
0.384
0.444
0.244
0.381
0.244
0.22
0.444
0.244
0.357
0.444
0.444
0.444
0.444
0.384
0.362
0.362
0.152
0.444
0.381
0.571
0.357
0.381
0.444
0.362
0.444
0.22
0.357
0.244
0.152

Pgv
8.2
12.9
12.9
8.6
15.2
19.4
14.5
16.5
14.5
6.9
19.4
14.5
10.8
19.4
19.4
19.4
19.4
15.2
19.1
19.1
6.5
19.4
16.5
12.7
10.8
16.5
19.4
19.1
19.4
6.9
10.8
14.5
3.2

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-2
ID
9154
9088
9195
9144
9191
9078
9197
9122
9121
9027
9120
9109
9123
9003
9156
9001
9193
9079
9002
9074
9108

Lat
34.047
34.156
34.048
34.058
34.058
34.058
34.049
34.185
34.185
34.179
34.157
34.061
34.048
34.154
34.169
34.179
34.143
34.051
34.179
34.062
34.062

Long
-118.445
-118.480
-118.444
-118.445
-118.460
-118.446
-118.462
-118.597
-118.597
-118.605
-118.255
-118.414
-118.445
-118.464
-118.609
-118.605
-118.361
-118.460
-118.605
-118.433
-118.417

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted By Height (As of 11/99) (continued)
Sty
16
16
16
17
17
17
17
18
18
20
20
20
21
21
22
22
24
24
26
27
28

Area
225500
233600
290600
252000
295600
391000
400000
345600
345600
385387
470000
510300
300966
369715
17500
450000
nr
448800
586000
364500
672000

Conn
640
558
962
480
432
704
416
274
278
880
488
1040
672
612
34
624
1104
1004
900
766
1232

Insp
32
121
50
28
42
12
61
114
251
867
8
10
34
388
10
603
67
106
874
22
18

Bot
0
19
0
0
0
0
0
14
23
5
0
0
0
43
0
21
0
0
41
0
0

Top
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
61
0
0

B-18

B&T
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
12
9
0

Total
0
19
0
0
0
0
0
14
23
5
0
0
0
48
0
21
0
0
114
9
0

DR
0.00
0.16
0.00
0.00
0.00
0.00
0.00
0.12
0.09
0.01
0.00
0.00
0.00
0.12
0.00
0.03
0.00
0.00
0.13
0.41
0.00

Shr
0
3
0
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0
0
0

PZ
0
7
0
0
0
0
0
0
0
0
0
0
0
2
0
2
0
0
0
0
0

MMI
8.2
7.8
8.2
7.8
7.9
7.8
8.2
8.4
8.4
8.3
7.0
7.5
8.2
8.1
8.1
8.3
7.2
8.2
8.3
7.6
7.5

Pga
0.357
0.381
0.357
0.243
0.171
0.243
0.243
0.362
0.362
0.362
0.22
0.238
0.357
0.333
0.362
0.362
0.187
0.243
0.362
0.363
0.238

Pgv
10.8
16.5
10.8
9.7
8.8
9.7
9.7
19.1
19.1
19.1
6.9
9.1
10.8
17.3
19.1
19.1
8.0
9.7
19.1
8.6
9.1

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table B-3
ID
9028
9067
9070
9165
9191
9116
9054
9171
9198
9013
9193
9149
9048
9120
9112
9036
9061
9109
9108
9144
9078
9162
9175
9197
9079
9117
9115
9124
9135
9034
9138
9066
9182

Lat
34.053
34.916
34.077
34.066
34.058
34.072
34.139
34.143
34.142
34.131
34.143
34.066
34.157
34.157
34.156
34.194
34.069
34.061
34.062
34.058
34.058
34.053
34.047
34.049
34.051
34.152
34.151
34.154
34.155
34.165
34.144
34.168
34.143

Long
-118.242
-118.391
-118.475
-118.460
-118.460
-118.362
-118.353
-118.361
-118.361
-118.344
-118.361
-118.469
-118.255
-118.255
-118.255
-118.900
-118.400
-118.414
-118.417
-118.445
-118.446
-118.470
-118.465
-118.462
-118.460
-118.340
-118.341
-118.368
-118.369
-118.374
-118.393
-118.395
-118.402

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted by Peak Ground Acceleration (As of 11/99)
Sty
11
15
5
6
17
4
3
3
4
4
24
1
8
20
14
4
3
20
28
17
17
3
3
17
24
4
6
8
6
8
6
4
2

Area
230560
315000
70800
45000
295600
73600
261000
865000
nr
26200
nr
15000
152000
470000
262440
80500
108000
510300
672000
252000
391000
100000
25000
400000
448800
72000
267600
116000
70000
232000
61200
68000
11000

Conn
528
720
100
156
432
96
135
220
240
136
1104
12
224
488
1344
86
202
1040
1232
480
704
28
46
416
1004
48
312
92
180
208
168
170
10

Insp
521
13
100
23
42
17
120
39
21
133
67
8
8
8
19
41
24
10
18
28
12
4
10
61
106
48
312
17
42
66
56
25
5

Bot
53
0
13
0
0
0
33
0
0
0
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
10
33
0
0
0
0
1
0

B-19

Top
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

B&T
2
0
2
0
0
0
10
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0

Total
55
0
15
0
0
0
45
0
0
0
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
11
33
0
4
4
0
2
0

DR
0.11
0.00
0.15
0.00
0.00
0.00
0.38
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.25
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.23
0.11
0.00
0.10
0.06
0.00
0.08
0.00

Shr
0
0
4
0
0
0
6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

PZ
7
0
0
0
0
0
8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8
0
0
0
0
0
0
0

MMI
6.7
6.9
7.3
7.5
7.9
7.6
7.2
7.2
7.2
7.2
7.2
7.6
7.0
7.0
7.1
6.5
7.5
7.5
7.5
7.8
7.8
8.1
8.2
8.2
8.2
7.4
7.4
7.4
7.5
7.5
7.6
7.7
7.8

Pga
0.152
0.152
0.171
0.171
0.171
0.185
0.187
0.187
0.187
0.187
0.187
0.202
0.220
0.220
0.220
0.237
0.238
0.238
0.238
0.243
0.243
0.243
0.243
0.243
0.243
0.244
0.244
0.244
0.244
0.244
0.244
0.244
0.244

Pgv
6.5
3.2
8.8
8.8
8.8
7.4
8.0
8.0
8.0
8.0
8.0
10.4
6.9
6.9
6.9
8.0
9.1
9.1
9.1
9.7
9.7
9.7
9.7
9.7
9.7
14.5
14.5
14.5
14.5
14.5
14.5
14.5
14.5

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-3
ID
9147
9050
9131
9055
9143
9076
9004
9148
9183
9127
9035
9202
9037
9192
9029
9051
9084
9016
9030
9003
9119
9118
9092
9189
9194
9157
9152
9022
9190
9139
9043
9041
9154

Lat
34.158
34.158
34.157
34.156
34.149
34.149
34.150
34.154
34.186
34.195
34.201
34.177
34.202
34.159
34.159
34.286
34.279
34.165
34.154
34.154
34.064
34.064
34.198
34.051
34.048
34.048
34.046
34.042
34.049
34.047
34.041
34.044
34.047

Long
-118.408
-118.422
-118.441
-118.431
-118.437
-118.437
-118.441
-118.444
-118.466
-118.462
-118.466
-118.466
-118.468
-118.501
-118.499
-118.744
-118.737
-118.466
-118.465
-118.464
-118.197
-118.197
-118.602
-118.434
-118.442
-118.443
-118.448
-118.469
-118.445
-118.446
-118.470
-118.467
-118.445

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted by Peak Ground Acceleration (As of 11/99) (continued)
Sty
5
6
3
4
3
3
3
15
5
5
4
5
5
3
6
2
6
4
5
21
4
5
3
4
3
3
3
3
5
9
13
14
16

Area
50000
120000
nr
42400
16000
45000
52000
270000
79000
92000
150000
77000
151000
35000
223000
22600
239400
94000
126020
369715
36800
68000
11100
93502
20000
22000
nr
50240
134600
70000
244000
344400
225500

Conn
70
228
480
74
34
36
114
576
220
144
208
210
268
54
252
16
568
304
1014
612
240
480
24
120
36
24
28
100
462
440
378
420
640

Insp
8
72
33
73
20
26
114
133
24
21
37
16
58
16
207
6
560
154
96
388
30
32
12
18
5
8
13
100
73
44
113
60
32

Bot
0
43
0
5
0
5
17
0
0
0
0
0
0
0
0
0
104
49
0
43
0
0
0
0
0
0
0
13
0
0
0
0
0

B-20

Top
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0

B&T
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
1
0
0
0
0
0

Total
0
43
0
5
5
5
17
3
1
0
0
0
0
0
0
0
104
49
14
48
0
0
0
0
0
0
0
14
2
0
0
1
0

DR
0.00
0.60
0.00
0.07
0.25
0.19
0.15
0.02
0.04
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.19
0.32
0.15
0.12
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.14
0.03
0.00
0.00
0.02
0.00

Shr
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

PZ
0
0
0
0
0
0
9
0
0
0
0
0
0
0
0
0
36
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0

MMI
7.8
8.0
8.1
8.1
8.2
8.2
8.2
8.2
7.8
7.8
7.9
7.9
7.9
7.4
7.4
7.7
7.7
7.9
8.1
8.1
7.1
7.1
8.5
7.9
8.1
8.1
8.2
8.2
8.2
8.2
8.2
8.2
8.2

Pga
0.244
0.244
0.244
0.244
0.244
0.244
0.244
0.244
0.320
0.320
0.320
0.320
0.320
0.328
0.328
0.328
0.328
0.333
0.333
0.333
0.336
0.336
0.353
0.357
0.357
0.357
0.357
0.357
0.357
0.357
0.357
0.357
0.357

Pgv
14.5
14.5
14.5
14.5
14.5
14.5
14.5
14.5
12.7
12.7
12.7
12.7
12.7
12.9
12.9
12.9
12.9
17.3
17.3
17.3
6.9
6.9
16.7
10.8
10.8
10.8
10.8
10.8
10.8
10.8
10.8
10.8
10.8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table B-3
ID
9195
9123
9160
9167
9170
9164
9018
9024
9156
9169
9186
9063
9106
9166
9025
9126
9184
9188
9137
9134
9200
9130
9133
9163
9178
9007
9056
9057
9005
9141
9032
9128
9060

Lat
34.048
34.048
34.158
34.167
34.170
34.170
34.167
34.168
34.169
34.169
34.173
34.171
34.172
34.172
34.172
34.175
34.175
34.174
34.179
34.174
34.175
34.175
34.175
34.175
34.174
34.175
34.174
34.174
34.171
34.174
34.177
34.174
34.174

Long
-118.444
-118.445
-118.592
-118.584
-118.606
-118.606
-118.584
-118.605
-118.609
-118.576
-118.603
-118.606
-118.605
-118.604
-118.603
-118.589
-118.589
-118.591
-118.599
-118.591
-118.589
-118.592
-118.589
-118.590
-118.587
-118.590
-118.587
-118.587
-118.657
-118.587
-118.597
-118.592
-118.587

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted by Peak Ground Acceleration (As of 11/99) (continued)
Sty
16
21
2
3
3
3
3
5
22
3
3
3
3
3
3
1
1
1
1
1
1
1
1
1
1
2
3
3
3
3
4
5
5

Area
290600
300966
48000
nr
25600
30000
82000
376000
17500
20400
60000
63000
84000
120000
120000
16300
17700
19130
20000
20710
21500
22400
23400
23500
24670
62800
48000
48000
59700
75000
46375
58500
130000

Conn
962
672
5
76
62
50
108
960
34
54
80
68
132
140
216
8
8
8
12
8
8
8
12
8
10
100
84
84
36
114
112
100
288

Insp
50
34
5
8
16
13
65
621
10
10
14
68
33
124
106
8
8
6
12
8
8
8
9
6
10
97
16
79
36
22
112
97
284

Bot
0
0
0
0
0
0
11
72
0
0
0
9
0
0
30
0
0
0
0
0
0
0
0
0
0
7
0
1
7
0
0
0
23

B-21

Top
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3

B&T
0
0
0
0
0
0
0
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2

Total
0
0
0
0
0
0
11
81
0
0
0
9
0
3
30
0
0
0
0
0
0
0
0
0
0
7
0
1
7
0
0
7
28

DR
0.00
0.00
0.00
0.00
0.00
0.00
0.17
0.13
0.00
0.00
0.00
0.13
0.00
0.02
0.28
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.07
0.00
0.01
0.19
0.00
0.00
0.07
0.10

Shr
0
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1

PZ
0
0
0
0
0
0
6
0
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
0
0
5
0
0
0
0
0
0
1

MMI
8.2
8.2
7.9
8.1
8.1
8.1
8.1
8.1
8.1
8.2
8.2
8.2
8.2
8.2
8.2
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3

Pga
0.357
0.357
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362

Pgv
10.8
10.8
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-3
ID
9059
9012
9011
9009
9010
9107
9027
9001
9002
9087
9168
9122
9121
9071
9077
9074
9015
9082
9038
9081
9014
9080
9088
9177
9129
9159
9125
9185
9006
9153
9142
9040
9136

Lat
34.174
34.176
34.178
34.179
34.179
34.179
34.179
34.179
34.179
34.183
34.185
34.185
34.185
34.067
34.067
34.062
33.976
34.156
34.156
34.157
34.156
34.156
34.156
34.155
34.187
34.187
34.187
34.187
34.201
34.201
34.201
34.234
34.240

Long
-118.587
-118.597
-118.597
-118.605
-118.605
-118.600
-118.605
-118.605
-118.605
-118.597
-118.606
-118.597
-118.597
-118.445
-118.445
-118.433
-118.393
-118.482
-118.482
-118.485
-118.477
-118.482
-118.480
-118.472
-118.502
-118.502
-118.502
-118.502
-118.495
-118.495
-118.495
-118.562
-118.570

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted by Peak Ground Acceleration (As of 11/99) (continued)
Sty
5
6
6
11
11
13
20
22
26
4
5
18
18
6
6
27
6
6
6
13
6
12
16
3
3
3
3
3
2
3
3
3
6

Area
250000
93514
93514
216640
216640
332800
385387
450000
586000
106800
79000
345600
345600
120000
240000
364500
172000
126000
150386
247000
567000
186000
233600
43000
50000
50000
50000
50000
49000
40000
70000
17700
250000

Conn
372
192
192
342
342
520
880
624
900
136
210
274
278
415
360
766
180
378
204
572
964
392
558
96
57
57
69
75
64
64
64
46
50

Insp
35
192
192
342
342
507
867
603
874
102
57
114
251
33
36
22
137
8
32
63
189
94
121
20
15
13
20
14
64
64
11
9
50

Bot
0
59
55
25
45
46
5
21
41
32
0
14
23
0
0
0
15
0
0
0
14
1
19
0
0
0
0
0
4
0
0
0
0

B-22

Top
0
0
0
6
4
4
0
0
61
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

B&T
0
1
7
2
1
1
0
0
12
0
0
0
0
0
0
9
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0

Total
0
60
62
33
50
51
5
21
114
32
1
14
23
0
0
9
16
0
0
0
15
1
19
1
1
2
2
0
5
4
0
0
0

DR
0.00
0.31
0.32
0.10
0.15
0.10
0.01
0.03
0.13
0.31
0.02
0.12
0.09
0.00
0.00
0.41
0.12
0.00
0.00
0.00
0.08
0.01
0.16
0.05
0.07
0.15
0.10
0.00
0.08
0.06
0.00
0.00
0.00

Shr
0
6
1
9
3
6
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
0
0

PZ
0
0
0
6
4
0
0
2
0
18
0
0
0
0
0
0
0
0
0
0
0
0
7
0
0
0
0
0
0
0
0
0
0

MMI
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.3
8.4
8.4
8.4
8.4
7.4
7.4
7.6
7.4
7.7
7.7
7.7
7.8
7.8
7.8
8.0
8.1
8.1
8.1
8.1
8.2
8.2
8.2
8.9
8.9

Pga
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.362
0.363
0.363
0.363
0.366
0.381
0.381
0.381
0.381
0.381
0.381
0.381
0.384
0.384
0.384
0.384
0.384
0.384
0.384
0.384
0.384

Pgv
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
19.1
8.6
8.6
8.6
8.2
16.5
16.5
16.5
16.5
16.5
16.5
16.5
15.2
15.2
15.2
15.2
15.2
15.2
15.2
15.2
15.2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Table B-3
ID
9008
9155
9110
9094
9096
9031
9026
9095
9093
9097
9021
9020
9132
9039
9187
9049
9064
9172
9161
9201
9140
9045
9044
9023
9103
9102
9104
9100
9101
9199
9019
9033
9046

Lat
34.239
34.265
34.019
34.236
34.236
34.239
34.239
34.236
34.243
34.243
34.153
34.153
34.202
34.173
34.171
34.410
34.038
34.213
34.221
34.047
34.047
34.040
34.040
34.040
34.032
34.032
34.032
34.032
34.032
34.036
34.033
34.038
34.033

Long
-118.564
-118.573
-118.457
-118.528
-118.528
-118.568
-118.564
-118.528
-118.532
-118.532
-118.462
-118.462
-118.630
-118.561
-118.543
-118.459
-118.375
-118.475
-118.466
-118.434
-118.435
-118.438
-118.438
-118.438
-118.456
-118.456
-118.456
-118.456
-118.456
-118.444
-118.450
-118.440
-118.450

FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted by Peak Ground Acceleration (As of 11/99) (continued)
Sty
10
3
2
2
2
2
2
4
4
3
4
4
6
3
3
2
4
3
6
3
4
10
10
11
3
3
3
3
3
6
8
9
13

Area
275000
22000
112000
17930
31800
34600
98444
27880
124200
45900
61000
63640
50000
25000
53000
36000
80000
68000
75000
nr
58000
183600
222400
409300
13500
16800
21000
33600
51000
250000
230000
180000
330000

Conn
688
114
172
72
68
50
84
96
240
172
88
40
108
22
57
68
228
162
190
96
298
440
503
920
72
102
80
152
312
700
188
278
823

Insp
626
17
15
27
68
50
82
96
240
168
88
40
20
7
9
59
21
16
35
23
43
369
232
917
32
44
32
76
154
87
163
40
750

Bot
156
0
0
1
0
0
22
11
59
26
9
10
0
0
0
1
3
0
0
0
0
0
0
123
1
4
0
11
12
0
17
0
0

