2021 LATBSDC Proceedings

Proceedings of the 2021 Conference
THE 2021 LOS ANGELES TALL BUILDINGS

CONFERENCE
LOS ANGELES, CALIFORNIA
NOVEMBER 12, 2021

PROCEEDINGS
LOS ANGELES TALL BUILDINGS
STRUCTURAL DESIGN COUNCIL
THE 2021 LOS ANGELES TALL

BUILDINGS CONFERENCE
NOVEMBER 12, 2021
Email: Info@LATallBuildings.org
Tel: (626) 389-1888

https://www.LATallBuildings.org
Copyright ©2021 by the Los Angeles Tall Buildings Structural Design Council
All rights reserved. No part of this book may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, recording, or by any information
storage and retrieval system, without permission from the publisher.
The Council was formed in 1988 to provide a forum for the discussion of issues relating to the design of
tall buildings. Its members seek to advance state-of-the-art structural design through interaction with other
professional organizations, building departments, and university researchers as well as recognize significant
contributions to the structural design of tall buildings. The Los Angeles Council is an affiliate of the
Council on Tall Buildings and Urban Habitat (CTBUH).
ii
The Council is a nonprofit California corporation whose members are those individuals who have
demonstrated exceptional professional accomplishments in the structural design of tall buildings. The
Annual Meeting of the Council presents a program to engineers, architects, contractors, building officials
and students that include research reports on areas of emerging importance, case studies of current structural
design and consensus documents by the membership on contemporary design issues.
Active Members
Mr. Shahen Akelyan Dr. Ifa Kashefi Mr. Barry Schindler

L.A. Dept. of Building & Safety KM Engineering and Consulting John A. Martin & Associates
Dr. Gregg E. Brandow Prof. Kristijan Kolozvari Prof. John W. Wallace

Brandow & Nastar, Inc. California State University, Fullerton University of California, Los Angeles
Dr. Lauren D. Carpenter Dr. Marshall Lew Mr. Philip Yin

WHL International, Inc. Wood Environment & Infrastructure City of Long Beach
Dr. Chukwuma Ekwueme Dr. Michael Mehrain Mr. Nabih F. Youssef

Thornton Tomasetti Mehrain Naeim International, Inc. Nabih Youssef Associates
Mr. Tony Ghodsi Dr. Farzad Naeim Prof. Farzin Zareian

Englekirk Farzad Naeim, Inc. University of California, Irvine
Dr. Saiful Islam Dr. Thomas A. Sabol Mr. Atila Zekioglu

Saiful/Bouquet Inc. Englekirk ARUP
Emeritus Members
Mr. Lawrence Brugger Mr. Robert N. Harder Mr. John A. Martin, Jr.
Mr. John Gavan Mr. Richard Holguin Mr. Donald R. Strand
Dr. Robert E. Englekirk Dr. Sampson C. Huang Mr. Edward J. Teal
Mr. Colin Kumabe
Deceased Members
Mr. Brian L. Cochran Dr. Gary C. Hart Dr. George W. Housner Mr. Clarkson W. Pinkham
Mr. Nick Delli Quadri Mr. Roy G. Johnston Mr. John A, Martin, Sr.
iii
Los Angeles Tall Buildings Structural Design Council
November 12, 2021
It is my great pleasure to welcome you all to the 2021 LA Tall Buildings Annual Conference!
I am so glad that we are able to do the conference in-person again. It’s not just the presentations, it is
also the interactions among all the attendees and all the technical discussions that occur during the
breaks that make the in-person conference so much more effective. I sincerely hope we are coming out
of this Covid crisis now, a crisis that none of us will ever forget.
This year’s conference, like others in the past, continues a tradition of providing a forum for design
professionals, researchers, academics, building officials, students and our industry partners to share the
latest and greatest on seismic design of tall buildings and special structures.
We have put together an excellent program with distinguished speakers addressing an array of
important topics ranging from case studies of some of the high profile projects in Southern California,
session on new methods and procedures which includes new ACI Provisions for Shear Walls, and a
session on Seismic Resilient Design and Structural Health Monitoring. I want to thank all the
distinguished speakers for their contributions.
I am very glad to be holding our conference this year at this great venue. I am especially thankful to all
of our corporate sponsors, exhibitors, and our Grand Event Sponsor. This is the first time ever we have
had that many corporate sponsors, 16 in total.
And finally, I want to thank all of you for attending our conference today.
Sincerely,
Saiful Islam, PH.D., S.E.