B-23

Top
5
0
0
0
0
0
0
0
3
0
1
0
0
0
0
0
0
0
0
0
0
0
0
5
0
0
0
0
0
0
0
0
0

B&T
2
0
0
0
0
0
3
1
16
3
2
3
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
2
0
0
0
0
0

Total
163
0
0
1
0
0
25
12
78
29
12
13
0
0
0
1
3
0
2
0
1
32
38
131
1
4
0
13
12
2
17
0
130

DR
0.26
0.00
0.00
0.04
0.00
0.00
0.30
0.13
0.33
0.17
0.14
0.33
0.00
0.00
0.00
0.02
0.14
0.00
0.06
0.00
0.02
0.09
0.16
0.14
0.03
0.09
0.00
0.17
0.08
0.02
0.10
0.00
0.17

Shr
0
0
0
0
0
0
2
3
0
0
1
6
0
0
0
0
0
0
0
0
0
0
0
3
0
0
0
0
0
0
4
0
0

PZ
23
0
0
0
0
0
5
0
4
3
2
1
0
0
0
0
0
0
0
0
0
0
0
19
0
0
0
0
0
0
0
0
0

MMI
8.9
9.0
8.0
8.9
8.9
8.9
8.9
8.9
9.0
9.1
8.1
8.1
8.2
8.2
8.1
8.6
7.6
8.1
8.1
8.0
8.0
8.0
8.0
8.0
8.1
8.1
8.1
8.1
8.1
8.1
8.1
8.1
8.1

Pga
0.384
0.384
0.395
0.403
0.403
0.403
0.403
0.403
0.403
0.403
0.414
0.414
0.416
0.417
0.418
0.425
0.428
0.439
0.439
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444
0.444

Pgv
15.2
15.2
11.7
18.3
18.3
18.3
18.3
18.3
18.3
18.3
14.5
14.5
19.7
12.9
13.0
16.7
14.1
13.7
13.7
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4
19.4

 FEMA-355E
Appendix B: Northridge Earthquake
WSMF Building Damage

Table B-3
ID
9105
9047
9146
9111
9065
9075
9062
9017
9090
9114
9089
9091
9098
9069
9068
9053
9086
9085
9179
9158

Lat
34.033
34.042
34.034
34.015
34.020
34.019
34.020
34.013
34.412
34.412
34.412
34.412
34.412
34.410
34.410
34.029
34.272
34.272
34.261
34.262

Long
-118.451
-118.444
-118.456
-118.491
-118.497
-118.498
-118.498
-118.493
-118.554
-118.554
-118.554
-118.554
-118.557
-118.574
-118.574
-118.480
-118.469
-118.469
-118.502
-118.503

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

185 Inspected WSMF Buildings Affected by the 1994 Northridge Earthquake
Sorted by Peak Ground Acceleration (As of 11/99) (continued)
Sty
13
3
9
3
4
6
13
5
1
2
2
2
3
1
4
4
4
4
3
3

Area
486720
54000
164311
23100
52000
90000
221000
102400
7500
15400
19400
57600
81600
27000
86000
54200
64000
64000
42000
67000

Conn
764
146
320
48
78
100
1042
96
30
16
40
480
nr
20
112
132
76
76
90
28

Insp
107
38
73
6
49
100
299
96
18
16
14
77
48
20
112
110
72
70
19
9

Bot
34
16
0
0
0
51
10
17
1
8
4
0
0
10
84
9
4
12
0
0

B-24

Top
2
0
0
0
0
1
0
16
1
0
0
0
0
0
0
1
0
0
0
0

B&T
12
1
0
0
0
7
0
15
2
0
0
0
0
0
0
0
0
0
0
0

Total
48
17
1
0
0
59
10
48
4
8
4
0
0
10
84
10
4
12
6
0

DR
0.45
0.45
0.01
0.00
0.00
0.59
0.03
0.50
0.22
0.50
0.29
0.00
0.00
0.50
0.75
0.09
0.06
0.17
0.32
0.00

Shr
1
0
0
0
0
0
0
16
0
0
0
0
0
0
41
5
0
1
0
0

PZ
2
0
0
0
0
0
0
10
0
4
0
0
0
9
21
0
2
4
0
0

MMI
8.1
8.2
8.2
7.9
7.9
7.9
7.9
7.8
9.0
9.0
9.0
9.0
9.0
9.1
9.1
8.1
9.9
9.9
9.4
9.4

Pga
0.444
0.444
0.444
0.571
0.571
0.571
0.571
0.572
0.586
0.586
0.586
0.586
0.586
0.586
0.586
0.642
0.686
0.686
0.841
0.841

Pgv
19.4
19.4
19.4
12.7
12.7
12.7
12.7
12.7
33.1
33.1
33.1
33.1
33.1
33.1
33.1
21.1
47.5
47.5
27.6
27.6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

APPENDIX C. OVERVIEW OF DAMAGE TO STEEL BUILDING
STRUCTURES OBSERVED IN THE 1995 KOBE EARTHQUAKE
THE 1995 HYOGOKEN-NANBU (KOBE) EARTHQUAKE
Masayoshi Nakashima
DISASTER PREVENTION RESEARCH INSTITUTE, KYOTO UNIVERSITY

C.1

Summary

This appendix presents an overview of damage to steel building structures observed
following the 1995 Hyogoken-Nanbu (Kobe) earthquake. Damage statistics are presented with
respect to the number of stories, type of structural framing, location of damaged elements and
severity of damage. Standard practices exercised in Japan before the earthquake and causes of
damage discussed immediately after the earthquake are introduced in terms of materials,
welding, beam-to-column connection details and seismic design forces. Efforts are made to
compare these with corresponding U.S. practices. A partial summary of post-Kobe research
activities in Japan on steel structures is also presented.
C.2

Introduction

The Hyogoken-Nanbu (Kobe) earthquake shook Kobe and surrounding areas on January 17,
1995, exactly one year after the Northridge earthquake. More than 5,000 people were killed,
35,000 people were injured and 300,000 more were rendered homeless by this earthquake. The
direct cost of structural damages caused by the earthquake exceeded $150 billion. Substantial
damage was experienced by reinforced concrete and steel highway structures, as well as by
wood, concrete and steel buildings.
In the long history of large Japanese earthquakes, the Kobe earthquake was the first to cause
widespread and serious damage to modern steel buildings. Numerous steel buildings had been
shaken strongly by other recent earthquakes, such as the 1978 Miyagiken-oki earthquake that
struck urban Sendai City. However, these recent earthquakes have caused only minimal damage
to steel buildings.
Why did the Kobe earthquake damage modern steel buildings so badly? Many suggest that
ground motions in Kobe were much larger than those experienced during previous earthquakes
in Japan. Moreover, the Kobe area is one of the earliest urban developments in Japan and,
consequently, contained a large inventory of older steel buildings designed when design criteria
were not as advanced as today. Whatever reasons may be given, the fact remains that modern
steel buildings experienced significant damage, refuting the popular myth in Japan that steel
buildings are immune to strong earthquakes.
To help put this situation into perspective, this section presents (1) an overview of damage to
steel buildings observed following the Kobe earthquake, (2) plausible causes of this damage as
discussed by the Japanese engineering community immediately after the earthquake, (3) some

C-1

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

comparisons between the types of damage observed and the post-earthquake actions taken in the
U.S. and Japan, and (4) a partial summary of the efforts being conducted in Japan with respect to
the improvement of seismic safety of modern steel buildings.
C.3

Damage to Steel Buildings

Steel Building Construction in Japan
Steel is a very popular structural material in Japanese building construction. Figure C-1(a)
compares the total floor area of steel buildings constructed each year with that employed in
construction using other structural materials. Wood has ranked first for years, but it is used
almost exclusively for residential houses. Steel ranks second, followed by reinforced concrete
(RC) and steel-encased reinforced concrete (designated SRC in Japan). Figure C-1(b) shows the
total floor area of steel buildings constructed each year with respect to the number of stories in
each building. This figure suggests that the vast majority of steel buildings are shorter than five
stories in height. In fact, most of steel buildings in Japan are low-rise, used for offices, shops
and mixed occupancies, as well as for industrial and manufacturing structures.
Damage to Older Steel Buildings
The Architectural Institute of Japan (AIJ) conducted a preliminary field reconnaissance of
Kobe from January 24 to 26, 1995, and identified 4,530 engineered buildings that were damaged,
including 1,067 that collapsed or were damaged beyond repair (AIJ 1995a, 1997a). The Kobe
area contains many engineered buildings constructed before the major economic growth of the
post-war era. As a consequence, a large stock of steel buildings more than 35 years old were
subjected to the effects of the ground shaking.
Figure C-2 shows examples of damage to such older steel buildings. As shown in Figure C2(b), these buildings generally were constructed with bundled light-gauged sections for columns;
beams were typically fabricated using shallow trusses consisting of light-gauge rolled sections
and round bars. Unfortunately, these old buildings lacked significant earthquake resistance, in
comparison with modern seismic design codes. Many of these buildings also suffered
considerable corrosion and other material deterioration, as shown in Figure C-2(c). According
to a preliminary estimate, over 70 % of the damaged steel buildings located in Kobe City were of
this older construction type.
Damage to Newer Steel Buildings
From mid-February to mid-March, 1995, the Steel Committee of the Kinki Branch of the
Architectural Institute of Japan conducted a detailed survey of the damage to modern steel
buildings and located 988 damaged steel buildings (AIJ 1995b). Among those buildings, 90
were rated as having collapsed, 332 as being severely damaged, 266 as moderately damaged, and
300 had minor damage. Figure C-3 shows the number of buildings with respect to the damage
level, indicating that most of the collapsed buildings were 2 to 5 stories tall. No building with
seven or more stories collapsed.

C-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

Damaged buildings were classified as having rigid moment-resisting (R) frames or braced
(B) frames. Thus, considering the two principal framing orientations of a building results in the
following framing designations: R-R (an unbraced moment resisting frame in two horizontal
directions), R-B (an unbraced frame in one horizontal direction and a braced frame in the other
direction), and B-B (a braced frame in both horizontal directions). Considering the 988 damaged
steel buildings, 432 were R-R, 134 were R-B and 34 were B-B, with 388 with an unidentified
framing system. These statistics indicate that the majority of damaged buildings had momentresisting frames. Although these statistics were for damaged buildings only, they are also
believed to reflect the general distribution among framing systems used in modern Japanese steel
buildings.
Table C-1 indicates cross-sectional types used for columns, beams, and braces. Beams
consisted almost exclusively of wide-flange sections, either rolled or built-up. For columns,
wide-flange (H) sections were used most extensively, followed by square-tube sections. During
the past 25 years, it is notable that square-tube (commonly cold-formed for low- to moderate-rise
construction) sections have been used more frequently for columns than have wide flange
sections. In braced frames, rods, angles, flat bars, round-tubes, wide-flanges, square-tubes, and
channels were used for the braces.
Table C-2 indicates the type of connection details found in the damaged buildings. In
moment-resisting frames, short stubs of wide flange beams are generally shop welded to the
columns in a so-called Christmas tree arrangement. The beam stubs were then field bolted
(using high-tension bolts) to the central portion of the beam. Column splices were mostly
accomplished by welding. In braced frames, braces were connected mostly by bolting, except
for small rod and flat bar braces, which are generally welded.
Figure C-4 shows two typical types of Japanese beam-to-column connections, namely the
through-diaphragm connection and the interior diaphragm connection. Of these connections, the
through-diaphragm connection is by far most popular, as indicated in Table C-2(c). In the
through-diaphragm connection, the square-tube used for the column is cut longitudinally into
three pieces: one used for the column of the lower story, one for the connection's panel zone (a
short piece, often called a ‘dice’ in Japan), and one for the column of the upper story. Two
diaphragm plates are inserted between the three separated pieces and shop-welded all around.
Then, short segments of beams are welded to the panel zone: the beam flanges are welded
directly to the diaphragms and the web to the side of the dice. The entire piece (often called a
Christmas tree) is transported to the construction site and connected to the mid-portion of the
beam by high-tension bolts.
Figure C-5 shows three typical types of column base connections, including the standard base
plate connection, the concrete encased column base connection, and the embedded column base
connection. As evidenced in Table C-2(d), standard base plate connections were most
commonly used.

C-3

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

C.4

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

General Damage Statistics for Modern Steel Buildings

Figure C-6 shows the correlation between damage level and structural framing type, which is
further sub-divided according to the type of column used. This figure indicates that no
significant difference existed in damage level with respect to the structural framing type (R-R, RB, and B-B). Interestingly, buildings with wide-flange columns suffered somewhat more serious
damage in comparison with buildings having other column types. This may be attributed to
building age, since square-tube sections are generally used in recent construction. Figure C-7
shows the location of damage (columns, beams, beam-to-column connections, braces, and
column bases) as a function of frame type. Major observations from the data in this figure are as
follows: (1) columns in unbraced frames suffered the most damage relative to other frame
elements (in terms of the number of buildings), while braces in braced frames were the most
frequently damaged structural element; (2) in unbraced frames, damage to beam-to-column
connections and column bases was also significant; (3) damage to beam-to-column connections
was most significant for unbraced frames having square-tube columns; and (4) damage to
columns was most significant for unbraced frames having wide-flange columns. Unfortunately,
these particular statistics are limited in that they do not include information on the building age
or the types of connections and members used in non-damaged steel buildings.
The Building Research Institute (BRI) of the Ministry of Construction of Japan conducted a
more detailed survey of about 630 steel buildings (not including old steel buildings made of
light-gauged sections) located in a severely shaken area (Midorikawa et al. 1997). According to
the BRI survey, the approximate distribution of damage severity was 17% collapse/severe, 17%
moderate, 33% minor, and 33% no damage. The incidence of damage to columns, braces, and
column bases were found to be significantly lower for buildings constructed after 1981 than for
those constructed before. This may be a function of the major changes in Japan’s Seismic
Design Code that occurred in 1981, or other changes in construction practices that occurred over
time. On the other hand, no significant difference was found in the BRI study between damage
to beam-to-column connections in buildings constructed before and after 1981.
C.5

Damage to Members in Modern Buildings

Columns
Damage occurred in many columns. Most of this damage occurred near beam-to-column
connections. Damage to columns included plastification, excessive distortion and local buckling
near the column ends, as well as fractures in the base metals and at column splices. Many wideflange columns sustained excessive bending about their weak-axis. Figure C-8 shows a fracture
in the base metal of a cold-formed square tube having dimensions of 450 mm x 450 mm x 25
mm. The fracture surface was mostly brittle, but significant local buckling appeared above the
fractured section, suggesting that the column had been loaded well beyond its yield stress.
A cluster of modern high-rise residential steel buildings, located near the seashore in
Ashiyahama, exhibited fractures in over 50 large-sized columns and braces The fractured
columns were made of square sections (not cold-formed sections), with an outer dimension of
500-550 mm and a wall thickness of 50-55 mm. Fracture occurred in the base metals (Figure C-

C-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

9(a)), at welded column splices, and at beam-to-brace connections (Figure C-9(b)). In these
fractures, surfaces were rather rough, exhibiting shear lips and tear ridges, which confirmed that
the fracture was brittle involving only limited plastification.
Braces
Damage to braces was found to be more severe in relatively smaller cross-sections (rods,
angles, and flat plates). Although this was not quantified, the size of cross-sections appears to
correlate strongly with the building age, with small cross-sections used more frequently in older
buildings. Damage to braces with larger cross-sections was concentrated mostly at their
connections with the adjoining beams or columns. Figure C-10(a) shows an example of such
damage, in which connection bolts were broken. Another example, in which a beam connected to
a pair of braces sustained significant web-plate buckling and out-of-plane distortion, is shown in
Figure C-10(b). On the other hand, it was not unusual for braced frames located in severely
shaken areas to have suffered only minimal damage; in these cases, the braces were firmly
connected to the adjoining beams and columns. In summary, damage to braces having large
cross-sections was strongly correlated with the connection details, and poor connection details
suffered more severe damage.
Column Bases
Damage to column bases was relatively common. Most of the damage to column bases was
observed for the standard base plate connections. Figure C-11 shows the damage level with
respect to the location of damage in the standard base plate connections, indicating that the
majority of damage occurred to the anchor bolts. In present seismic design practice, standard
base plate connections are commonly designed assuming them to be pin supports (meaning that
no moment transfer is considered for design at the column base). Regardless of such
assumptions, column bases must securely withstand shear forces under load reversals. The
design profession in Japan has come to recognize that some of the practices used for the design
of these connections before the Kobe earthquake were inadequate.
Beam-to-Column Connections
As stated earlier, many fractures were observed in beam-to-column welded connections.
Fractures to beam-to-column connections were essentially divided into two types. Figure C-12
shows the first type of fracture, which occurred in beam-to-column (shop-welded) connections
where columns, beam, and connection panels were fillet-welded using rather small sized welds.
At a glance, it was understood that such small welds could hardly transfer the stresses exerted on
the connections. In fact, they generally fractured without any observed plastification in the
neighboring columns and beams.
The second type of fracture was observed in many beam-to-column connections that
employed full-penetration welding. Fractures were mostly brittle and occurred in the weld
deposit, heat-affected zones, base metals (initiating from the toe of the weld access holes), and
diaphragm plates (Figure C-13). From instances where such fractures occurred, the following
observations could be made: (1) residual story drifts were not significant; (2) damage to interior

C-5

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

and exterior finishes were minimal; (3) fractures occurred mostly at beam bottom flanges only;
(4) significant yielding, plastification and local buckling of beam bottom flanges were observed,
indicating that the beams dissipated some energy before fracture; and (5) such plastification
occurred only in the beams while the adjoining columns remained almost elastic. The reason for
observation (5) may be that many designers adopted strong column – weak beam proportioning
concepts, as well as the fact that the real yield strength of a square tube column was usually
significantly higher than the nominal value due to the cold-forming of the steel during the
manufacture of square tube sections.
Figure C-14 shows a set of results of tests conducted for the base metal of a fractured beam
(Nakashima et al. 1998). The Charpy V-notch test shows that the material could absorb more
than 50 Joules of energy prior to fracture at zero degree centigrade. The base metal near the
fractured surface was significantly hardened, which suggests the base metal in the fractured
connection might have sustained significant plastic strain before fracture.
C.6

Design and Construction Practices Before Kobe Earthquake Damage

Among various types of the damage described above, the damage to welded beam-to-column
connections of modern steel buildings has posed one of the most serious problems in the steel
building community. The following summarizes pre-Kobe design and fabrication practices for
Japanese steel buildings. Also noted are some relations between these practices and damage as
discussed by the Japanese design community immediately after the Kobe earthquake.
Materials
Japan uses both the integrated casting and continuous casting (mini-mill) steels for beam
members. The market share of continuous casting steels is higher for smaller, thinner sections
(commonly less than 450 mm deep and thinner than 40 mm). Immediately after the Kobe
earthquake, some concerns were raised about the fracture toughness of such steels, particularly at
the flange-web junction. However, it was confirmed that the steels in recent continuous casting
production are equivalent in general quality to corresponding steels from integrated casting mills.
Further, it is notable that large construction companies establish their own in-house regulations
for quality assurance when using steels from continuous casting mills.
At the time of the Kobe earthquake, Japan was in a process of introducing new types of
structural steels, called the SN steels. The development of SN steel initiated long before the
Kobe earthquake. Both upper and lower limits are specified for the yield and maximum
strengths of this steel, and an upper bound of 0.8 is specified for the ratio of the yield to the
maximum strength in order to ensure good ductility. These steels have been commercially
available and, particularly after the Kobe earthquake, designers have been more interested in the
use of these steels.
Connection Details
As previously discussed, the design and fabrication of recent Japanese beam-to-column
connections are characterized as follows: (1) square tube (often cold-formed) sections are used

C-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

for columns; (2) the through-diaphragm connection (Figure C-4(a)) is by far most popular means
for accomplishing the beam-to-column connection; and (3) welding is accomplished in the shop,
where automatic welding robots are increasingly used in some shops.
A constructor conducted a detailed survey on several damaged steel buildings in which shopwelded through-diaphragm connections were used (1997a). Out of 2,396 connections surveyed,
79 were found to have damage including complete fractures and partial cracks. The damage
percentage is 3.3%. Figure C-15 shows the fracture and crack distribution with respect to the
damage location, indicating that 20.5 % of the damaged connections fractured in the base metal.
In these instances, the fracture initiated from the toe of the weld access hole. It was speculated
that this type of fracture was attributed primarily to a combined effect of stress and strain
concentration at the toe and low fracture toughness of the material at the flange-web junction.
After the Kobe earthquake, extensive studies have been undertaken to determine how to avoid
this type of fracture, and various modified connection details have been proposed, as will be
explained later. Most of the modified details reduce the size of the weld access holes, aiming to
mitigate the stress concentration at the toe.
In the through-diaphragm connection, the beam web is shop-welded directly to the column
flange, but normally the column is supplied without any vertical stiffener or diaphragm at the
back of the beam web (Figure C-4(a)). Because of the flexibility of the column flange in out-ofplane deformation, the moment resisted by the beam web is reduced compared to that developed
for an ideal rigid support, causing stress concentrations in the beam flanges. Concerns were
raised about this issue after the earthquake.
As stated earlier, most of the fractures in beam-to-column connections occurred at beam
bottom flanges only, which is very similar to what has been observed from the 1994 Northridge
earthquake. In the U.S., the coincidence of the weld root, backing bar, and most stressed fiber
due to local deformations was said to be a source of bottom flange fractures. On the other hand,
Japanese shop welding enables welding of the beam bottom flange from the bottom side, as
shown in Figure C-4(a), where the weld root is located on the interior side of the bottom flange.
The observation that many of Japanese shop welded beam-to-column connections also fractured
only at bottom flanges indicates that the root location is not the sole cause of bottom flange
fractures. Composite action with floor slabs, another source being addressed, is a strong suspect
because in Kobe some beam-to-column connections fractured in both the top and bottom flanges
when floor slabs were not present.
Cold-formed tubes are limited to thicknesses less than 40 mm and dimensions less than 1000
mm. Therefore, columns for tall buildings and columns with large axial loads are generally
constructed as heavy built-up columns with the connection shown in Figure C-4(b). The built-up
box sections require extensive fabrication labor because of the longitudinal seam and internal
diaphragm welds. This welding is completed in the shop. When such large, heavy box columns
are used, connections with shop-attached beam stubs projecting from the column (Figure C-4(a))
are no longer feasible, because of difficulties in transportation and the size of the required bolted
beam splice. Thus, connections similar to the U.S. practice (i.e. field welded beam flange to
column joints and bolted web attachments) are employed. No serious damage was reported for
this type of connection (AIJ 1995b) during the Kobe earthquake.
C-7

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Welding
In recent Japanese construction, semi-automatic CO2- (or sometimes argon-mixed) shielded
metal arc welding has been used almost exclusively for the welding of beam flanges. This
practice is common not only in shop welding, but in field welding as well. Self-shielded fluxcored electrodes were introduced in the late 1960s and used in the early 1970s primarily for field
welding, with electrodes developed in Japan. It lost favor, however, and gas shielded metal arc
welding has been used almost exclusively in recent years.
As shown in Figure C-15, 24.4 % of the damaged connections had complete fracture along
the weld metal, 10.3 % had cracks at craters, and 37.2 % had cracks initiating from run-off tabs.
This large percentage of damage associated with welding was striking, and serious concerns
were raised about the present welding practice. Higher voltages and larger deposition rates than
those stipulated in regulations and excessive weaving were thought to be the likely causes.
Seismic Design Forces
The present Japanese seismic design code, adopted in 1981, provides two levels (Levels-I
and -II) of design earthquake forces. Level-I is for small to medium earthquakes with maximum
ground accelerations ranging between 0.8 and 1.0 m/sec2. To ensure serviceability, structural
systems are required to remain elastic during such earthquakes. The Level-II design earthquake
represents a large earthquake with the maximum ground acceleration ranging approximately
from 3.0 to 4.0 m/sec2. For such large earthquakes, collapse prevention is a typical design
consideration, and some damage to structures (meaning plastic deformation in members and
connections) is permitted. Based upon these maximum ground accelerations, Japan’s Seismic
Design Code stipulates 0.2g and 1.0g (g is the acceleration of gravity) as the maximum design
base shear for Levels-I and -II, respectively. For Level-II, a trade-off between strength and
ductility capacities is taken into account, and a reduction factor of 4.0 is adopted for most ductile
steel moment frames. Figure C-16 shows the Level-II design base shear unreduced for ductility
corresponding to medium soil conditions. Also drawn in this figure are pseudo-acceleration
spectra (with 2% damping) for a dozen large ground motions recorded in the Kobe earthquake.
This plot clearly indicates that some of response accelerations are significantly larger than code
specified values. Considering such large recorded ground motions as well as Japan’s seismic
design philosophy, it was understood that steel buildings located in severely shaken regions were
expected to sustain significant structural damage even if they were built in conformance with the
present design and construction practices.
C.7

Comparison of Building Damage in the U.S. and Japan

Damage Similarities
In both the Northridge and Kobe earthquakes, steel buildings sustained significant damage,
and many similarities in damage patterns were disclosed. Some notable similarities are
summarized as follows.

C-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

1. Steel buildings in Japan and the U.S. had not experienced much damage in previous
earthquakes. These two earthquakes exposed for the first time in each country the potential
significant damage in welded steel moment resisting frame buildings.
2. Many modern building structures designed and constructed with present practices were
damaged. Thus, damage was not exclusively associated with old technologies and design
practices.
3. Much damage was found. However, no building constructed using the most recent design
and construction practices collapsed.
4. Many welded beam-to-column connections failed by fracturing, indicating that welded
connections were one of the weakest locations in steel moment frames.
Differences in Damage, Design and Construction
Differences in damage patterns and sources were also observed. Notable differences are
summarized as follows.
1. Beam-to-column connections fractured, but in many instances fractures in Japanese
structures were preceded by significant plastification and local buckling, meaning that the
beams dissipated some energy before fracture. The vast majority of fractures in the U.S.
involved virtually no plastification in either beams or columns. Thus, the degree of plastic
rotation capacity of steel beams-to-column connections constructed using pre-Northridge and
pre-Kobe practices may have been significantly different.
2. Steel materials used may also be different. It appears that Japan has placed relatively more
attention to the importance of material strength and strain hardening for securing beam
plastic rotation capacity. Development of SN steels before the Kobe earthquake may be an
indirect indication of this. The use of so-called dual-certified steels in the U.S prior to the
Northridge earthquake suggests less concern in this regard.
3. Welding processes and procedures are significantly different between the two countries.
Japan almost exclusively employs gas-shielded metal arc welding with solid wires, whereas
self-shielded flux-cored welding is commonly used in the U.S. Japanese welding is often
conducted in the shop as shown in Figure C-4(a), whereas the critical welding of beams to
columns in the U.S. is commonly done in the field.
4. Connection details are also different. Japan construction typically uses square tube (box)
columns, whereas wide-flange columns are usually employed in the U.S. This difference is
accompanied by many differences in local connection details, such as through-diaphragm
connections in Japan versus through-column connections in the U.S., and welded web to
column joints in Japan versus bolted web to column in pre-Northridge connections in the
U.S.
5. The redundancy of the moment frame system is not the same in the two countries. All
beam-to-column connections are rigidly connected in Japan, whereas in the U.S. rigid
connections are commonly assigned only to selected locations. In addition to the degree of
redundancy, this difference affects member sizes, the importance of gravity loads relative to
seismic loads, and the stress condition (bi- versus uni-directional bending) in columns.

C-9

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

C.8

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Partial Summary of Post-Kobe Japanese Research

Research Efforts
As a natural consequence of the damage disclosed in Kobe by the Kobe earthquake, major
research and development programs have been undertaken in Japan. Japanese post-Kobe steel
research efforts aim at (1) reevaluation and upgrading of plastic rotation capacity of welded steel
connections and (2) characterization of plastic rotation demanded of these connections. The
latter effort is closely associated with a revision of the Japan’s Building Law. The new law
passed the Japanese parliament in July 1998, and a detailed design code that supports the
implementation of the revised law will become available in the summer of 2000.
The complete body of the research leading to these code changes, conducted in universities
and by government and industry organizations, is too extensive to summarize herein. A partial
description of large coordinated research/development efforts conducted after the Kobe
earthquake follows. The Japanese Ministry of Education provided a four-year grant-in-aid
(1996-1999) for studying urban disaster mitigation measures. The principal investigator of this
project is Prof. Kenzo Toki of Kyoto University. This project is broad based, ranging from
ground motion research to studies of societal impacts and risk management. Part of this project
deals with steel structures. The Japanese Ministry of Construction (MOC) undertook a threeyear comprehensive project (1996-1998) for improving Japanese steel construction, in which
issues related to materials and welding, beam-to-column connections, and plastic rotation
demands were investigated. Prof. Koichi Takanashi of Chiba University chaired the oversight
committee of this project. Three volumes of the project report were released in the spring of
1999 (MOC 1999), and efforts still continue to develop guidelines for design and fabrication of
steel moment frames. The Steel Committee of the Kinki Branch of the Architectural Institute of
Japan conducted a two-year study (1996-1997) on welded beam-to-column connections. The
project leader was Prof. Kazuo Inoue of Osaka University. In that study, 86 full-scale beamcolumn sub-assemblages were tested considering various parameters including connection
details (weld access holes, run-off tabs, etc.), welding procedures, rate of loading, and
temperature (AIJ 1997b). The Japan Society of Steel Construction (JSSC) conducted a two-year
study (1995-1996), lead by Prof. Ben Kato of Toyo University, on improvement of welded
connections, and published a guideline for design of steel moment frames (JSSC 1997). The
Japan Welding Engineering Society (JWES) has been conducting a comprehensive research
project on both demand and capacity issues for steel moment frames and their welded
connections. Prof. Koichi Takanashi of Chiba University heads the project. In 1996, they
released an interim report consisting of three volumes (JWES 1997). From 1995 to 1997, a
research group lead by Prof. Hiroshi Akiyama of University of Tokyo conducted a series of
dynamic loading tests on full-scale steel connections using a large shaking table (15 m by 15 m).
The results have been published in several professional journals (for example, see Akiyama et al.
1998). A book that completely describes this project is planned. In 1997, the Science and
Technology Agency (STA) started constructing a larger multi-axis shaking table having a
dimension of 20 m by 15 m. Completion of the table and associated facilities is expected in the
year of 2005, and various structures and structural components are planned for testing.

C-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

Code Changes
In response to the urgent need to upgrade design and fabrication practices for steel building
structures, changes have already been made in a few Japanese design codes and guidelines.
A steel fabrication specification, called “JASS-6,” published by the Architectural Institute of
Japan (AIJ 1996), was revised in 1996. It contains new recommendations with respect to the
shapes and sizes of weld access holes. Many of the newly recommended details utilize smaller
hole sizes so that stress and strain concentrations would be mitigated at the toe of the weld access
hole. Whether or not backing bars and run-off tabs should be removed has remained a subject of
continuing debate in Japan, but the revised specification does not require their removal.
The Japan seismic design code is reviewed regularly, and minor revisions are made every
few years. The last revision was made in 1997 (BCJ 1997), in which many sections related to
steel buildings were amended to reflect the damage observed in the Kobe earthquake. Notable
changes include among other items: (1) introduction of SN steels, (2) new design procedures for
cold-formed steel tubes, (3) description of required material properties, and (4) detailed design
procedures for column bases.
Japan’s Building Law was revised in July 1998, and the Ministry of Construction is
undertaking efforts to establish a detailed design code that supports the implementation of the
revised law. In the new code, deformation demand and capacity are to be considered more
explicitly. The code is to be enforced beginning in the summer of 2000.
Differences in Post-Earthquake Actions in U.S. and Japan
Damage to steel buildings in Northridge and Kobe was believed to have occurred because of
a mixture of various sources related to design, materials, welding, connection details, and
structural systems. This understanding is shared between the U.S. and Japan. However,
solutions being provided after few years of postearthquake efforts appear to be significantly
different in many aspects between the two countries. Some examples of differences, particularly
related to beam-to-column connections, are summarized below.
Materials
Use of materials with larger ductility can be a solution toward higher seismic performance of
steel buildings. Japan developed a new type of steels having a good margin between the yield
and ultimate stresses, a smaller variation in these specified stresses, and larger fracture
toughness. Use of the new steels is still optional at the present time, but their use has been
increasing significantly. The U.S. also introduced a new steel grade and revised specifications
for testing of materials to be used in seismic details. It appears that the utilization of special high
toughness steels is gaining acceptance faster in Japan than in the U.S.

C-11

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Welding
Fractures at weld metals were very serious in the U.S., and use of different electrodes having
a larger toughness and controlled deposition rate is now mandatory. In Japan, fractures at weld
metals were also disclosed in many instances, and welding with stringer bead placement to avoid
too large heat input is strongly recommended. Efforts to develop tougher electrodes are also
underway in Japan (for example, see JWES 1996). In general, however, the U.S. is more explicit
as to the changes in welding and inspection practices.
Connection Details
Regarding connection details, the U.S. has pursued three courses: moving the plastic hinge
away from the beam end, improving the local details and in situ material properties for
conventional unreinforced connections, and substitution of welded connections by bolted
connections. Many believe that moving the plastic hinge region is the most secure way to
improve the ductility capacity of beam-to-column connections. Many new details have been
developed along this line, such as strengthening of beam ends by cover plates, haunches, ribs,
etc. or trimming beam flanges at a location away from the column face (named the Reduced
Beam Section (RBS) connection). Such strengthening is considered as a possible solution also in
Japan, but the general sentiment is that sufficient ductility capacity can be ensured by modifying
connection details combined with good welding. Many efforts have been made to modify details
by changing the size and shape of weld access holes, etc (Figure C-17). After five years of
extensive studies, it has been felt that connection details without any weld access holes (shown in
Figure C-17(c)) can ensure the most ductile performance among the various post-Kobe
connection details considered.
Of immediate interest is whether a U.S.-style RBS connection or a Japanese-style connection
without weld access holes (designated the no-hole connection) has larger deformation capacity.
To provide some quantitative information on this issue, an experimental study was conducted as
part of a U.S.-Japan Cooperative Research Project on Urban Disaster Mitigation (Suita et al.
1999). In the tests, all conditions including material properties, sectional properties, fabricator,
welder, and loading history were identical, except for the connection details as shown in Figure
C-18. The design procedure proposed by Engelhardt (1999) was adopted for trimming the beam
flanges in the RBS connection. The no-hole connection is one of the recommended connections
adopted in the AIJ’s steel fabrication specification (AIJ 1996). Examples of test results in terms
of the beam end moment versus rotation relationship are shown in Figure C-19. In both the nohole and RBS connections (Figure C-19(a), (b)), no early fracture occurred at or in the vicinity of
welds, and no strength deterioration was observed up to the plastic rotation of about 0.03 to 0.04
radians. The strength reduction during cycles with larger rotation amplitudes was gradual,
caused primarily by the progress of local buckling to beam flanges. Ductility capacity of the two
connections was judged to be nearly the same in this test program. For comparison purposes, an
additional specimen was fabricated with the pre-Kobe conventional detail having the then
standard weld access holes (Figure C-17(a)). The specimen failed in fracture initiated from the
toe of the weld access hole, with the rotation capacity significantly smaller than the specimens
having other two connections (Figure C-19(c)).