President, Los Angeles Tall Buildings Structural Design Council
Chairman & CEO, Saiful Bouquet, Inc.
545 South Figueroa Street, Suite 1223

Los Angeles, CA 90017 www.latallbuildings.org
CONFERENCE PROCEEDINGS
TABLE OF CONTENTS
CASE STUDY – THE GRAND LA 1
Patrick Lindblom and Scott Erickson
CASE STUDY – SoFi STADIUM, INGLEWOOD, CALIFORNIA 13

Rafael Sabelli, Mark Waggoner and Ozgur Atlayan
SEISMIC RETROFIT OF A PRE-NORTHRIDGE STEEL MOMENT FRAME BUILDING 49

USING ASCE 41-17
Kevin O’Connell, David Gonzalez, Guzhao Li, Emily McCarthy, Russell A. McLellan
ASCE/SEI 41 ASSESSMENT OF REINFORCED CONCRETE BUILDINGS: 71

BENCHMARKING ASCE/SEI 41 LINEAR AND NONLINEAR DYNAMIC PROCEDURES
WITH EMPIRICAL DAMAGE OBSERVATIONS
Russ Berkowitz, Andrew Sen and Dustin Cook
NEW NONLINEAR MODELING PARAMETERS AND ACCEPTANCE CRITERIA 72

FOR RC STRUCTURAL WALLS
Saman A. Abdullah and John W. Wallace
ATC-145 UPDATE: DRAFT GUIDELINE FOR POST-EARTHQUAKE ASSESSMENT, 86

REPAIR AND RETROFIT OF BUILDINGS
K.J. Elwood and J.P. Moehle
PHENOMENOLOGICAL NONLINEAR MODELING FOR PERFORMANCE-BASED 95

DESIGN OF HIGH-RISE SHEAR WALL BUILDINGS
Ian McFarlane, Juan D. Piotrowski, Kevin Aswegan, Juan D. Pozo, Kristijan Kolovari and John
Hooper
MULTI-PERIOD RESPONSE SPECTRA 110

Sanaz Rezaeian and Nicolas Luco
THE PROPOSED ASCE 7-22 MULTI-PERIOD RESPONSE SPECTRA – IMPACT ON THE 130
LOS ANGELES METROPOLITAN AREA
Marshall Lew and Kenneth S. Hudson
SEISMIC STRUCTURAL HEALTH MONITORING AS AN ESSENTIAL INGREDIENT OF 144

RESILIENT DESIGN
Farzad Naeim
CHILE S2HM EXPERIENCE 156

Ruben Boroschek
LESSONS LEARNED FROM SIX-YEAR EXPERIENCE ON MARKET-BASED 165

IMPLEMENTATION OF SSHM NAMED q-NAVI
Masayoshi Nakashima, Katsuhisa Kanda, Saori Ogasawara and Yu Fukutomi
NEW ACI PROVISIONS FOR SHEAR WALL DESIGN AND CASE STUDIES 177
John Wallace, Saiful Islam, Tony Ghodsi, Vladimir Volnyy, Rishabh Singhvi, Akshay Patil
v
LosAngeles Tall Buildings Structural Design Council
FINAL PROGRAM FOR THE 2021 CONFERENCE