C-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

Structural Systems
As to the structural system employed, it is not likely that Japan will abandon square tube
(box) columns and switch to wide flange columns at least for the foreseeable future. Similarly,
the practice of using rigid connections for all beam-to-column connections will also likely
remain unchanged. By the same token, the U.S. practices of using wide-flange shapes for
columns and providing moment resisting connections in a small portion of the structure will
likely remain unchanged as well.
C.9

Conclusions

This section has provided an overview of damage to steel building structures observed in the
1995 Hyogoken-Nanbu (Kobe) earthquake, and postearthquake activities being conducted in
Japan. Although the U.S. and Japan experienced similar damage, a closer look indicates
significant differences in the causes of damage. These differences most likely stem from the
variation in design, detailing, fabrication, and construction practices between the two countries.
It also appears that postearthquake approaches to resolve problems encountered in Kobe and
Northridge also appear to be different in many aspects. After the Northridge and Kobe
earthquakes, extensive technical exchange has been conducted between the U.S. and Japan.
Japanese researchers and professional engineers have learned much from this exchange, and
have benefited a great deal from U.S. efforts by the FEMA/SAC Steel Project and others.
Nevertheless, Japanese approaches differ in many aspects from U.S. approaches, primarily as a
result of differences in construction culture and philosophy as a whole. The writer wishes that
U.S. readers recognize these differences when referring to Japanese literature.
Table C-1

Cross-Sections Used In Damaged Steel Buildings; (a) Columns; (b) Beams;
(c) Braces
(a)

(b)

C-13

(c)

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Table C-2

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Types Of Connections Used In Damaged Buildings;

(a) Columns; (b) Beams; (c) Beam-To-Column Connections;
(d) Column Bases
(a)

(b)

(c)

(d)

Annual construction (floor area × million m2)
120
Wood

Steel

100
80
60
RC

40
20

SRC

0
70 72 74 76 78 80 82 84 86 88 90 92 94
Year

(a)

(b)

Figure C-1 Market Share For Japanese Steel Building Construction; (a) Floor Areas
with Respect to Structural Material; (b) Floor Areas with Respect to Number of Stories

C-14

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

(a)

(b)

(c)
Figure C-2

Damage to Old Steel Buildings; (a) Collapse; (b) Construction with LightGauged Sections; (c) Corroded Sections

C-15

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Figure C-3

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Damage Level with Respect to Number of Stories

(a)
(b)
Figure C-4 Types of Beam-to-Column Connections Popular in Japan; (a) ThroughDiaphragm Connection; (b) Interior-Diaphragm Connection

C-16

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

(a)

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

(b)

(c)

Figure C-5 Types of Column Bases Popular in Japan; (a) Base Plate Connection;
(b) Concrete Encased Column Base; (c) Embedded Column Base

Figure C-6

Distribution of Damage Level with Respect to Structural Type

C-17

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Figure C-7

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Damage to Structural Members with Respect to Structural Type

Figure C-8

Fracture at Cold-Formed Square Tube Column

C-18

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Figure C-9

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

(a)
(b)
Fracture of Square Tube Jumbo Columns; (a) Fracture at Base Metal;
(b) Fracture at Brace-To-Column Connection

(a)
(b)
Figure C-10 Damage To Brace Connections; (a) Breakage of Bolts; (b) Beam Web
Buckling and Out-of-Plane Deformation

C-19

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Figure C-11 Damage Location and Level of Column Base Connections

(a)
(b)
Figure C-12 Fracture at Beam-To-Column Connections with Fillet Welding of Small
Sizes; (a) Fracture at Column Top; (b) Fracture at Beam End

C-20

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

(a)
(b)
Figure C-13 Fracture at Beam-To-Column Connections with Full Penetration Groove
Welding; (a) Fracture at Base Metal Initiating from Toe of Weld Access Hole; (b) Fracture
Involving Yielding and Local Buckling

(a)

(b)

Figure C-14 Material Properties of Base Metal Near Fractured Surface; (a) Charpy VNotch Test; (b) Hardness Test

C-21

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

40
(1)
(2)
(3)
(4)
(5)
(6)

30
20
10

complete fracture at base metal
complete fracture at weld metal
crack at crater
crack from run-off tab
crack to web initiated from weld access hole
crack to diaphragm plate initiated from weld

0
(1)
1

(2)
2

(3)
3

(4)
4

(5)
5

(6)
6

Figure C-15 Distribution of Damage To Beam-To-Column Connections with Respect to
Type and Location

Pseudo Acceleration Response
(m/s/s)

40

30

20

10

0

0

1

2
Period (T)

3

4

Figure C-16 Design Base Shear of Level-I Japanese Seismic Design Code and Pseudo
Acceleration Response Spectra of Large Ground Motions Recorded in Kobe Earthquake

C-22

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

(a)

FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
Observed in 1995 Kobe Earthquake

(b)

(c)

Figure C-17 Weld Access Hole Details Proposed after Kobe Earthquake; (a) Pre-Kobe
Standard Detail; (b) Modified Detail with Smaller Hole; (c) No-Hole Detail

(a)

(b)

Figure C-18 Comparison Between U.S. RBS Connection and Japanese No-Hole
Connection; (a) No-Hole Connection Specimen; (b) RBS Connection Specimen

C-23

 FEMA-355E
Appendix C: Overview of Damage to Steel Building Structures
in 1995 Kobe Earthquakes

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

M m /M p

M m /M p

M m /M p

1.0

1.0

1.0

θ m (rad)

-0.04

0.04

-0.04

θ m (rad)

0.04

θ m (rad)

-1.0

-1.0

-1.0

RBS1

NOHOLE1

(a)

-0.04

0.04

(b)

Conventional

(c)

Figure C-19 Beam End Moment Versus Beam Rotation Relationships Obtained from
Test; (a) No-Hole Connection; (b) RBS Connection; (c) Japanese Pre-Kobe Connection

C-24

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix D: Damage to Steel Buildings Due to the
September 21, 1999 Ji Ji, Taiwan Earthquake

APPENDIX D. DAMAGE TO STEEL BUILDINGS DUE TO THE
SEPTEMBER 21, 1999 JI JI, TAIWAN EARTHQUAKE
(Compiled by Steve Mahin with contributions from K.C. Tsai, National Center for Earthquake
Engineering Research, National Taiwan University; S.J. Chen, National Taiwan University of
Science and Technology; C.M. Uang, University of California, San Diego.)
A series of damaging earthquakes struck central Taiwan starting on September 21, 1999.
The largest shock was assigned a 7.6 Richter magnitude by USGS. More than 9,000 aftershocks
followed the initial event. The epicenter was located in Nantou County, about 150 km south of
Taipei. Many structures were heavily shaken along the 60-km long path of the fault that
ruptured through or near several major cities in Taichung and Nantou Counties. Permanent fault
offsets of up to nine meters were measured following the earthquake in both the vertical and
horizontal directions. Prior to the earthquake, this area was thought to be one of moderate
seismic hazard, and design forces used for building structures were lower than in other areas of
Taiwan thought to have higher seismic risk.
Reinforced concrete is the most prevalent building material used in Taiwan for both
buildings and bridges. Most commercial and residential structures are made from reinforced
masonry, or reinforced concrete framing infilled with masonry or architectural concrete.
Nonetheless, numerous mid- to high-rise steel structures have been constructed in the capital,
Taipei (up to nearly 100 stories tall) as well as in other major cities. In Taichung, the largest city
in the heavily shaken region, several high-rise dual systems and welded moment frame buildings
have been constructed. These include several steel buildings in the 20 to 50-story range that were
under construction at the time of the earthquake.
As indicated in Table D-1, damage was most prevalent in reinforced concrete and masonry
buildings. Although damage to architectural features and localized yielding has been noted in
steel structures, no significant fracture-related damage has been reported to date in welded steel
moment-resisting frames.
Little information is available about the specifics of damage to steel buildings. However,
some steel structures suffered localized yielding. In a few cases, some steel frames exhibited
considerable ductility and lateral deformations that caused significant nonstructural damage,
Figure D-1. It is expected that many steel frame buildings saw severe ground shaking in
Taichung, Dali, Nantou and other large cities near the fault rupture. The only significant fracturerelated damage reported was apparently detected in a couple of high-rise steel braced frames
located in Taichung. Reports indicated brittle fractures occurred in some of the brace
connections. Precise details of this damage are not currently available.
In Taiwan the most common grade of steel used for beams is A36 (or SM400) and A572 Gr.
50 (or SM490) for columns. Recently, higher strength steel plates (A572 Gr. 50, Gr.60 and Gr.
65, SM570) are also being used for both girders and columns.

D-1

 FEMA-355E
Appendix D: Damage to Steel Buildings Due to the
September 21, 1999 Ji Ji, Taiwan Earthquake

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Multistory moment frame structures generally have welded moment connections provided at
all beam-to-column connections, in both framing directions, thereby giving the structures a high
degree of redundancy. Columns are thus frequently built-up box sections. Beams are usually
rolled sections, but built-up plate girders are used more frequently than in the U.S. In some
cases, reduced beam section configurations are employed. The reduction generally differs from
that used in the US profile (Chen, et al, 1997). Because of the high degree of redundancy, and the
smaller seismic design forces used relative to the highest seismic hazard regions in the U.S.,
member sizes for comparable height buildings are smaller than would be typical in California. A
typical steel frame building under construction is shown in Figure D-2.
Beam flanges are generally field welded to the columns utilizing complete joint penetration
welds with E7016 using the shielded metal arc welding (SMAW) process. Interestingly, backing
bars and run-off tabs are usually left in place following construction. A variety of bolted or
welded connections of the beam web to shear tab/column flange is utilized. A typical connection
detail of wide flange beams to box column is shown in Figure D-3. Shop welding is often done
with ER70S or E70XX flux core electrode.
Low-rise steel structures up to three stories in height were used throughout the shaken area
for commercial and residential construction. Welded and bolted (end plate) connections in light
framing members are used. Reportedly, such structures are not designed by engineers, but rather
by local, and often non-certified, fabricators and contractors. Some evidence of local working
and yielding in bolted connections was observed on occasion, see Figure D-4. These light
buildings generally performed well and continued to function following the earthquake, unless
they were subjected to differential settlement, pounding damage or fault rupturing.

Table D-1
Location
Nantou
County
Taichung
County

Statistics on Damage Due to September 21, 1999 Taiwan Earthquake, by
Type of Building Material (NCREE, 1999)
Reinforced
Concrete/SRC
2291/9

Masonry

Wooden

Steel/Light Steel

Other

Total

1069

67

25 / 16

954

4431

1337/4

688

43

16 / 49

658

2785

D-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Figure D-1

Figure D-2

FEMA-355E
Appendix D: Damage to Steel Buildings Due to the
September 21, 1999 Ji Ji, Taiwan Earthquake

Permanent Lateral Displacement in Small Steel Frame

Typical Taiwanese High-Rise Structure under Construction in Tiachung

D-3

 FEMA-355E
Appendix D: Damage to Steel Buildings Due to the
September 21, 1999 Ji Ji, Taiwan Earthquake

Figure D-3

Figure D-4

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Welded Connection Detail of Beams to Box Column

Bolted End Plate Connection in Light Steel Frame Building Showing
Evidence of Local Yielding

D-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
Appendix D: Damage to Steel Buildings Due to the
September 21, 1999 Ji Ji, Taiwan Earthquake

References
“http://921., 1999, ncree.gov.tw,” National Center for Research on Earthquake Engineering,
Taipei, Taiwan.
Chen, S.J., Chu, J.M. and Chou, Z.L., 1997, Dynamic Behavior of Steel Frames with Beam
Flanges Shaved Around Connection, J. Constructional Steel Research, 42 (1) 49-69.

D-5

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

REFERENCES, FEMA REPORTS, SAC REPORTS, AND ACRONYMS
References.
AISC, 1994, “Earthquake Safety,” Modern Steel Construction. (Preliminary report prepared by
the Technology and Research Department of AISC), American Institute of Steel
Construction, Chicago, Illinois.
AISC, 1997, Seismic Provisions for Structural Steel Buildings, American Institute of Steel
Construction, Inc., Chicago, Illinois
Albert C. Martin & Associates, 1973, “Union Bank Square,” in Murphy, Leonard M, editor, San
Fernando, California, Earthquake of February 9, 1971, Volume I, Part B, National Oceanic
and Atmospheric Administration, Washington, D.C.
Anderson, J. C. and Bertero, V.V., 1997, Implications of the Landers and Big Bear Earthquakes
on Earthquake Resistant Design of Structures, EERC-97/08.
Anderson, J.C., Johnston, R.G., and Partridge, J.E., 1995, Post Earthquake Studies of a Damaged
Low Rise Office Building (CE95-07), Department of Civil Engineering, University of
Southern California.
Anderson, J.C. and Linderman, R.R., 1991, Post Earthquake Repair of Welded Moment
Connections (CE 91-04), Department of Civil Engineering, University of Southern
California.
Astaneh-Asl, A. and Liu, J., 1999, “Tests on Simple Connection Including Slab Effects (SAC
Subtask 7.04),” unpublished progress report. (To be published as Liu and Astaneh-Asl,
“Cyclic Tests on Simple Connections Including Effects of the Slab,” SAC/BD-00/03.)
AWS, 1995, “Structural Welding Committee Position Statement on Northridge Earthquake
Welding Issues,” American Welding Society.
Bear, J., 1999, City of Santa Clarita Building Official. Personal correspondence.
Becker, R., Naeim, F., and Teal, E.J., 1988, “Seismic Design Practice for Steel Buildings,” Steel
Tips, Structural Steel Education Council.
Beedle, L.S., 1963, “Tests of Steel Moment Connections,” Proceedings, Structural Engineers
Association of California.
Beedle, L.S., 1965, “Reversed Loading of Steel Frames – Preliminary Tests,” Proceedings,
Structural Engineers Association of California.
Berg, G.V., Bolt, B.A., Sozen, M.A., and Rojahn, C., 1980, Earthquake in Romania, March 4,
1977, National Academy Press.
Berg, G.V., and Stratta, J. L., 1964, Anchorage and the Alaska Earthquake of March 27, 1964, ),
American Institute of Steel Construction, Chicago, Illinois.
Bertero, V.V., Anderson, J.C., Krawinkler, H., 1994, “Performance of Steel Building Structures
during the Northridge Earthquake,” EERC, UCB/EERC-94/09.
Bertero, V.V. and Miranda, E., 1989, “Performance of Low-Rise Buildings in Mexico City,”
Earthquake Spectra.
Bertero, V.V., and Popov, E.P., 1965, “Effects of Large Alternating Strains of Steel Beams,”
Journal of the Structural Division, ASCE.

R-1

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Bertero, V.V., Popov, E.P., and Krawinkler, H., 1972, “Beam-Column Subassemblages Under
Repeated Loading,” Journal of the Structural Division, ASCE.
Blodgett, O.W., 1963, Design of Weldments, The James F. Lincoln Arc Welding Foundation.
Bonowitz, D, 1998, “The SAC Survey: Understanding and Using Steel Frame Damage Data,” in
Proceedings of the NEHRP Conference and Workshop On Research on the Northridge,
California Earthquake of January 17, 1994, Volume III-B, California Universities for
Research in Earthquake Engineering.
Bonowitz, D., 1999(a), “SAC Database of WSMF Connection Tests.” Contractor’s report
submitted to the SAC Joint Venture.
Bonowitz, D., 1999(b), “The SAC Database: Understanding and Using Steel Frame Test
Results,” in Proceedings of the 1999 Structures Congress, ASCE, New Orleans.
Bonowitz, D. and Youssef, N., 1995, “SAC Survey of Steel Moment-Resisting Frame Buildings
Affected by the 1994 Northridge Earthquake,” in SAC 95-06.
Bouwkamp, J.G., and Clough, R.W., 1965, “Dynamic Properties of a Steel Frame Building,”
SEAOC Proceedings.
Building Standards, 1994, “ICBO Board Approves Emergency Structural Design Provision,”
p. 26.
California Institute of Steel Construction, Southern California Division, Why Take a Chance?
(collection of news articles), California Institute of Steel Construction.
Cary, H.B., 1970, “Flux-cored Arc Welding – Advances and Applications in the U.S.A.,”
Welding and Metal Fabrication.
Carydis, P. G., Tilford, N., Brandow, G., and Jirsa, J., 1982, The Central Greece Earthquakes of
February-March 1981, National Academy Press.
Castro, J.B., 1997, “Cosmetically Repaired Welded Steel Moment Frames: A Clear Risk of
Disaster,” in California Construction Law Reporter, pp 52-62 and 67-68.
City of Burbank, Municipal Code, 1999, “Sec. 7-140. Earthquake Damage Repair of Welded
Steel Moment Frame Buildings.”
City of Los Angeles Building Bureau, 1994a, “Welding Procedures for Moment Ductile
Frames—Both New and Repair Work,” Interdepartmental Correspondence.
City of Los Angeles Building Bureau, 1994b, “Repair of Cracked Moment Connections in Steel
Frame Structures and Requirements for Connections in New Buildings,” Interdepartmental
Correspondence.
City of Los Angeles Building Bureau, 1994c, “In New Construction the Use of Welded
Connections in Buildings with Special Moment-Resisting Steel Frames (SMRF),”
Interdepartmental Correspondence.
City of Los Angeles Building Bureau, 1994d, “Repair of Cracked Moment Frame Connections in
Welded Steel Frame Structures and Requirements for SMRF-Welded Connections in New
Buildings,” Interdepartmental Correspondence.
City of Santa Monica, Municipal Code, “Chapter 8.76 Seismic Strengthening Provisions for
Existing Welded Steel Moment Frame Structures,” undated. Copy received December 13,
1999.