9:00-9:10 am WELCOME & INTRODUCTIONS
Dr. Saiful Islam, President, LATBSDC
Session on Project Case Studies
9:10-9:40 CASE STUDY #1: GRAND AVENUE PROJECT, LOS ANGELES
Patrick Lindblom and Scott Erickson, DCI Engineers
9:40-10:10 CASE STUDY #2: SOFI STADIUM, LOS ANGELES – STRUCTURAL DESIGN AND
SEISMIC INSTRUMENTATION
Rafael Sabelli, Mark Waggoner, and Ozgur Atlayan, Walter P. Moore
Session on Seismic Retrofit of Structures
10:10-10:40 SEISMIC RETROFIT OF A PRE-NORTHRIDGE STEEL MOMENT FRAME BUILDING
USING ASCE 41-17
K D. O'Connell, D. Gonzalez, G. Li, E. McCarthy, and R.A. McLellan, SGH
10:40-10:50 BREAK (10 min)
10:50-11:40 ASCE/SEI 41 ASSESSMENT OF REINFORCED CONCRETE BUILDINGS:
BENCHMARKING ASCE/SEI 41 LINEAR AND NONLINEAR DYNAMIC PROCEDURES
WITH EMPIRICAL DAMAGE OBSERVATIONS
Russ Berkowitz, Andrew Sen and Dustin Cook
11:40-12:10 UPDATES TO ATC-140 PROVISIONS AND EXAMPLES
Saman Abdullah and John Wallace, UCLA
12:10-1:00 pm LUNCH BREAK
Session on New Methods and Procedures
1:00-1:30 ATC-145 UPDATE: DRAFT GUIDELINE FOR POST-EARTHQUAKE ASSESSMENT,
REPAIR, AND RETROFIT OF BUILDINGS
K.J. Elwood and J.P. Moehle
1:30-2:00 NEW ACI PROVISIONS FOR SHEAR WALL DESIGN AND CASE STUDIES
John Wallace, Saiful Islam, Tony Ghodsi, Vladimir Volnyy, Rishabh Singhvi, Akshay Patil
2:00-2:30 PHENOMENOLOGICAL NONLINEAR MODELING FOR PERFORMANCE BASED
DESIGN OF HIGH-RISE SHEAR WALL BUILDINGS
Ian McFarlane, Juan D. Piotrowski, Kevin Aswegan, Juan D. Pozo, Kristijan Kolovari and John
Hooper
Session on Design Ground Motions
2:30-3:00 MULTI-PERIOD RESPONSE SPECTRA
Sanaz Rezaeian and Nicolas Luco, USGS
3:00-3:10 BREAK (10 min)
3:10-3:40 THE PROPOSED ASCE MULTI-PERIOD RESPONSE SPECTRA – IMPACT ON THE LOS
ANGELES METROPOLITAN AREA
Marshall Lew and Kenneth S. Hudson, Wood
Session on Resilient Design and Structural Health Monitoring
3:40-3:45 LATBSDC PROGRESS REPORT
Atila Zekioglo, Arup
3:45-4:00 SEISMIC STRUCTURAL HEALTH MONITORING AS AN ESSENTIAL INGREDIENT OF
RESILIENT DESIGN
Farzad Naeim
4:00-4:30 CHILE S2HM EXPERIENCE
Ruben Boroschek, University of Chile
4:30-5:00 LESSONS LEARNED FROM SIX-YEAR EXPERIENCE ON MARKET-BASED
IMPLEMENTATION OF SSHM NAMED “q-NAVI”
M. Nakashima, K. Kanda, S. Ogasawara, and Y. Fukutomi, Kobori Research Complex Inc.
5:00 ADJOURN
vi
Thanks to the Corporate Sponsors of the Conference
GRAND SPONSOR:
GOLD SPONSORS:
SILVER SPONSORS:
vii
Proceedings of the 2021 Los Angeles Tall Buildings Structural Design Council Conference
CASE STUDY – THE GRAND LA
Patrick Lindblom, P.E., S.E., Scott Erickson, P.E., S.E.

DCI Engineers
Abstract
The Grand LA is a multi-tower mixed-use development in the Bunker Hill neighborhood of downtown
Los Angeles. The project consists of a 500-foot (150 m) tall, 45-story residential tower and a 285-foot (85
m) tall, 28-story hotel tower, located in the full city block bounded by South Grand Avenue, South Olive
Street, West 1st Street, and West 2nd Street, and across Grand Ave from the iconic Walt Disney Concert
Hall. The $1 billion project is scheduled to open in early 2022 and will include 1.6 million ft2 of
residential, hotel, amenity, retail, food and beverage, and parking uses. This paper touches the surface of
some of the highlights of the journey from concept to completion.
Figure 1 – Architectural Rendering
History
In 2001, a public-private committee was formed with the goal of “Reimagining Grand Avenue”. A Grand
Avenue Joint Powers Authority was formed in 2003, consolidating agencies and properties owned by the
City of Los Angeles and Los Angeles County to form The Grand Avenue Project. Related Companies
1
was granted rights to develop multiple parcels in 2004, and in 2007 the City and County approved the
Grand Avenue project. 2012 saw the renovated 16-acre Grand Park open and in 2014, parcel M, called
The Emerson, a 19-story residential development, was completed. Parcel L, The Broad Museum, was
completed in 2015. Parcel Q, aka, The Grand, will be completed in 2022.
Project Team
DCI Engineers started design on The Grand in 2014 with Related and Gehry Partners. Early on, the
project went through various massing studies and then a period of waiting to figure out how to
accomplish the ambitious concepts. Excavation reached the bottom of the hole in the middle of 2019 and
the building will be complete in 2022. Approximately 5 years of design and 3 years of construction
through covid, and all along the way, the project team members have remained consistent, of the highest
quality, and focused on bringing this amazing project to reality.
Building Gravity Framing Overview
Both towers are built in stacks that slide, rotate, and shift as the building goes up. Each stack is required
to appear to be sitting on the stack below, as if it could slide right off; so the structure and the cladding at
the interface was coordinated to achieve this. Coordinating the columns to thread through each different
residential layout stack was a huge undertaking. Not to mention column locations in the retail spaces,
parking levels, amenity areas, open space, and the stack sliding surfaces. This was one of the most
challenging aspects of the design. Figure 2 shows an overlay of the different stacks and how the floor
plates change through the height of the buildings. Multiple walking columns and transfer beams were
used to support the changing slabs and accommodate the floor layouts.
Figure 2 – Floor Plan Overlay
The building is mostly cast in place concrete with some steel framing in the western corner (bottom right)
of the site at the upper portion of the podium levels. Each tower is supported by a concrete mat
2
foundation, with some columns supported by spread footings. The mat foundations were required to
control settlement. Of particular note is an existing drainage tunnel that runs below the residential tower.
Studies were done to determine that we were not going to place any more pressure on the tunnel with the
new construction than was existing.
Parking levels exist from the lowest below grade level up to the first level below Upper Grand Avenue.
The site is accesses from 1st and 2nd Avenues (you can drive straight through from one side to the other at
the mid-block entrances. There is also access from Olive Street and via Lower Grand Avenue to the
loading dock. All of these entrances are fairly concealed by the surrounding architecture and features of
the site. The parking floors are all mild reinforced concrete slabs.
The need for flexibility demanded that the retail slabs be mild reinforced concrete as well. However, at
the main plaza slab off of Upper Grand Avenue, as well as floors above and below this, there are a fair
amount of transfer beams that required post-tensioning. Because of this, there are some areas that are not
quite as flexible as others. A tenant leasing plan was developed to show these transfer beam areas to help
with future space planning needs.
The residential and hotel tower floors are post-tensioned concrete flat slabs.
The eastern corner of the site was originally designed with steel framing to span across the theatres and
several steel trusses were used to span across and support the ballroom and amenity deck and pool above.
During construction, this portion of the project was redesigned, as described below and became mostly
concrete with some steel framed levels for the last two floors. Steel trusses still span over the ballroom
and support the pool and amenity deck above.
Finally, there is a raised central platform area between the towers that rises up from an opening in the
Upper Grand level slab, supported from the level below. This platform is a steel truss moment frame
system supported on 4ft diameter concrete columns. The platform will have restaurant / bar space on it
and will be topped with multiple 50ft tall sail sculpures. The platform is accessed via steel truss bridges
that span from both the residential tower and the hotel tower. There is also another bridge that spans from
one tower to the other at the Upper Grand level; which is supported off of the truss moment frame
columns. Figure 4 shows the central platform with the connecting bridges and various pedestrian
pathways to circulate through the site.
3
Firgure 4 – Podium Layering & Circulation
Building Lateral Framing Overview and Seismic Performance Based Design (SPBD)
Both towers and the podium are all one building. There is no seismic joint anywhere, except at the
interface to the adjoing Upper Grand Avenue bridge that abuts the property and between the bridges that
extend from the central platform and the towers they land on. We decided to pursue this approach to offer
flexibility in the layout of the lateral system below grade and avoid back to back lateral elements down
the center of the building. This also avoided a seismic jointing in the plaza finishes.
Both towers are above the code prescriptive height limit for concrete shear wall buildings, so a Seismic
Performance Based Design was undertaken. This gave us the flexibility to also take other code exceptions
to make the design as efficient as possible. Code exceptions of note are the prescriptive height limit,