R-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Coffman, J.L. and von Hake, C.A., 1973, editors. Earthquake History of the United States
(Publication 41-1, Revised Edition through 1970), Washington: U.S. Department of
Commerce.
Collin, A.L., 1992, “Letter to the Research Committee,” Research Bulletin Board, SEAONC
Research Committee.
Cornell, C.A. and Luco, N., 1999, “The Effect of Connection Fractures on Steel Moment
Resisting Frame Seismic Demands and Safety: A Report on SAC Phase II Task 5.4.6,” Final
Draft (To be published as SAC/BD-99/03.)
Couch, M.F. and Olsson, D.A., 1965, “Weld Quality and Sonic Inspection,” in Proceedings: 34th
Annual Convention, Structural Engineers Association of California.
County of Los Angeles, 1994, “Emergency Regulations for Welded Special Moment Frame
Connections and Concentrically Braced Frames (Building Code Manual No. 27.06),” County
of Los Angeles Department of Public Works.
Crosby, P., 1999, personal communication.
Dames and Moore., 1998, Survey of Damaged Steel Moment Frame Buildings. Contractor’s
report to the SAC Joint Venture, Phase 2, Task 3.1.2.
Daniels, L. H. and Collin, A. L., 1972, “Design for Material Performance,” Proceedings,
Structural Engineers Association of California.
Deierlein, G., 1995, “Summary of Building Analysis Studies,” in SAC 95-04, Part I.
Deppe, K., 1994, “Northridge Earthquake Overall Damage Review,” in Proceedings of the 63rd
Annual Convention, Structural Engineers Association of California.
DSA, 1995, “DSA Emergency Code Change For Steel Moment Frame Girder-to-Column
Connection (IR 27-8),” California Division of the State Architect.
Durkin, M., 1995, “Inspection, Damage, and Repair of Steel Frame Buildings Following the
Northridge Earthquake,” in SAC 95-06.
Durkin, M., 1999, SAC Task 3.1 Contractor's Report, in progress.
Earthquake Spectra, May, 1985; November, 1985; February 1986; February 1988; August 1989;
October 1991; April 1995 (Supplements A and B to Volume 11).
EERI, 1970, Peru Earthquake of May 31, 1970: Preliminary Report, Earthquake Engineering
Research Institute.
EERI, 1976, “The Guatemala Earthquake of February 4, 1976 (A Preliminary Report), EERI
Newsletter, v10, n2B, Earthquake Engineering Research Institute.
EERI, 1978, Miyagi-Ken-Oki, Japan Earthquake, June 12, 1978, Peter I Yanev, editor,
Earthquake Engineering Research Institute.
EERI, 1980, Montenegro, Yugoslavia Earthquake, April 15, 1979, Arline Leeds, editor,
Earthquake Engineering Research Institute.
EERI, 1983, El-Asnam, Algeria Earthquake of October 10, 1980, Earthquake Engineering
Research Institute
EERI, 1996, Public Policy and Building Safety, Earthquake Engineering Research Institute,
Earthquake Engineering Research Institute
El-Tawil, S., Vidarsson, E., Mikesell, T., and Kunnath, S., 1999, “Inelastic Behavior and Design
of Steel Panel Zones,” Journal of Structural Engineering, pp. 183-193.
R-3

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Engelhardt, M.D. and Husain, A.S., 1993, “Cyclic Loading Performance of Welded FlangeBolted Web Connections,” Journal of Structural Engineering, ASCE.
Engelhardt, M.D., Kim, K., Sabol, T., Ho, L., Kim, H.-I., Uzarski, J., and Husain, A., 1995,
“Analysis of a Six Story Steel Moment Frame Building in Santa Monica,” in SAC 95-04,
Part 1.
Engelhardt, M.D. and Sabol, T.A., 1994, “Testing of Welded Steel Moment Connections in
Response to the Northridge Earthquake,” in Northridge Steel Update I, American Institute of
Steel Construction.
ENR, 1997a, “Suit Fuels Debate Over West’s Once Popular Weld Metal.”
ENR, 1997b, “Unusual Bonds to Fund Quake Reconstruction.”
ENR, 1997c, “Researcher Says No One Heeded His Warnings About Connection.”
ENR, 1997d, “Lincoln Settles First Lawsuit.”
ENR, 1998, “Parties Settle in Lawsuit Over Cracking of Welded Connections.”
EQE, 1990, The July 16, 1990 Philippines Earthquake: A Quick Look Report, EQE.
FEMA-267, 1995, Interim Guidelines: Evaluation, Repair, Modification, and Design of Welded
Steel Frame Moment Frame Structures, Federal Emergency Management Agency,
Washington, DC.
FEMA-267A, 1997, Interim Guidelines Advisory No. 1 (Supplement to FEMA 267), Federal
Emergency Management Agency, Washington, DC.
FEMA-273, 1997, NEHRP Guidelines for the Seismic Rehabilitation of Buildings, Federal
Emergency Management Agency, Washington, DC.
FEMA-274, 1997, NEHRP Commentary on the Guidelines for the Seismic Rehabilitation of
Buildings, Federal Emergency Management Agency, Washington, DC.
FEMA-288, 1997, Background Reports: Metallurgy, Fracture Mechanics, Welding, Moment
Connections and Frame Systems Behavior, Federal Emergency Management Agency,
Washington, DC.
FEMA-302 and FEMA-303, 1998, NEHRP Recommended Provisions for Seismic Regulations
for New Buildings and Other Structures (Part 1–Provisions) and (Part 2–Commentary),
Federal Emergency Management Agency, Washington, DC.
Ferch, R., 2000, unpublished correspondence with SAC.
Forell, N.F. and Nicoletti, J.P., 1980, Mexico Earthquakes, Earthquake Engineering Research
Institute.
Foutch, D.A., Housner, G.W., and Jennings, P.C. 1975, Dynamic Responses of Six Multistory
Buildings During the San Fernando Earthquake (Report No. EERL 75-02). California
Institute of Technology, Earthquake Engineering Research Laboratory.
Freeman, J.R., 1932, Earthquake Damage and Earthquake Insurance, McGraw-Hill Book
Company, New York.
Garreau, J., 1991, Edge City, Doubleday, New York.
Gates, W.E., 1973a, “Kajima International Building,” in Murphy, Leonard M, editor, San
Fernando, California, Earthquake of February 9, 1971, Volume I, Part B, National Oceanic
and Atmospheric Administration, Washington, D.C.

R-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Gates, W.E., 1973b, “KB Valley Center,” in Murphy, L.M, editor, San Fernando, California,
Earthquake of February 9, 1971, Volume I, Part B, National Oceanic and Atmospheric
Administration, Washington, D.C.
Gates, W.E. and Morden, M., 1995, “Lessons From Inspection, Evaluation, Repair and
Construction of Welded Steel Moment Frames Following the Northridge Earthquake,” in
SAC 95-06.
Goel, S.C., 1968, “Response of Multistory Steel Frames to Earthquake Forces,” in Steel
Research for Construction (Bulletin No. 12), American Iron and Steel Institute.
Goltz, J.D. and Weinberg, P., 1998, “The Impact of Earthquakes on Welded Steel Moment
Frame Buildings: Experience in Past Earthquakes,” draft report for the SAC Steel Project.
(To be published as Weinberg and Goltz, SAC/BD-99/11.)
Graham, J.D., Sherbourne, A. N., and Khabbaz, R. N., 1959, Welded Interior Beam-to-Column
Connections, American Institute of Steel Construction, Inc., New York.
H. J. Degenkolb Associates, Engineers, The Whittier Narrows Earthquake, October 1, 1987, H.
J. Degenkolb Associates, Engineers, not dated.
Habibullah, A., 1984, “Personal Computers and the Structural Engineer,” Proceedings: 1984
Convention, Structural Engineers Association of California.
Hadley, J. M. How Structures Withstood the Japanese Earthquake and Fire, Portland Cement
Association.
Hanson, R.D. and Degenkolb, H.J., 1969, The Venezuela Earthquake, July 29, 1967, American
Iron and Steel Institute.
Hart, G.C., 1973, “1901 Avenue of the Stars Building,” in Murphy, Leonard M, editor, San
Fernando, California, Earthquake of February 9, 1971, Volume I, Part B, National Oceanic
and Atmospheric Administration, Washington, D.C.
Hart, G.C., Huang, S., Lobo, R.F., Van Winkle, M., Jain, A., 1995, “Earthquake Response of
Strengthened Steel Special Moment Resisting Frames,” in SAC 95-04, Part 1.
Holmes, W.T., 1998, “The History of U.S. Seismic Code Development,” ’98 EERI Annual
Meeting, Earthquake Engineering Research Institute.
ICBO, 1949 through 1997, Uniform Building Code.
ICBO, 1999, 1999 City of Los Angeles Building Code, International Conference of Building
Officials and California Building Standards Commission.
Islam, M.S., Youssef, N.F.G., and Lam, S.T., 1995, Analysis of Two Steel Moment-Resisting
Frame Buildings Damaged During the Northridge Earthquake. Draft report to NIST.
J. H. Wiggins Company, 1971, “Arco Plaza Earthquake Risk Evaluation,” Unpublished report
prepared for Arthur Young and Company.
Jacobsen, L.S. and Ayre, R.S., 1938, “Experimentally Determined Dynamic Shears in a SixteenStory Model,” Bulletin of the Seismological Society of America, V28.4.
The James F. Lincoln Arc Welding Foundation, 1997, Fabricators’ and Erectors’ Guide to
Welded Steel Construction, the James F. Lincoln Arc Welding Foundation.
Jenks, K.B. and Ritts, G.J., 1997, “Welded Steel Moment Frames: In an Earthquake, We’d
Rather Be in a Steel-Framed Building,” in California Construction Law Reporter, pp 62-66.

R-5

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Jennings, P.C., 1971, editor. Engineering Features of the San Fernando Earthquake of February
9, 1971 (EERL 71-02), Earthquake Engineering Research Laboratory, California Institute of
Technology, Pasadena.
John A. Blume & Associates, 1973, “Bunker Hill Tower,” in Murphy, L.M., editor, San
Fernando, California, Earthquake of February 9, 1971, Volume I, Part A, National Oceanic
and Atmospheric Administration Washington, D.C.
Johnson, M.Q., “Evaluation of the Effect of Welding Procedure on the Mechanical Properties of
FCAW-S and SMAW Weld Metal Used in the Construction of Seismic Moment Frames,” inprogress research to be published by SAC as SAC/BD-00/12.
Kariotis, J. and Eimani, T.J., 1995, “Analysis of a 16-Story Steel Frame Building at Site 5 for the
Northridge Earthquake,” in SAC 95-04, Part 2.
Kaufmann, E.J., Fisher, J.W., Di Julio, R.M. Jr., and Gross, J.L., 1997, Failure Analysis of
Welded Steel Moment Frames Damaged in the Northridge Earthquake (NISTIR 5944),
National Institute of Science and Technology.
Kellam, H.S. “Recommended Lateral Force Requirements By SEAOC, 1966, Part I—
Background of Paragraph (J),” 35th Annual Convention Proceedings, Structural Engineers
Association of California.
Krawinkler, H., 1985, “Tentative Lateral Force Requirements Background to Selected Steel
Provisions,” Proceedings: 54th Annual Convention, Structural Engineers Association of
California.
Krawinkler, H., 1997, “Systems Behavior of Structural Steel Frames Subjected to Earthquake
Ground Motion,” in FEMA-288.
Krawinkler, H., 2000, State of the Art Report: Systems Performance of Moment-Resisting Steel
Frames Subject to Earthquake Ground Shaking (FEMA-355C).
Krawinkler, H., Alali, A., Thiel, C.C., Dunlea, J.M., 1995, “Analysis of a Damaged Four-Story
Building and an Undamaged Two-Story Building,” in SAC 95-04, Part 1.
LACOTAP (Los Angeles County Technical Advisory Panel), 1997, SMRF Bulletin No. 3, Steel
Moment Resisting Frame (SMRF) Connection Systems.
Layne, H.M. (moderator), et al., 1963, “Panel Discussion: Section J of the SEA Code—Should It
Be Retained, Modified, or Eliminated?” Proceedings, Structural Engineers Association of
California.
Lee, K.-H., Stojadinovic, B., Goel, S., Margarian, A., Choi, J., Wongkaew, A., Reyher, B., and
Lee, D.-Y., “Parametric Tests on Unreinforced Connections.” Contractors’ report to the SAC
Joint Venture, to be published as SAC/BD-00/01.
Los Angeles Department of Building & Safety, “Status of Buildings Ordered to Comply with
Ord. 170406 as of 5/4/98,” unpublished.
Mahin, S.A., Malley, J.O., Hamburger, R.O., 1996, “Phase 2 of the SAC Steel Project,”
Proceedings: 65th Annual Convention, Structural Engineers Association of California.
Maison, B.F. and Bonowitz, D., 1999, “How Safe Are Pre-Northridge WSMFs? A Case Study of
the SAC Los Angeles Nine-Story Building,” in Earthquake Spectra 15.
Maison, B.F. and Bonowitz, D., 2000, “Simplified Loss Estimation For Pre-Northridge WSMF
Buildings,” in-progress report to SAC.
Maison, B.F. and Kasai, K., 1997, “Analysis of Northridge Damaged Thirteen-Story WSMF
Building,” in Earthquake Spectra 13.
R-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

McDonald, G., 1999, City of Simi Valley Building Official. Personal correspondence.
McGuire, W., 1988, “Introduction,” in Journal of Constructional Steel Research, volume 10
(special issue on steel beam-to-column building connections), Elsevier Applied Science.
Melnick, S., 1991, “Criticizing Steel,” Modern Steel Construction.
Mendizabal, M., 2000, City of Santa Monica Building Official. Personal correspondence.
Mendoza, B., 1999, City of San Fernando Building Official. Personal correspondence.
Modern Steel Construction, 1994, “Localized Steel Damage,” Modern Steel Construction.
Moon, S., 1999, City of Beverly Hills Building Official. Personal correspondence.
Naeim, F., DiJulio, R., Benuska, K., Reinhorn, A.M., and Chen, L., 1995, “Evaluation of
Seismic Performance of an 11-Story Steel Moment Frame Building During the 1994
Northridge Earthquake,” in SAC 95-04, Part 2.
Naeim, F., Lobo, R., Skliros, K., and Sgambelluri, M., 1999, Seismic Performance of Four
Instrumented Moment Resisting Steel Structures During the January 17, 1994 Northridge,
California Earthquake. Final Draft report to California Division of Mines and Geology,
SMIP.
Osteraas, J., and Krawinkler, H., 1989, “The Mexico Earthquake of September 19, 1985:
Behavior of Steel Buildings,” Earthquake Spectra, v 5.1, Earthquake Engineering Research
Institute.
Paret, T.F., 1999, “Studies Directed Towards Explaining the W1 Issue; Clarifying the Extent of
Northridge Induced Weld Fracturing; and Examining the Related Issue of UT Reliability,”
Contractor’s report to the SAC Joint Venture. (To be published as SAC/BD-99/10.)
Paret, T.F., Nine Studies Aimed at Explaining the W1 Issue and the Related Issue of UT
Inspection Reliability, Contractor’s report to the SAC Joint Venture, undated.
Paret, T. and Sasaki, K.K., 1995, “Analysis of a 17-Story Steel Frame Building Damaged by the
Northridge Earthquake,” in SAC 95-04, Part 2.
Phipps, M.T., 1998, “Task 3.1, Objective 5: Investigation of Damage to WSMF by Earthquakes
other than Northridge,” Contractor’s report to the SAC Joint Venture, (to be published as
SAC/BD-99/09.)
Popov, E.P., 1987, “Seismic Moment Connections for MRF’s,” Journal of Constructional Steel
Research (U.K.). Republished in Journal of Constructional Steel Research, volume 10
(special issue on steel beam-to-column building connections), Elsevier Applied Science,
1988.
Popov, E.P., Amin, N.R., Louie, J.J.C., and Stephen, R.M., 1985, “Cyclic Behavior of Large
Beam-Column Assemblies,” Earthquake Spectra, Earthquake Engineering Research
Institute.
Popov, E.P. and Bertero, V.V., 1970, “Research on Seismic Behavior of High-Rise MomentResisting Steel Frames,” Proceedings, Structural Engineers Association of California.
Popov, E.P. and Bertero, V.V., 1973, “Cyclic Loading of Steel Beams and Connections,”
Journal of the Structural Division, ASCE.
Popov, E.P. and Franklin, H.A., 1965, “Steel Beam-to-Column Connections,” SEAOC
Proceedings.
Popov, E.P. and Pinkney, A.M., 1969, “Cyclic Yield Reversal in Steel Building Connections,”
Journal of the Structural Division, ASCE.
R-7

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Popov, E.P. and Stephen, R.M., 1972, “Cyclic Loading of Full-Size Steel Connections,” Steel
Research for Construction (Bulletin No. 21), AISI.
Popov, E.P. and Tsai, K.C., 1987, “Performance of Large Seismic Steel Moment Connections
Under Cyclic Loads,” SEAONC Proceedings, 56th Annual Convention.
Popov, E.P., Tsai, K.C., and Engelhardt, M.D., 1988, “On Seismic Steel Joints and
Connections,” in Proceedings: 57th Annual Convention, Structural Engineers Association of
California.
Popov, E.P., Yang, T.-S., and Grigorian, C.E., 1993, “New Directions in Structural Seismic
Designs,” Earthquake Spectra.
Preece, F.R., 1981, Structural Steel in the 80s—Materials, Fastening and Testing. The Steel
Committee of California.
Preece, F.R. and Collin, A.L., 1991, Structural Steel Construction in the ‘90s. Structural Steel
Education Council.
Procedure Handbook of Arc Welding, The, 1973, Twelfth Edition, The Lincoln Electric
Company, Cleveland, Ohio.
Putkey, J.J., 1993, “Common Steel Erection Problems and Suggested Solutions,” in Steel Tips,
Structural Steel Education Council.
Reynolds, R.E., editor. Landers: Earthquakes and Aftershocks, Winter 1993 issue of San
Bernardino County Museum Association Quarterly.
Roeder, C.W. and Foutch, D.A., 1996, “Experimental Results for Seismic Resistant Steel
Moment Frame Connections,” Journal of Structural Engineering.
SAC, 1998, “Observations of the Behavior of Steel Frame Structures in Historic Earthquakes,”
draft report to the SAC Joint Venture Partnership.
SAC 95-01, 1995, Advisory No. 3, SAC Joint Venture Partnership.
SAC 95-04, 1995, Analytical and Field Investigations of Buildings Affected by the Northridge
Earthquake of January 17, 1994, SAC Joint Venture Partnership.
SAC 95-06, 1995, Surveys and Assessment of Damage to Buildings Affected by the Northridge
Earthquake of January 17, 1994, SAC Joint Venture Partnership.
SAC 95-07, 1995, Case Studies of Steel Moment Frame Building Performance in the Northridge
Earthquake of January 17, 1994, SAC Joint Venture Partnership.
SAC Steel Project, 1996, “Notice to Building Officials and Design Professionals in the San
Francisco Bay Area.”
Schneider, S.P., Roeder, C.W., 1993, Carpenter, James E., “Seismic Behavior of Moment
Resisting Steel Frames: Experimental Study,” Journal of Structural Engineering, ASCE.
SEAOC, 1959; 1968 (with revisions); 1974 (with 1975 and 1980 revisions); 1988; 1990; 1996,
Recommended Lateral Force Requirements and Commentary (Blue Book), Structural
Engineers Association of California.
SEAOC Plan Review, 1996, “SEAOC Represented in Legislative Process.”
SEAONC, 1998, SMRF Connection Qualification Commentary, Structural Engineers
Association of Northern California Seismology Committee.

R-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Seligson, H. and Eguchi, R., 1999, “Normalizing Database of SMRF Structures within Los
Angeles County for Use in SAC Damage Model Development Projects,” EQE Project
Number 250825.01.
Sloan, T., 2000, City of Burbank Plan Check Engineer. Personal correspondence.
Smith, D., 1995, “Inspections, Repairs Lag for Steel-Frame Buildings,” Los Angeles Times.
Song, J. and Ellingwood, B.R., 1999, “Seismic Reliability of Special Moment Steel Frames With
Welded Connections: I,” Journal of Structural Engineering.
SSDA, 1996, “Slotted Beam-to-Column Connection Designs,” Seismic Structural Design
Associates, Inc.
Steel Committee of California, 1979 (revised); 1983; 1986, Steel Connections/Details and
Relative Costs.
Steinbrugge, K.V. and Moran, D.F., 1954, “An Engineering Study of the Southern California
Earthquake of July 21, 1952, and Its Aftershocks,” Bulletin of the Seismological Society of
America.
Steinbrugge, K.V., Schader, E.E., Bigglestone, H.C., Weers, C.A., 1971, “San Fernando
Earthquake, February 9, 1971,” Pacific Fire Rating Bureau.
Strand, D.R., 1984, “Code Development Between 1927 and 1980,” Proceedings, Structural
Engineers Association of California.
Stratta, J.L., Escalante, L.E., Krinitzsky, Ellis L., and Morelli, U., 1981, Earthquake in
Campania-Basilicata, Italy, November 23, 1980, National Academy Press.
Stratta, J.L. and Wyllie, L.A., 1979, Friuli, Italy Earthquakes of 1976, Earthquake Engineering
Research Institute.
Thornton, W.A., 1992, “Designing for Cost Efficient Fabrication,” in Steel Tips, Structural Steel
Education Council.
Tide, R., 2000, unpublished correspondence with SAC, January 28, 2000, and personal
correspondence.
Tierney, K.J., 1985, Report on the Coalinga Earthquake of May 2, 1983. Seismic Safety
Commission.
Tobriner, S., 1984, “The History of Building Codes to the 1920’s,” Proceedings, Structural
Engineers Association of California.
Tom, S., 2000, City of Glendale Building Official. Personal correspondence.
Tsai, K.C. and Popov, E.P., 1988, Steel Beam-Column Joints in Seismic Moment Resisting
Frames (UCB/EERC-88/19), Earthquake Engineering Research Center, Richmond,
California.
Uang, C.M., Yu, Q.S., Sadre, A., Bonowitz, D., Youssef, N., 1995, Performance of a 13-Story
Steel Moment-Resisting Frame Building Damaged in the 1994 Northridge Earthquake,
SSRP-95/04, University of California, San Diego.
United States Geological Survey (USGS), 1907, The San Francisco Earthquake and Fire of
April 18, 1906, Washington: Government Printing Office.
United States Steel Corporation, circa 1980, Beam-to-Column Flange Connections—Restrained
Members: A Design Aid, United States Steel Corporation.

R-9

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

Wilson, E.L., 1984, “Selection of Microcomputer Hardware for Structural Analysis and Design,”
Proceedings: 1984 Convention, Structural Engineers Association of California.
Wood, F.J., editor, 1967, The Prince William Sound, Alaska, Earthquake of 1964 and
Aftershocks (Coast and Geodetic Survey Publication 10-3), Washington D.C.: United States
Government Printing Office.
Yanev, P.I., Gillengerten, J.D., and Hamburger, R.O., 1991, The Performance of Steel Buildings
in Past Earthquakes, The American Iron and Steel Institute.
Youssef, N., Bonowitz, D., and Gross, J., 1995, A Survey of Steel Moment-Resisting Frame
Buildings Affected by the 1994 Northridge Earthquake (NISTIR 5625), National Institute of
Science and Technology.
Zeevaert, L., 1962, “The Development of Shear Displacement Meters and Accelerometers to
Measure Earthquake Forces in Buildings,” in Proceedings: 31st Annual Convention,
Structural Engineers Association of California.
Zhang, J. and Dong, P., 2000, “Residual Stresses in Welded Moment Frames and Implications
for Structural Performance,” Journal of Structural Engineering.
Zsutty, T.C., 1989, “The 1989 Blue Book Commentary,” in Proceedings: 58th Annual
Convention, Structural Engineers Association of California.
FEMA Reports.
FEMA reports are listed by report number.
FEMA-178, 1992, NEHRP Handbook for the Seismic Evaluation of Existing Buildings,
developed by the Building Seismic Safety Council for the Federal Emergency Management
Agency, Washington, DC.
FEMA-267, 1995, Interim Guidelines, Inspection, Evaluation, Repair, Upgrade and Design of
Welded Moment Resisting Steel Structures, prepared by the SAC Joint Venture for the
Federal Emergency Management Agency, Washington, DC. Superseded by FEMA 350
to 353.
FEMA-267A, 1996, Interim Guidelines Advisory No. 1, prepared by the SAC Joint Venture for
the Federal Emergency Management Agency, Washington, DC. Superseded by FEMA 350
to 353.
FEMA-267B, 1999, Interim Guidelines Advisory No. 2, prepared by the SAC Joint Venture for
the Federal Emergency Management Agency, Washington, DC. Superseded by FEMA 350
to 353.
FEMA-273, 1997, NEHRP Guidelines for the Seismic Rehabilitation of Buildings, prepared by
the Applied Technology Council for the Building Seismic Safety Council, published by the
Federal Emergency Management Agency, Washington, DC.
FEMA-274, 1997, NEHRP Commentary on the Guidelines for the Seismic Rehabilitation of
Buildings, prepared by the Applied Technology Council for the Building Seismic Safety
Council, published by the Federal Emergency Management Agency, Washington, DC.
FEMA-302, 1997, NEHRP Recommended Provisions for Seismic Regulations for New Buildings
and Other Structures, Part 1 – Provisions, prepared by the Building Seismic Safety Council
for the Federal Emergency Management Agency, Washington, DC.

R-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

FEMA-303, 1997, NEHRP Recommended Provisions for Seismic Regulations for New Buildings
and Other Structures, Part 2 – Commentary, prepared by the Building Seismic Safety
Council for the Federal Emergency Management Agency, Washington, DC.
FEMA-310, 1998, Handbook for the Seismic Evaluation of Buildings – A Prestandard, prepared
by the American Society of Civil Engineers for the Federal Emergency Management
Agency, Washington, DC.
FEMA-350, 2000, Recommended Seismic Design Criteria for New Steel Moment-Frame
Buildings, prepared by the SAC Joint Venture for the Federal Emergency Management
Agency, Washington, DC.
FEMA-351, 2000, Recommended Seismic Evaluation and Upgrade Criteria for Existing Welded
Steel Moment-Frame Buildings, prepared by the SAC Joint Venture for the Federal
Emergency Management Agency, Washington, DC.
FEMA-352, 2000, Recommended Postearthquake Evaluation and Repair Criteria for Welded
Steel Moment-Frame Buildings, prepared by the SAC Joint Venture for the Federal
Emergency Management Agency, Washington, DC.
FEMA-353, 2000, Recommended Specifications and Quality Assurance Guidelines for Steel
Moment-Frame Construction for Seismic Applications, prepared by the SAC Joint Venture
for the Federal Emergency Management Agency, Washington, DC.
FEMA-354, 2000, A Policy Guide to Steel Moment-Frame Construction, prepared by the SAC
Joint Venture for the Federal Emergency Management Agency, Washington, DC.
FEMA-355A, 2000, State of the Art Report on Base Metals and Fracture, prepared by the SAC
Joint Venture for the Federal Emergency Management Agency, Washington, DC.
FEMA-355B, 2000, State of the Art Report on Welding and Inspection, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Washington, DC.
FEMA-355C, 2000, State of the Art Report on Systems Performance of Steel Moment Frames
Subject to Earthquake Ground Shaking, prepared by the SAC Joint Venture for the Federal
Emergency Management Agency, Washington, DC.
FEMA-355D, 2000, State of the Art Report on Connection Performance, prepared by the SAC
Joint Venture for the Federal Emergency Management Agency, Washington, DC.
FEMA-355E, 2000, State of the Art Report on Past Performance of Steel Moment-Frame
Buildings in Earthquakes, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Washington, DC.
FEMA-355F, 2000, State of the Art Report on Performance Prediction and Evaluation of Steel
Moment-Frame Buildings, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Washington, DC.
SAC Joint Venture Reports.
SAC Joint Venture reports are listed by report number, except for SAC 2000a through 2000k;
those entries that do not include a FEMA report number are published by the SAC Joint
Venture.
SAC 94-01, 1994, Proceedings of the Invitational Workshop on Steel Seismic Issues, Los
Angeles, September 1994, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Washington, DC.

R-11

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

SAC 94-01, 1994b, Proceedings of the International Workshop on Steel Moment Frames,
Sacramento, December, 1994, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Washington, DC.
SAC 95-01, 1995, Steel Moment Frame Connection Advisory No. 3, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Washington, DC.
SAC 95-02, 1995, Interim Guidelines: Evaluation, Repair, Modification and Design of Welded
Steel Moment Frame Structures, prepared by the SAC Joint Venture for the Federal
Emergency Management Agency, Report No. FEMA-267, Washington, DC.
SAC 95-03, 1995, Characterization of Ground Motions During the Northridge Earthquake of
January 17, 1994, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Washington, DC.
SAC 95-04, 1995, Analytical and Field Investigations of Buildings Affected by the Northridge
Earthquake of January 17, 1994, prepared by the SAC Joint Venture for the Federal
Emergency Management Agency, Washington, DC.
SAC 95-05, 1995, Parametric Analytic Investigations of Ground Motion and Structural
Response, Northridge Earthquake of January 17, 1994, prepared by the SAC Joint Venture
for the Federal Emergency Management Agency, Washington, DC.
SAC 95-06, 1995, Technical Report: Surveys and Assessment of Damage to Buildings Affected
by the Northridge Earthquake of January 17, 1994, prepared by the SAC Joint Venture for
the Federal Emergency Management Agency, Washington, DC.
SAC 95-07, 1995, Technical Report: Case Studies of Steel Moment-Frame Building
Performance in the Northridge Earthquake of January 17, 1994, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Washington, DC.
SAC 95-08, 1995, Experimental Investigations of Materials, Weldments and Nondestructive
Examination Techniques, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Washington, DC.
SAC 95-09, 1995, Background Reports: Metallurgy, Fracture Mechanics, Welding, Moment
Connections and Frame Systems Behavior, prepared by the SAC Joint Venture for the
Federal Emergency Management Agency, Report No. FEMA-288, Washington, DC.
SAC 96-01, 1996, Experimental Investigations of Beam-Column Subassemblages, Part 1 and 2,
prepared by the SAC Joint Venture for the Federal Emergency Management Agency,
Washington, DC.
SAC 96-02, 1996, Connection Test Summaries, prepared by the SAC Joint Venture for the
Federal Emergency Management Agency, Report No. FEMA-289, Washington, DC.
SAC 96-03, 1997, Interim Guidelines Advisory No. 1 Supplement to FEMA-267 Interim
Guidelines, prepared by the SAC Joint Venture for the Federal Emergency Management
Agency, Report No. FEMA-267A, Washington, DC.
SAC 98-PG, Update on the Seismic Safety of Steel Buildings – A Guide for Policy Makers,
prepared by the SAC Joint Venture for the Federal Emergency Management Agency,
Washington, DC.
SAC 99-01, 1999, Interim Guidelines Advisory No. 2 Supplement to FEMA-267 Interim
Guidelines, prepared by the SAC Joint Venture, for the Federal Emergency Management
Agency, Report No. FEMA-267B, Washington, DC.

R-12

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

SAC, 2000a, Recommended Seismic Design Criteria for New Steel Moment-Frame Buildings,
prepared by the SAC Joint Venture for the Federal Emergency Management Agency, Report
No. FEMA-350, Washington, DC.
SAC, 2000b, Recommended Seismic Evaluation and Upgrade Criteria for Existing Welded Steel
Moment-Frame Buildings, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Report No. FEMA-351, Washington, DC.
SAC, 2000c, Recommended Postearthquake Evaluation and Repair Criteria for Welded Steel
Moment-Frame Buildings, prepared by the SAC Joint Venture for the Federal Emergency
Management Agency, Report No. FEMA-352, Washington, DC.
SAC, 2000d, Recommended Specifications and Quality Assurance Guidelines for Steel MomentFrame Construction for Seismic Applications, prepared by the SAC Joint Venture for the
Federal Emergency Management Agency, Report No. FEMA-353, Washington, DC.
SAC, 2000e, A Policy Guide to Steel Moment-Frame Construction, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Report No. FEMA-354,
Washington, DC.
SAC, 2000f, State of the Art Report on Base Metals and Fracture, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Report No. FEMA-355A,
Washington, DC.
SAC, 2000g, State of the Art Report on Welding and Inspection, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Report No. FEMA-355B,
Washington, DC.
SAC, 2000h, State of the Art Report on Systems Performance, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Report No. FEMA-355C,
Washington, DC.
SAC, 2000i, State of the Art Report on Connection Performance, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Report No. FEMA-355D,
Washington, DC.
SAC, 2000j, State of the Art Report on Past Performance of Steel Moment-Frame Buildings in
Earthquakes, prepared by the SAC Joint Venture for the Federal Emergency Management
Agency, Report No. FEMA-355E, Washington, DC.
SAC, 2000k, State of the Art Report on Performance Prediction and Evaluation, prepared by the
SAC Joint Venture for the Federal Emergency Management Agency, Report No. FEMA355F, Washington, DC.
SAC/BD-96/01, Selected Results from the SAC Phase 1 Beam-Column Connection Pre-Test
Analyses, submissions from B. Maison, K. Kasai, and R. Dexter; and A. Ingraffea and G.
Deierlein.
SAC/BD-96/02, Summary Report on SAC Phase 1 - Task 7 Experimental Studies, by C. Roeder
(a revised version of this document is published in Report No. SAC 96-01; the original is no
longer available).
SAC/BD-96/03, Selected Documents from the U.S.-Japan Workshop on Steel Fracture Issues.
SAC/BD-96/04, Survey of Computer Programs for the Nonlinear Analysis of Steel Moment
Frame Structures.
SAC/BD-97/01, Through-Thickness Properties of Structural Steels, by J. Barsom and S.
Korvink.

R-13

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

SAC/BD-97/02, Protocol for Fabrication, Inspection, Testing, and Documentation of BeamColumn Connection Tests and Other Experimental Specimens, by P. Clark, K. Frank, H.
Krawinkler, and R. Shaw.
SAC/BD-97/03, Proposed Statistical and Reliability Framework for Comparing and Evaluating
Predictive Models for Evaluation and Design, by Y.-K. Wen.
SAC/BD-97/04, Development of Ground Motion Time Histories for Phase 2 of the FEMA/SAC
Steel Project, by. P. Somerville, N. Smith, S. Punyamurthula, and J. Sun.
SAC/BD-97/05, Finite Element Fracture Mechanics Investigation of Welded Beam-Column
Connections, by W.-M. Chi, G. Deierlein, and A. Ingraffea.
SAC/BD-98/01, Strength and Ductility of FR Welded-Bolted Connections, by S. El-Tawil, T.
Mikesell, E. Vidarsson, and S. K. Kunnath.
SAC/BD-98/02, Effects of Strain Hardening and Strain Aging on the K-Region of Structural
Shapes, by J. Barsom and S. Korvink
SAC/BD-98/03, Implementation Issues for Improved Seismic Design Criteria: Report on the
Social, Economic, Policy and Political Issues Workshop by L. T. Tobin.
SAC/BD-99/01, Parametric Study on the Effect of Ground Motion Intensity and Dynamic
Characteristics on Seismic Demands in Steel Moment Resisting Frames by G. A. MacRae.
SAC/BD-99/01A, Appendix to: Parametric Study on the Effect of Ground Motion Intensity and
Dynamic Characteristics on Seismic Demands in Steel Moment Resisting Frames by G. A.
MacRae.
SAC/BD-99/02, Through-Thickness Strength and Ductility of Column Flange in Moment
Connections, by R. Dexter and M. Melendrez.
SAC/BD-99/03, The Effects of Connection Fractures on Steel Moment Resisting Frame Seismic
Demands and Safety, by C. A. Cornell and N. Luco.
SAC/BD-99/04, Effects of Strength/Toughness Mismatch on Structural and Fracture Behaviors
in Weldments, by P. Dong, T. Kilinski, J. Zhang, and F.W. Brust.
SAC/BD-99/05, Assessment of the Reliability of Available NDE Methods for Welded Joint and
the Development of Improved UT Procedures, by G. Gruber and G. Light.
SAC/BD-99/06, Prediction of Seismic Demands for SMRFs with Ductile Connections and
Elements, by A. Gupta and H. Krawinkler.
SAC/BD-99/07, Characterization of the Material Properties of Rolled Sections, by T. K. Jaquess
and K. Frank.
SAC/BD-99/08, Study of the Material Properties of the Web-Flange Intersection of Rolled
Shapes, by K. R. Miller and K. Frank.
SAC/BD-99/09, Investigation of Damage to WSMF Earthquakes other than Northridge, by M.
Phipps.
SAC/BD-99/10, Clarifying the Extent of Northridge Induced Weld Fracturing and Examining
the Related Issue of UT Reliability, by T. Paret.
SAC/BD-99/11, The Impact of Earthquakes on Welded Steel Moment Frame Buildings:
Experience in Past Earthquakes, by P. Weinburg and J. Goltz.
SAC/BD-99/12, Assessment of the Benefits of Implementing the New Seismic Design Criteria
and Inspection Procedures, by H. A. Seligson and R. Eguchi.