redundancy factor (), grade 80 longitudinal reinforcing for the verticals in the lateral system, and the
building risk category.
The occupancy of each tower was calculated, from an egress perspective, by adding the occupants from
top to bottom, and the floor where the occupancy exceeded 5,000 was determined. For both towers this
occurred at level 6. Therefor at level 6 and below, Risk Category III was assigned, and above level 6,
Risk Category II was assigned. Per LATBSDC, the deformation-controlled actions from level 6 to level 8
also used the Risk Category III designation. Figure 5 shows the delineation of risk categories.
4
Figure 5 –Risk Category Diagram
Data from a previous tower design of ours, along with the fact that we had the same peer reviewers on
this project, allowed us to pursue using Grade 80 vertical reinforcing in the shear walls before it was
introduced into the ACI code.
We are grateful to Farzad Naeim, John Wallace, Paul Somerville, Sophia Gavridou, and Kristijan
Kolozvari for their peer review efforts and insight on this complicated project. And we are thankful for
Marty Hudson’s work on providing the time histories and geotechnical needs for the development.
The building lateral system was kept as simple as possible; utilizing a central concrete core through the
height of both towers and adding concrete blade walls at the hotel tower wings and buckling restrained
braces (BRBs) at the residential tower wings. The BRBs were used to more easily tune the structure
stiffness as well as to allow them to transfer out below the Upper Grand Avenue diaphragm (below the
seismic base). This allowed much more architectural freedom in the parking levels below. The eastern
corner of the site used concrete blade walls for future “flexibility” to take steel framing out and replan the
space. That flexibility came in handy when we redesigned the entire area during construction to a mostly
concrete framed portion of the project. We also used some concrete blade walls up against the Grand
Avenue side of the project to resist unbalanced soil loads.
Conventional, diagonal, and Steel Reinforced Composite (SRC) coupling beams were all used to further
tune the response of the building and meet the rotational demands induced by the ground motions. Figure
6 shows two different types of coupling beams being used in one coupling beam bay to meet the
rotational demands.
5
Figure 6 – Coupling Beam S3 Rotational Response
Diaphragm design was made immensly simpler by our use of diaphragm heat maps and various section
cuts throughout the diaphragms. The Perform 3D model floors were discretized appropriately and each
heat map exported showed the maximum shear value at each element. This was helpful in understanding
and focusing our attention on the flow of force, which helped us reinforce elements properly. The section
cuts used throughout were able to capture combined axial, flexure, and shear as needed at critical
sections. This combined approach was very important due to the unusal shapes of most of the
diaphragms, especially near the seismic base of the building.
Wind Considerations
The Grand wind tunnel analysis was performed by CPP. They developed the structural wind loads for the
project, came up with strategies to mitigate sensitive outdoor terrace areas, and evaluated the building for
occupant comfort accelerations. They were also extremely helpful evaluating the various architectural
concepts developed for the artwork / sculpture that sits on top of the central platform structure, which is
tucked in between both towers, but behaves completely differently from a typical “building” due to the
truss moment frame it is supported by.
6
It is always helpful to estimate how stiff to make the lateral system before finalizing the design, so no
other structural modifications are required for occupant comfort. In general, this is typically done by
looking at primary building periods, aspect ratios for the lateral system, considering the shape of the
building, and by experience. For the residential tower, which was of primary consideration, the height of
the building (from the roof to the base at Olive Street) divided by the out to out of the concrete core is
10:1 in the long direction and 13.5:1 in the short direction. These are measured from the average
thickness of the walls at mid-height of the building. Of note are the wings of the tower, which gradually
wedding cake outward from approximately half-way up the building downward towards the base. These
are supplementally supported by BRBs to stabilize the sway of the wings as they extend further out from
the main tower. The braces stop at Upper Grand and do not extend down below this, but the columns
continue down to the foundation. Aspect ratios for these vary, but in general, they are on the order of 9.5
to 12:1 (measured from centerline of columns) with the height taken as only the height of the braces.
Figure 7 shows the wind tunnel model and setup in the testing tunnel.
Figure 7 – Wind Tunnel Testing
Predicted wind accelerations satisfied residential occupant comfort criteria and aligned closely, indicating
that the lateral system layout was sized just right.
Building Information Modeling
Gehry Partners has a unique documentation standard. The intent is for the 3D model to be the primary
construction document. All of the building geometry is to be modeled explicitly in the 3D model. This
may seem like the natural next step in the evolution of BIM, but those that are familiar with current
modeling practice may be aware of some of the subtleties between the model being the final construction
document and 2D drawings.
• Grid systems were used only for orientation and not tied to any specific building element.
• Sizing of structural elements (columns, shear walls, foundations, etc…) shown on plan are only
for reference, and the modeled sizes are the controlling dimension.
• Vertical structural elements are broken at each floor, rather than extended full height.
• Slab and soffit steps are modeled and any space between elevations must be filled.
• Tolerance for slab extents/geometry have to be reviewed exhaustively with the architect.
• If design changes are to be issued after the construction documents, an as-built model has to be
maintained.
Despite the effort required to build and maintain the 3D Revit model, it was invaluable in coordination
and clash detection. Particularly at the podium levels where the framing and complexity would have been
7
very difficult to convey on 2D drawings in an efficient manner. As construction proceeded, Visicon

software was used to overlay multiple architectural and structural models daily to continually relate to
conditions in the field in a fast, virtual environment. Figure 8 shows the Architectural building model in
Visicon.
Figure 8 – Architectural Building Model in Visicon
Cladding Coordination
To ease installation and increase efficiency of the cladding, drilled in connections into pieces of
embedded tubes at the top and bottom of the slab edges were used. This in turn increased the complexity
of the post-tension tendon installation, as the live end stressing anchor heads had to fit in between the
perimeter embed tubes. This became even more complicated at the sliding stack joint floors where the
cladding was offset from one floor to the next and the embed tubes were also offset. This is just one
example of the complexity of the building and how the entire team endeavored to work together. Figure 9
shows the coordination between the cladding embed detailer and the PT subcontractor.
8
Figure 9 – Embed and PT subcontrator coordination layout

Re-Designs
To keep tenant flexibility, the podium floors were largely framed with mild reinforced slabs. However, at
the corner of Olive and 1st, the program was intended to have a multi-floor cinema. To maintain open
sight-lines and theater seating, this sector was framed with steel and composite metal deck.
As construction began, theater operators were hesitant to lease the space, so the owner made the decision
to change this to a more conventional retail space. This meant changing the cinema floors from steel to
concrete, re-laying out the columns, and completely changing the vertical transportation scheme. This
also had major implications for the lateral system. To add to the urgency, the foundation was scheduled to
be poured in two months. Figure 10 shows sketches and notes about what was revised for this ‘Cinema
Re-stack.’
9
Figure 10 – Sketch of Cinema Re-stack