R-14

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

SAC/BD-99/13, Earthquake Loss Estimation for WSMF Buildings, by C. A. Kircher.
SAC/BD-99/14, Simplified Loss Estimation for Pre-Northridge WSMF Buildings, by B. F.
Maison and D. Bonowitz.
SAC/BD-99/15, Integrative Analytical Investigations on the Fracture Behavior of Welded
Moment Resisting Connections, by G. G. Deierlein and W.-M. Chi.
SAC/BD-99/16, Seismic Performance of 3- and 9- Story Partially Restrained Moment Frame
Buildings, by B. F. Maison and K. Kasai.
SAC/BD-99/17, Effects of Partially-Restrained Connection Stiffness and Strength on Frame
Seismic Performance, by K. Kasai, B. F. Maison, and A. Mayangarum.
SAC/BD-99/18, Effects of Hysteretic Deterioration Characteristics on Seismic Response of
Moment Resisting Steel Structures, by F. Naeim, K. Skliros, A. M. Reinhorn, and M. V.
Sivaselvan.
SAC/BD-99/19, Cyclic Instability of Steel Moment Connections with Reduced Beam Section, by
C.-M. Uang and C.-C. Fan.
SAC/BD-99/20, Local and Lateral-Torsion Buckling of Wide Flange Beams, by L.
Kwasniewski, B. Stojadinovic, and S. C. Goel.
SAC/BD-99/21, Elastic Models for Predicting Building Performance, by X. Duan and J. C.
Anderson.
SAC/BD-99/22, Reliability-Based Seismic Performance Evaluation of Steel Frame Buildings
Using Nonlinear Static Analysis Methods, by G. C. Hart and M. J. Skokan.
SAC/BD-99/23, Failure Analysis of Welded Beam to Column Connections, by J. M. Barsom and
J. V. Pellegrino.
SAC/BD-99/24, Weld Acceptance Criteria for Seismically-Loaded Welded Connections, by W.
Mohr.
SAC/BD-00/01, Parametric Tests on Unreinforced Connections, Volume I – Final Report, by
K.-H. Lee, B. Stojadinovic, S. C. Goel, A. G. Margarian, J. Choi, A. Wongkaew, B. P.
Reyher, and D.-Y, Lee.
SAC/BD-00/01A, Parametric Tests on Unreinforced Connections, Volume II – Appendices, by
K.-H. Lee, B. Stojadinovic, S. C. Goel, A. G. Margarian, J. Choi, A. Wongkaew, B. P.
Reyher, and D.-Y, Lee.
SAC/BD-00/02, Parametric Tests on the Free Flange Connections, by J. Choi, B. Stojadinovic,
and S. C. Goel.
SAC/BD-00/03, Cyclic Tests on Simple Connections Including Effects of the Slab, by J. Liu and
A. Astaneh-Asl.
SAC/BD-00/04, Tests on Bolted Connections, Part I: Technical Report, by J. Swanson, R. Leon,
and J. Smallridge.
SAC/BD-00/04A, Tests on Bolted Connections, Part II: Appendices, by J. Swanson, R. Leon,
and J. Smallridge.
SAC/BD-00/05, Bolted Flange Plate Connections, by S. P. Schneider and I. Teeraparbwong.
SAC/BD-00/06, Round Robin Testing of Ultrasonic Testing Technicians, by R. E. Shaw, Jr.
SAC/BD-00/07, Dynamic Tension Tests of Simulated Welded Beam Flange Connections, by J.
M. Ricles, C. Mao, E. J. Kaufmann, L.-W. Lu, and J. W. Fisher.
R-15

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

SAC/BD-00/08, Design of Steel Moment Frame Model Buildings in Los Angeles, Seattle and
Boston, by P. Clark.
SAC/BD-00/09, Benchmarking of Analysis Programs for SMRF System Performance Studies, by
A. Gupta and H. Krawinkler.
SAC/BD-00/10, Loading Histories for Seismic Performance Testing of SMRF Components and
Assemblies, by H. Krawinkler, A. Gupta, R. Medina, and N. Luco.
SAC/BD-00/11, Development of Improved Post-Earthquake Inspection Procedures for Steel
Moment Frame Buildings, by P. Clark.
SAC/BD-00/12, Evaluation of the Effect of Welding Procedure on the Mechanical Properties of
FCAW-S and SMAW Weld Metal Used in the Construction of Seismic Moment Frames, by
M. Q. Johnson.
SAC/BD-00/13, Preliminary Evaluation of Heat Affected Zone Toughness in Structural Shapes
Used in the Construction of Seismic Moment Frames, by M. Q. Johnson and J. E. Ramirez.
SAC/BD-00/14, Evaluation of Mechanical Properties in Full-Scale Connections and
Recommended Minimum Weld Toughness for Moment Resisting Frames, by M. Q. Johnson,
W. Mohr, and J. Barsom.
SAC/BD-00/15, Simplified Design Models for Predicting the Seismic Performance of Steel
Moment Frame Connections, by C. Roeder, R. G. Coons, and M. Hoit.
SAC/BD-00/16, SAC Phase 2 Test Plan, by C. Roeder.
SAC/ BD-00/17, Behavior and Design of Radius-Cut, Reduced Beam Section Connections, by
M. Engelhardt, G. Fry, S. Jones, M. Venti, and S. Holliday.
SAC/BD-00/18, Test of a Free Flange Connection with a Composite Floor Slab, by M. Venti
and M. Engelhardt.
SAC/BD-00/19, Cyclic Testing of a Free Flange Moment Connection, by C. Gilton, B. Chi, and
C. M. Uang.
SAC/BD-00/20, Improvement of Welded Connections Using Fracture Tough Overlays, by James
Anderson, J. Duan, P. Maranian, and Y. Xiao.
SAC/BD-00/21, Cyclic Testing of Bolted Moment End-Plate Connections, by T. Murray, E.
Sumner, and T. Mays.
SAC/BD-00/22, Cyclic Response of RBS Moment Connections: Loading Sequence and Lateral
Bracing Effects, by Q. S. Yu, C. Gilton, and C. M. Uang.
SAC/BD-00/23, Cyclic Response of RBS Moment Connections: Weak Axis Configuration and
Deep Column Effects, by C. Gilton, B. Chi, and C. M. Uang.
SAC/BD-00/24, Development and Evaluation of Improved Details for Ductile Welded
Unreinforced Flange Connections, by J. M. Ricles, C. Mao, L.-W. Lu, and J. Fisher.
SAC/BD-00/25, Performance Prediction and Evaluation of Steel Special Moment Frames for
Seismic Loads, by K. Lee and D. A. Foutch.
SAC/BD-00/26, Performance Prediction and Evaluation of Low Ductility Steel Moment Frames
for Seismic Loads, by S. Yun and D. A. Foutch.
SAC/BD-00/27, Steel Moment Resisting Connections Reinforced with Cover and Flange Plates,
by T. Kim, A. S. Whittaker, V. V. Bertero, A. S. J. Gilani, and S. M. Takhirov.
SAC/BD-00/28, Failure of a Column K-Area Fracture, by J. M. Barsom and J. V. Pellegrino.
R-16

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

SAC/BD-00/29, Inspection Technology Workshop, by R. E. Shaw, Jr.
SAC/BD-00/30, Preliminary Assessment of the Impact of the Northridge Earthquake on
Construction Costs of Steel Moment Frame Buildings, by Davis Langdon Adamson.
Acronyms.
BSSC, Building Seismic Safety Council
BUEP, Bolted Unstiffened End Plate
(connection)
C, carbon
CA, California
CAC-A, air carbon arc cutting
CAWI, Certified Associate Welding
Inspector
CGHAZ, coarse-grained HAZ
CJP, complete joint penetration (weld)
CMU, concrete masonry unit, concrete
block
COD, crack opening displacement
“COV,” modified coefficient of variation, or
dispersion
CP, Collapse Prevention (performance level)
Connection Performance (team)
Cr, chromium
CSM, Capacity Spectrum Method
CTOD, crack tip opening dimension or
displacement
CTS, controlled thermal severity (test)
Cu, copper
CUREe, California Universities for
Research in Earthquake Engineering
CVN, Charpy V-notch
CWI, Certified Welding Inspector
D, displacement response, dead load
DMRSF, ductile, moment-resisting, space
frame
DNV, Det Norske Veritas
DRAIN-2DX, analysis program
DRAIN-3DX, analysis program
DRI, direct reduced iron
DST, Double Split Tee (connection)
DTI, Direct Tension Indicator
EAF, electric-arc furnace
EBT, eccentric bottom tapping
EE, electrode extension

2-D, two-dimensional
3-D, three-dimensional
A, acceleration response, amps
A2LA, American Association for
Laboratory Accreditation
ACAG, air carbon arc gouging
ACIL, American Council of Independent
Laboratories
AE, acoustic emission (testing)
AISC, American Institute for Steel
Construction
AISI, American Iron and Steel Institute
AL, aluminum
ANSI, American National Standards
Institute
API, American Petroleum Institute
ARCO, Atlantic-Richfield Company
As, arsenic
ASD, allowable stress design
ASME, American Society of Mechanical
Engineers
ASNT, American Society for
Nondestructive Testing
ASTM, American Society for Testing and
Materials
ATC, Applied Technology Council
AWS, American Welding Society
B, boron
BB, Bolted Bracket (connection)
BD, background document
BF, bias factor
BFO, bottom flange only (fracture)
BFP, Bolted Flange Plates (connection)
BM, base metal
BO, Boston, Massachusetts
BOCA, Building Officials and Code
Administrators
BOF, basic oxygen furnace
BSEP, Bolted Stiffened End Plate
(connection)
R-17

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

LA, Los Angeles, California
LACOTAP, Los Angeles County Technical
Advisory Panel
LAX, Los Angeles International Airport
LB, lower bound (building)
LBZ, local brittlezone
LDP, Linear Dynamic Procedure
LEC, Lincoln Electric Company
LMF, ladle metallurgy furnace
LRFD, load and resistance-factor design
LS, Life Safety (performance level)
LSP, Linear Static Procedure
LTH, linear time history (analysis)
LU, Lehigh University
M, moment
MAP, modal analysis procedure
MAR, microalloyed rutile (consumables)
MCE, Maximum Considered Earthquake
MDOF, multidegree of freedom
MMI, Modified Mercalli Intensity
Mn, manganese
Mo, molybdenum
MRF, steel moment frame
MRS, modal response spectrum
MRSF, steel moment frame
MT, magnetic particle testing
N, nitrogen
Nb, niobium
NBC, National Building Code
NDE, nondestructive examination
NDP, Nonlinear Dynamic Procedure
NDT, nondestructive testing
NEHRP, National Earthquake Hazards
Reduction Program
NES, National Evaluation Services
NF, near-fault, near-field
Ni, nickel
NLP, nonlinear procedure
NLTH, nonlinear time history (analysis)
NS, north-south (direction)
NSP, Nonlinear Static Procedure
NTH, nonlinear time history (analysis)
NVLAP, National Volunteer Laboratory
Accreditation Program
O, oxygen
OHF, open hearth furnace

EERC, Earthquake Engineering Research
Center, UC Berkeley
EGW, electrogas welding
ELF, equivalent lateral force
EMS, electromagnetic stirring
ENR, Engineering News Record
ESW, electroslag welding
EWI, Edison Welding Institute
FATT, fracture appearance transition
temperature
fb, fusion boundary
FCAW-G, flux-cored arc welding – gasshielded
FCAW-S or FCAW-SS, flux-cored arc
welding – self-shielded
FEMA, Federal Emergency Management
Agency
FF, Free Flange (connection)
FGHAZ, fine-grained HAZ
FL, fusion line
FR, fully restrained (connection)
GBOP, gapped bead on plate (test)
gl, gage length
GMAW, gas metal arc welding
GTAW, gas tungsten arc welding
HAC, hydrogen-assisted cracking
HAZ, heat-affected zone
HBI, hot briquetted iron
HSLA, high strength, low alloy
IBC, International Building Code
ICBO, International Conference of Building
Officials
ICC, International Code Council
ICCGHAZ, intercritically reheated CGHAZ
ICHAZ, intercritical HAZ
ID, identification
IDA, Incremental Dynamic Analysis
IMF, Intermediate Moment Frame
IO, Immediate Occupancy (performance
level)
IOA, Incremental Dynamic Analysis
ISO, International Standardization
Organization
IWURF, Improved Welded Unreinforced
Flange (connection)
L, longitudinal, live load
R-18

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

SFRS, seismic-force-resisting system
Si, silicon
SMAW, shielded metal arc welding
SMF, Special Moment Frame
SMRF, special moment-resisting frame (in
1991 UBC)
SMRF, Steel Moment Frame
SMRSF, special moment-resisting space
frame (in 1988 UBC)
SN, strike-normal, fault-normal
Sn, tin
SP, Side Plate (connection)
SP, strike-parallel, fault-parallel
SP, Systems Performance (team)
SPC, Seismic Performance Category
SRSS, square root of the sum of the squares
SSPC, Steel Shape Producers Council
SSRC, Structural Stability Research Council
SUG, Seismic Use Group
SW, Slotted Web (connection)
SwRI, Southwest Research Institute
T, transverse
TBF, top and bottom flange (fracture)
Ti, titanium
TIGW, tungsten inert gas welding
TMCP, Thermo-Mechanical Processing
TN, Tennessee
TT, through-thickness
TWI, The Welding Institute
UB, upper bound (building)
UBC, Uniform Building Code
UCLA, University of California, Los
Angeles
UM, University of Michigan
URM, unreinforced masonry
US, United States of America
USC, University of Southern California
USGS, US Geological Survey
UT, ultrasonic testing
UTA, University of Texas at Austin
UTAM, Texas A & M University
V, vanadium
VI, visual inspection
w/o, without
WBH, Welded Bottom Haunch (connection)

OMF, Ordinary Moment Frame
OTM, overturning moment
P, axial load
P, axial load, phosphorus
Pb, lead
PGA, peak ground acceleration
PGV, peak ground velocity
PIDR, pseudo interstory drift ratio
PJP, partial joint penetration (weld)
PPE, Performance, Prediction, and
Evaluation (team)
PQR, Performance Qualification Record
PR, partially restrained (connection)
PR-CC, partially restrained, composite
connection
PT, liquid dye penetrant testing
PWHT, postweld heat treatment
PZ, panel zone
QA, quality assurance
QC, quality control
QCP, Quality Control Plan, Quality
Certification Program
QST, Quenching and Self-Tempering
(process)
RB, Rockwell B scale (of hardness)
RBS, Reduced Beam Section (connection)
RCSC, Research Council for Structural
Connections
RT, radiographic testing
S, sulphur, shearwave (probe)
SAC, the SAC Joint Venture; a partnership
of SEAOC, ATC, and CUREe
SAV, sum of absolute values
SAW, submerged arc welding
SBC, Standard Building Code
SBCCI, Southern Building Code Congress
International
SCCGHAZ, subcritically reheated CGHAZ
SCHAZ, subcritical HAZ
SCWB, strong column, weak beam
SCWI, Senior Certified Welding Inspector
SDC, Seismic Design Category
SDOF, single degree of freedom
SE, Seattle, Washington
SEAOC, Structural Engineers Association
of California
R-19

 FEMA-355E
References, FEMA Reports, SAC Reports, and Acronyms

WCPF, Welded Cover Plate Flange
(connection)
WCSB, weak column, strong beam
WF, wide flange
WFP, Welded Flange Plate (connection)
WFS, wire feed speed
WPQR, Welding Performance Qualification
Record
WPS, Welding Procedure Specification
WSMF, welded steel moment frame
WT, Welded Top Haunch (connection)
WTBH, Welded Top and Bottom Haunch
(connection)
WUF-B, Welded Unreinforced Flanges –
Bolted Web (connection)
WUF-W, Welded Unreinforced Flanges –
Welded Web (connection)

R-20

Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

SAC PHASE II PROJECT PARTICIPANTS
FEMA Project Officer

FEMA Technical Advisor

Michael Mahoney
Federal Emergency Management Agency
500 C St. SW, Room 404
Washington, DC 20472

Robert D. Hanson
Federal Emergency Management Agency
DFO Room 353
P.O. Box 6020
Pasadena, CA 91102-6020

Joint Venture Management Committee (JVMC)
William T. Holmes, Chair
Rutherford and Chekene
427 Thirteenth Street
Oakland, CA 94612

Christopher Rojahn
Applied Technology Council
555 Twin Dolphin Dr., Suite 550
Redwood City, CA 94065

Edwin T. Huston
Smith & Huston, Inc.
8618 Roosevelt Way NE
Seattle, WA 98115

Arthur E. Ross
Cole/Yee/Shubert & Associates
2500 Venture Oaks Way, Suite 100
Sacramento, CA 95833

Robert Reitherman
California Universities for Research in
Earthquake Engineering
1301 South 46th St.
Richmond, CA 94804

Robin Shepherd
Earthquake Damage Analysis Corporation
40585 Lakeview Drive, Suite 1B
P.O. Box 1967
Big Bear Lake, CA 92315

Project Management Committee (PMC)
Stephen A. Mahin, Project Manager
Pacific Earthquake Engr. Research Center
University of California
Berkeley, CA 94720

William T. Holmes, JVMC
Rutherford and Chekene
427 Thirteenth Street
Oakland, CA 94612

Ronald O. Hamburger, Project Director for
Project Development
EQE International
1111 Broadway, 10th Floor
Oakland, CA 94607-5500

Christopher Rojahn, JVMC
Applied Technology Council
555 Twin Dolphin Dr., Suite 550
Redwood City, CA 94065
Robin Shepherd, JVMC
Earthquake Damage Analysis Corporation
40585 Lakeview Drive, Suite 1B
P.O. Box 1967
Big Bear Lake, CA 92315

James O. Malley, Project Director for
Topical Investigations
Degenkolb Engineers
225 Bush St., Suite 1000
San Francisco, CA 94104-1737

S-1

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Peter W. Clark, Technical Assistant to PMC
SAC Steel Project Technical Office
1301 South 46th St.
Richmond, CA 94804
Project Administration
Allen Paul Goldstein, Project Administrator
Allen Paul Goldstein and Associates
1621B 13th Street
Sacramento, CA 95814

Lori Campbell, Assistant to the Project
Administrator
4804 Polo Court
Fair Oaks, CA 95628

Project Oversight Committee (POC)
James R. Harris
J.R. Harris and Co.
1580 Lincoln St., Suite 550
Denver, CO 80203-1509

William J. Hall, Chair
3105 Valley Brook Dr.
Champaign, IL 61821
Shirin Ader
International Conference of Building
Officials
5360 Workman Mill Rd.
Whittier, CA 90601-2298

Richard Holguin
520 Kathryn Ct.
Nipomo, CA 93444
Nestor Iwankiw
American Institute of Steel Construction
One East Wacker Dr., Suite 3100
Chicago, IL 60601-2001

John M. Barsom
Barsom Consulting, Ltd.
1316 Murray Ave, Suite 300
Pittsburgh, PA 15217

Roy Johnston
Brandow & Johnston Associates
1600 West 3rd St.
Los Angeles, CA 90017

Roger Ferch
Herrick Corporation
7021 Koll Center Parkway
P.O Box 9125
Pleasanton, CA 94566-9125

Leonard Joseph
Thornton-Tomassetti Engineers
641 6th Ave., 7th Floor
New York, NY 10011

Theodore V. Galambos
University of Minnesota
122 CE Building, 500 Pillsbury Dr. SE
Minnneapolis, MN 55455

Duane K. Miller
The Lincoln Electric Company
22801 St. Clair Ave.
Cleveland, OH 44117-1194

John L. Gross
National Institute of Stds. & Technology
Building and Fire Research Lab,
Building 226, Room B158
Gaithersburg, MD 20899

John Theiss
EQE/Theiss Engineers
1848 Lackland Hills Parkway
St. Louis, MO 63146-3572
S-2

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

John H. Wiggins
J.H. Wiggins Company
1650 South Pacific Coast Hwy, Suite 311
Redondo Beach, CA 90277
Team Leaders for Topical Investigations
Helmut Krawinkler
Department of Civil Engineering
Stanford University
Stanford, CA 94305

Douglas A. Foutch
University of Illinois
MC-250, 205 N. Mathews Ave.
3129 Newmark Civil Engineering Lab
Urbana, IL 61801

Charles W. Roeder
University of Washington
233-B More Hall FX-10
Dept. of Building and Safety
Seattle, WA 98195-2700

Karl H. Frank
University of Texas at Austin
10100 Bornet Rd.
Ferguson Lab, P.R.C. #177
Austin, TX 78758

L. Thomas Tobin
Tobin and Associates
134 California Ave.
Mill Valley, CA 94941

Matthew Johnson
Edison Welding Institute
1250 Arthur E. Adams Drive
Columbus, OH 43221