Another re-design that was more of a coordination effort was the Floor Re-label. The primary frontage for
the project is along Upper Grand Avenue, and the site slopes down 30’ to Olive Street on the east. The
Los Angeles Department of Building and Safety had approved the original floor labeling of Grand
Avenue being level L1, and Olive Street being L0, with two floors in between, which were too large to be
called mezzanines, so they were labeled Valet floors, V1 and V2.
During construction, while shop drawings were being submitted and checked, but before final permit
approval, the new fire marshal insisted that the lowest grade that fire trucks can pull up and enter the site
from, be labeled L1. After a lot of fruitless back and forth, it was determined to change the floor labeling
as requested, and to remove L13 in the mix. So, after a lot of additional coordination and re-labeling
during construction we came out with a building that had an additional four stories, without making it
taller. Figure 11 shows a section showing the floor re-labeling scheme.
Figure 11 – Floor Re-label

Construction
At the beginning of design in 2014, the Temporary Certificate of Occupancy (TCO) date was set by the
developer to be the first quarter of 2022. This date has not moved through design, financing, permitting,
construction, or the pandemic. The construction and design teams have been incredibly focused on
meeting the developer’s goal, and the building is nearing completion. The structure topped out in
February of 2021, and fit-out work is on track for the project delivery date.
Figure 12 shows the construction of the two towers from above. Figure 13 shows cantilevering elements
to support the twisting and turning floor transitions.
10
Figure 12 – Aerial photo of construction
Figure 13 – Cantilevering structural elements at floor transitions
11
Figure 14 shows the construction cam status on the day that California was ordered to stay at home for the
covid-19 pandemic, and where it stands today. Once the people of Los Angeles start venturing into
Bunker Hill again, they will be greeted with an incredible new venue for dining, hospitality, living and
more.
March 19, 2020 October 20, 2021

Figure 14 – Construction Camera
12
CASE STUDY: SoFi STADIUM, INGLEWOOD, CALIFORNIA
Rafael Sabelli, Mark Waggoner and Ozgur Atlayan

Walter P Moore
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
SEISMIC RETROFIT OF A PRE-NORTHRIDGE STEEL MOMENT

FRAME BUILDING USING ASCE 41-17
Kevin O’Connell, S.E.