Lead Guideline Writers
John D. Hooper
Skilling Ward Magnusson Barkshire, Inc.
1301 Fifth Avenue, Suite 3200
Seattle, WA 98101-2699

Robert E. Shaw
Steel Structures Technology Center, Inc.
42400 W Nine Mile Road
Novi, MI 48375-4132

Lawrence D. Reaveley
University of Utah
Civil Engineering Dept.
3220 Merrill Engineering Building
Salt Lake City, UT 84112

Raymond H. R. Tide
Wiss, Janney, Elstner Associates, Inc.
330 Pfingsten Road
Northbrook, IL 60062-2095
C. Allin Cornell, Associate Guideline Writer
Stanford University
Terman Engineering Center
Stanford, CA 94305-4020

Thomas A. Sabol
Englekirk & Sabol Consulting Engineers
P.O. Box 77-D
Los Angeles, CA 90007
C. Mark Saunders
Rutherford & Chekene
303 Second St., Suite 800 North
San Francisco, CA 94107

S-3

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Technical Advisory Panel (TAP) for Materials and Fracture
Dean C. Krouse*
705 Pine Top Drive
Bethelem, PA 18017

John M. Barsom, POC
Barsom Consulting, Ltd.
1316 Murray Ave, Suite 300
Pittsburgh, PA 15217
Serge Bouchard*
TradeARBED
825 Third Avenue, 35th Floor
New York, NY 10022

Frederick V. Lawrence
University of Illinois at Urbana-Champaign
205 N. Mathews Ave.
Room 2129 Newmark Lab
Urbana, IL 61801

Michael F. Engestrom*
Nucor-Yamato Steel
P.O. Box 678
Frederick, MD 21705-0678

Robert F. Preece
Preece, Goudie & Associates
100 Bush St., Suite 410
San Francisco, CA 94104

Karl H. Frank, Team Leader
University of Texas at Austin
10100 Bornet Rd.
Ferguson Lab, P.R.C. #177
Austin, TX 78758

Raymond H. R. Tide, Guideline Writer
Wiss, Janney, Elstner Associates, Inc.
330 Pfingsten Road
Northbrook, IL 60062-2095

Nestor Iwankiw, POC*
American Institute of Steel Construction
One East Wacker Dr., Suite 3100
Chicago, IL 60601-2001
TAP for Welding and Inspection
John M. Barsom, POC
Barsom Consulting, Ltd.
1316 Murray Ave, Suite 300
Pittsburgh, PA 15217

Matthew Johnson, Team Leader
Edison Welding Institute
1250 Arthur E. Adams Drive
Columbus, OH 43221

John W. Fisher
Lehigh University
117 ATLSS Drive
Bethlehem, PA 18015-4729

David Long
PDM Strocal, Inc.
2324 Navy Drive
Stockton, CA 95206

J. Ernesto Indacochea
University of Illinois at Chicago
Civil and Materials Engineering (mc 246)
842 West Taylor Street
Chicago, IL 60607

Duane K. Miller, POC
The Lincoln Electric Company
22801 St. Clair Ave.
Cleveland, OH 44117-1194

S-4

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Robert Pyle*
AISC Marketing
10101 South State Street
Sandy, Utah 84070

Richard I. Seals
P.O. Box 11327
Berkeley, CA 94712-2327
Robert E. Shaw, Guideline Writer
Steel Structures Technology Center, Inc.
42400 W Nine Mile Road
Novi, MI 48375-4132

Douglas Rees-Evans*
Steel Dynamics, Inc.
Structural Mill Division
2601 County Road 700 East
Columbia City, IN 46725

TAP for Connection Performance
Charlie Carter*
American Institute of Steel Construction
One East Wacker Drive, Suite 3100
Chicago, IL 60601-2001

Steve Powell*
SME Steel Contractors
5955 W. Wells Park Rd.
West Jordan, UT 84088

Robert H. Dodds
University of Illinois at Urbana-Champaign
205 N. Mathews Ave.
2129 Newmark Lab
Urbana, IL 61801

Charles W. Roeder, Team Leader
University of Washington
233-B More Hall FX-10
Dept. of Building and Safety
Seattle, WA 98195-2700

Roger Ferch, POC
Herrick Corporation
7021 Koll Center Parkway
P.O Box 9125
Pleasanton, CA 94566-9125

Stanley T. Rolfe
University of Kansas
Civil Engineering Department
2006 Learned Hall
Lawrence, KS 66045-2225

John D. Hooper, Guideline Writer
Skilling Ward Magnusson Barkshire, Inc.
1301 Fifth Avenue, Suite 3200
Seattle, WA 98101-2699

Rick Wilkinson*
Gayle Manufacturing Company
1455 East Kentucky
Woodland, CA 95695

Egor Popov
University of California at Berkeley
Department of Civil and Environmental
Engineering, Davis Hall
Berkeley, CA 94720

S-5

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

TAP for System Performance
Andrei M. Reinhorn
State University of New York at Buffalo
Civil Engineering Department
231 Ketter Hall
Buffalo, NY 14260

Jacques Cattan*
American Institute of Steel Construction
One East Wacker Drive, Suite 3100
Chicago, IL 60601-2001
Gary C. Hart
Hart Consultant Group
The Water Garden, Ste. 670E
2425 Olympic Blvd.
Santa Monica, CA 90404-4030

Arthur E. Ross, JVMC
Cole/Yee/Shubert & Associates
2500 Venture Oaks Way, Suite 100
Sacramento, CA 95833

Y. Henry Huang*
Los Angeles County Dept. of Public Works
900 S. Fremont Avenue, 8th Floor
Alhambra, CA 91803

C. Mark Saunders, Guideline Writer
Rutherford & Chekene
303 Second St., Suite 800 North
San Francisco, CA 94107

Helmut Krawinkler, Team Leader
Department of Civil Engineering
Stanford University
Stanford, CA 94305

W. Lee Shoemaker*
Metal Building Manufacturers Association
1300 Summer Avenue
Cleveland, OH 44115

Dennis Randall*
SME Steel Contractors
5955 West Wells Park Road
West Jordan, UT 84088

John Theiss, POC
EQE/Theiss Engineers
1848 Lackland Hills Parkway
St. Louis, MO 63146-3572

TAP for Performance Prediction and Evaluation
Vitelmo V. Bertero
University of California at Berkeley
Pacific Earthquake Engr. Research Center
1301 S. 46th St.
Richmond, CA 94804

Douglas A. Foutch, Team Leader
University of Illinois
MC-250, 205 N. Mathews Ave.
3129 Newmark Civil Engineering Lab
Urbana, IL 61801

Bruce R. Ellingwood
Johns Hopkins University
Department of Civil Engineering
3400 N. Charles St.
Baltimore, MD 21218

Theodore V. Galambos, POC
University of Minnesota
122 CE Building, 500 Pillsbury Dr. SE
Minnneapolis, MN 55455
Lawrence G. Griffis
Walter P. Moore & Associates
3131 Eastside, Second Floor
Houston, TX 77098

S-6

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Edwin T. Huston, JVMC
Smith & Huston, Inc.
8618 Roosevelt Way NE
Seattle, WA 98115

Thomas A. Sabol, Guideline Writer
Englekirk & Sabol Consulting Engineers
P.O. Box 77-D
Los Angeles, CA 90007

Harry Martin*
American Iron and Steel Institute
11899 Edgewood Road, Suite G
Auburn, CA 95603

Tom Schlafly*
American Institute of Steel Construction
One East Wacker Drive, Suite 3100
Chicago, IL 60601-2001
Technical Advisors

NormAbrahamson
Pacific Gas & Electric
P.O. Box 770000, MC N4C
San Francisco, CA 94177

Robert Kennedy
RPK Structural Mechanics Consultants
18971 Villa Terr
Yorba Linda, CA 92886

C.B. Crouse
URS – Dames and Moore
2025 First Avenue, Suite 500
Seattle, WA 98121
Social Economic and Policy Panel
Alan Merson
Morley Builders
2901 28th Street, Suite 100
Santa Monica, CA 90405

Martha Cox-Nitikman
Building and Owners and Managers
Association, Los Angeles
700 South Flower, Suite 2325
Los Angeles, CA 90017

Joanne Nigg
University of Delaware
Disaster Research Center
Newark, DE 19716

Karl Deppe
27502 Fawnskin Dr.
Rancho Palos Verdes, CA 90275

William Petak
University of Southern California
Lewis Hall, Room 201
650 Childs Way
Los Angeles, CA 90089

Eugene Lecomte
Institute for Business and Home Safety
6 Sheffield Drive
Billerica, MA 01821
James Madison
Attorney at Law, Mediator and Arbitrator
750 Menlo Avenue, Suite 250
Menlo Park, CA 94025

Francine Rabinovitz
Hamilton, Rabinovitz and Alschuler
1990 South Bundy Drive, Suite 777
Los Angeles, CA 90025

S-7

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Dennis Randall
SME Steel Contractors
5955 West Wells Park Road
West Jordan, UT 84088

Stephen Toth
TIAA-CREF
730 Third Avenue
New York, NY 10017-3206

David Ratterman
Stites and Harbison
400 West Market St., Suite 1800
Louisville, KY 40202-3352

John H. Wiggins, POC
J.H. Wiggins Company
1650 South Pacific Coast Hwy, Suite 311
Redondo Beach, CA 90277

L. Thomas Tobin, Panel Coordinator
134 California Ave.
Mill Valley, CA 94941
Performance of Steel Buildings in Past Earthquakes Subcontractors
Peter Maranian
Brandow & Johnston Associates
1660 West Third Street
Los Angeles, CA 90017

David Bonowitz
887 Bush, No. 610
San Francisco, CA 94108
Peter Clark
SAC Steel Project Technical Office
1301 South 46th St.
Richmond, CA 94804

Terrence Paret
Wiss Janney Elstner Associates, Inc.
2200 Powell St. Suite 925
Emeryville, CA 94602

Michael Durkin
Michael Durkin & Associates
22955 Leanora Dr.
Woodland Hills, CA 91367

Maryann Phipps
Degenkolb Engineers
225 Bush Street, Suite 1000
San Francisco, CA 94104

James Goltz
California Institute of Technology
Office of Earthquake Programs
Mail Code 252-21
Pasadena, CA 91125

Allan Porush
Dames & Moore
911 Wilshire Blvd., Suite 700
Los Angeles, CA 90017

Bruce Maison
7309 Lynn Ave
Elcerrito, CA 94530
Access Current Knowledge Subcontractors
David Bonowitz
887 Bush , No. 610
San Francisco, CA 94108

Stephen Liu
Colorado School of Mines
Mathematics and Computer Science
Department
Golden, CO 80401

S-8

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Materials and Fracture Subcontractors
Robert Dexter
University of Minnesota
122 Civil Engineering Building
500 Pillsbury Drive SE
Minneapolis, MN 55455-0116

Karl H. Frank
University of Texas at Austin
10100 Bornet Rd.
Ferguson Lab, P.R.C. #177
Austin, TX 78758

Welding and Inspection Subcontractors
Pingsha Dong / Tom Kilinski
Center for Welded Structures Research
Battelle Memorial Institute
501 King Avenue
Columbus, OH 43201-2693

Glenn M. Light / George Gruber
Southwest Research Institute
6220 Culebra Road, P. O. Drawer 28510
San Antonio, TX 78228-0510
William C. Mohr
Edison Welding Institute
1250 Arthur E. Adams Drive
Columbus, OH 43221

Matthew Johnson
Edison Welding Institute
1250 Arthur E. Adams Drive
Columbus, OH 43221

Connection Performance Subcontractors
Sherif El-Tawil / Sashi Kunnath
University of Central Florida
Civil and Environmental Engr. Department
Orlando, FL. 32816-2450

Gregory Deierlein
Stanford University
Terman Engineering Center
Department of Civil and Enviromental Engr.
Stanford, CA 94305-4020

Anthony Ingraffea
Cornell University
School of Civil Engineering
363 Hollister Hall
Ithaca, NY 14853

Charles W. Roeder
University of Washington
233-B More Hall FX-10
Seattle, WA 98195-2700

System Performance Subcontractors
Andrei M. Reinhorn
State University of New York at Buffalo
Civil Engineering Department
231 Ketter Hall
Buffalo, NY 14260

Paul Somerville
Woodward-Clyde Federal Services
566 El Dorado St., Suite 100
Pasadena, CA 91101-2560
Farzad Naeim
John A. Martin & Associates
1212 S. Flower Ave.
Los Angeles, CA 90015

C. Allin Cornell
Stanford University
Terman Engineering Center
Stanford, CA 94305-4020

S-9

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Helmut Krawinkler
Dept. of Civil Engineering
Stanford University
Stanford, CA 94305

Kazuhiko Kasai
Tokyo Institute of Technology
Structural Engineering Research Center
Nagatsuta, Midori-Ku
Yokohama 226-8503, JAPAN

Gregory MacRae
University of Washington
Civil Engineering Department
Seattle, WA 98195-2700

Bruce F. Maison
7309 Lynn Avenue
El Cerrito, CA 94530

Performance Prediction and Evaluation Subcontractors
James Anderson
University of Southern California
Civil Engineering Department
Los Angeles, CA 90089-2531

Gary C. Hart
Department of Civil and Environmental
Engineering
University of California
Los Angeles, CA 90095

Douglas A. Foutch
University of Illinois
MC-250, 205 N. Mathews Ave.
3129 Newmark Civil Engineering Lab
Urbana, IL 61801

Y.K. Wen
University of Illinois
3129 Newmark Civil Engineering Lab
205 N. Mathews Ave.
Urbana, IL 61801

Testing Subcontractors
Subhash Goel / Bozidar Stojadinovic
University of Michigan
Civil Engineering Department
Ann Arbor, MI 48109

Thomas Murray
Virginia Tech, Dept. of Civil Engineering
200 Patton Hall
Blacksburg, VA 24061

Roberto Leon
Georgia Institute of Technology
School of Civil & Environmental Engr.
790 Atlantic Ave.
Atlanta, GA 30332-0355

James M. Ricles / Le-Wu Lu
Lehigh University
c/o ATLSS Center
117 ATLSS Drive, H Building
Bethlehem, PA 18015-4729

Vitelmo V. Bertero / Andrew Whittaker
UC Berkeley
Pacific Earthquake Engr. Research Center
1301 S. 46th St.
Richmond, CA 94804

John M. Barsom
Barsom Consulting, Ltd.
1316 Murray Ave, Suite 300
Pittsburgh, PA 15217

S-10

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Hassan Astaneh
University of California at Berkeley
Dept. of Civil and Environmental Engr.
781 Davis Hall
Berkeley, CA 94720

Stephen Schneider
University of Ilinois at Urbana-Champaign
3106 Newmark Civil Engr. Lab, MC-250
205 N. Mathews Avenue
Urbana, IL 61801

Michael Engelhardt
University of Texas at Austin
Ferguson Laboratory
10100 Burnet Road, Building 177
Austin, TX 78712-1076

Matthew Johnson
Edison Welding Institute
1250 Arthur E. Adams Drive
Columbus, OH 43221
James Anderson
University of Southern California
Civil Engineering Department
Los Angeles, CA 90089-2531

Gary T. Fry
Texas A&M University
Department of Civil Engineering
Constructed Facilities Division, CE/TTI
Building, Room 710D
College Station, TX 77843-3136

Bozidar Stojadinovic
Dept. of Civil & Environmental Engr.
University of California
Berkeley, CA 94720

Chia-Ming Uang
University of California at San Diego
Dept. of AMES, Division of Structural Engr.
409 University Center
La Jolla, California 92093-0085

Inspection Procedure Consultants
Thomas Albert
Digiray Corporation
2235 Omega Road, No. 3
San Ramon, CA 94583

Andrey Mishin
AS & E High Energy Systems
330 Keller Street, Building 101
Santa Clara, CA 95054

Randal Fong
Automated Inspection Systems, Inc.
4861 Sunrise Drive, Suite 101
Martinez, CA 94553

Robert Shaw
Steel Structures Technology Center, Inc.
42400 W. Nine Mile Road
Novi, MI 48375-4132

Andre Lamarre
R.D Tech, Inc.
1200 St. Jean Baptiste, Suite 120
Quebec City, Quebec, Canada G2ZE 5E8

Carlos Ventura
Dept of Civil Engineering
University of British Columbia
2324 Main Hall
Vancouver, BC, Canada V6T 1Z4

Glenn Light
Southwest Research Institute
6220 Culebra Road
San Antonio, TX 78228

S-11

 Past Performance of
Steel Moment-Frame Buildings
in Earthquakes

FEMA-355E
SAC Phase II Project Participants

Guideline Trial Applications Subcontractors
John Hopper
Skilling Ward Magnusson Barkshire, Inc.
1301 Fifth Avenue, Suite 320
Seattle WA 98101-2699

Lawrence Novak
Skidmore, Owings, and Merrill
224 S. Michigan Ave, Suite 1000
Chicago, IL 60604

Leonard Joseph
Thornton-Tomassetti Engineers
641 6th Avenue, 7th Floor
New York, NY 10011

Maryann Phipps
Degenkolb Engineers
225 Bush Street, Suite 1000
San Francisco, CA 94104

Economic and Social Impact Study Subcontractors
Ronald Eguchi
EQE Engineering and Design
300 Commerce Dr., Ste. 200
Irvine, CA 92602

Charles Kircher
Charles Kircher & Associates
1121 San Antonio Road, Suite D-202
Palo Alto, CA 94303

Martin Gordon / Peter Morris
Adamson Associates
170 Columbus Avenue
San Francisco, CA 94133

Lizandro Mercado
Brandow & Johnston Associates
1600 West 3rd St.
Los Angeles, CA 90017

Richard Henige
Lemessurier Consultants Inc.
675 Massachusetts Ave.
Cambridge, MA 02139-3309

Greg Schindler
KPFF Consulting Engineers
1201 3rd Ave.
Seattle, WA 98101-3000

Report Production and Administrative Services
Carol Cameron, Publications Coordinator
Ericka Holmon, Admin. Assistant
California Universities for Research in
Earthquake Engineering
1301 S. 46th Street
Richmond, CA 94804

A. Gerald Brady, Technical Editor
Patricia A. Mork, Administrative Asst.
Peter N. Mork, Computer Specialist
Bernadette A. Mosby, Operations Admin.
Michelle S. Schwartzbach, Pub. Specialist
Applied Technology Council
555 Twin Dolphin Drive, Suite 550
Redwood City, CA 94065

*indicates industrial or organizational contact representative

S-12