David Gonzalez, S.E.
Guzhao Li, Ph.D., S.E.
Emily McCarthy, Ph.D., P.E.
Russell A. McLellan, P.E.
Simpson Gumpertz & Heger, Inc.
Abstract
This case study outlines the retrofit approach and challenges to bring an existing seventeen-story, 335,000
sq. ft., steel-framed building, located in Los Angeles County, with typical Pre-Northridge steel moment-
resisting connections constructed circa 1970 into general compliance with ASCE 41-17.
The voluntary seismic upgrade objectives are to meet the ASCE 41-17 Basic Performance Objective for
Existing Buildings (BPOE) and reduce the Probable Maximum Loss (PML) to less than 20%, based on the
Scenario Upper Loss (SUL) for an earthquake hazard with a 475-year return period.
The building consists of a fifteen-story tower above-grade with steel moment frames on all four sides of
the perimeter. The moment connections are Pre-Northridge Welded Unreinforced Flange – Welded Web
(WUF-W). There are two parking levels partially below-grade constructed with a gravity steel frame and
precast concrete shear walls around the perimeter.
We performed nonlinear static and nonlinear time history analyses following ASCE 41-17. The preferred
seismic retrofit approach involves installing piston-type fluid viscous dampers (FVDs) to protect the
existing moment connections and reduce demands on existing columns, column splices, and column bases.
Limited foundation retrofit of parking garage precast shear walls was required.
This paper presents the methodology followed to evaluate vulnerable WUF-W connections and panel zones
using the latest provisions of ASCE 41-17. Database scripting techniques were utilized to extract
information from the vast output files which significantly reduced the post-processing time. The project
was peer-reviewed by a team of structural and geotechnical engineers. Construction is now complete.
Introduction
Building Description
The building is a seventeen-story steel frame superstructure with two levels of partially underground
parking constructed circa 1970. The building’s lateral force resisting system (LFRS) comprises perimeter
steel moment-resisting frames utilizing “pre-Northridge” Welded Unreinforced Flange - Welded Web
(WUF-W) beam-column moment connections.
49
Photo 1. Exterior View of Building During Construction
The building is rectangular in plan, with the tower having overall dimensions of approximately 106 ft in
the north-south direction by 211 ft in the east-west direction (Figure 1). Two partially subterranean levels
have overall dimensions of approximately 268 ft in the north-south direction by 211 ft in the east-west
direction.
50
Figure 1. Typical Floor (3rd to 16th) Framing Plan (simplified for clarity)
The elevated floors consist of a 3-1/4 in. lightweight concrete fill slab reinforced with welded wire mesh
over steel deck spanning between steel wide flange beams. The lowest basement level floor is a concrete
slab-on-grade. Steel-wide flange beams and columns vertically support the floors and roof. Foundations
consist of isolated spread footings beneath interior columns and strip footings beneath perimeter retaining
walls.
Lateral loads are distributed to the perimeter steel moment frames by the metal deck and concrete-filled
slab acting as a diaphragm. The east and west moment frame lines have seven bays with a typical bay width
of 15 ft. The north and south moment frame lines have five bays with a typical bay width of 30 ft. All
columns in the moment frames are oriented on the strong axis. The first story height is 20 ft-6 in. The height
of the above stories is 13 ft-1 in. Typical moment connections consist of complete joint penetration (CJP)
welds between the beam flanges and the column flanges and welded shear tabs. The column splices consist
of partial joint penetration (PJP) welds and erection plates.
Objective and Approach
The structural performance objective of this voluntary seismic retrofit design is to meet the structural
requirements of the ASCE 41-17 Basic Performance Objective for Existing Buildings (BPOE). The
evaluation of nonstructural components to meet BPOE is not within the scope of this voluntary retrofit. The
subject building is classified as Risk Category II. In meeting BPOE, we found the building’s Probable
Maximum Loss (PML) to be less than 20%, based on the Scenario Upper Loss (SUL) for an earthquake
hazard with a 475-year return period.
We performed both linear dynamic analysis and nonlinear static (pushover) analysis for the existing
building to identify critical structural deficiencies (such as columns splices, panel zones, and substructure
shear walls) and confirm that seismic retrofit is required to meet the selected BPOE. Then, following the
ASCE 41-17 Nonlinear Dynamic Procedure (NDP), the retrofit components are designed, modeled, and
refined to reach compliance with the selected BPOE. The seismic retrofit intends to reduce potential damage
by improving building performance.
51
Seismic Ground Motions
Site-specific response spectra with 5% effective damping for the use with the linear dynamic procedure
(LDP) were developed by the project geotechnical consultant for both BSE-1E (Basic Safety Earthquake
1) and BSE-2E (Basic Safety Earthquake 2) conditions, as shown in Figure 2.
Figure 2. Site-Specific Response Spectra (5% damping) for the Use with LDP
The project geotechnical consultant also provided us with eleven pairs of acceleration-time histories for the
BSE-2E and BSE-1E, respectively, for the use with NDP per the requirements of ASCE 41-17. Each set of
time histories consists of two orthogonal horizontal ground motion components (rotated to fault normal and
fault parallel for near-fault events and modified to the building axes, i.e., N-S and E-W directions). The
geotechnical consultant selected the acceleration-time histories using the following criteria (among others)
based on the results of disaggregation of hazard of various periods:
• Earthquake magnitude range: selected records from 6.0 to 7.5 (local), 7.0 to 8.0 (distant)
• Joyner- Boore (Rjb) distance: selected records both from 0 to 25 kilometers (local), and 25 to 200
kilometers (distant)
• Fault mechanism: selected records from strike-slip, reverse, and reverse-oblique
• Initial scale factor for seed time history: 0.2 to 4.0
Structural Deficiencies and Seismic Retrofit
Deficiencies identified using LDP
To preliminarily identify deficiencies in the building’s LFRS, we performed a linear dynamic analysis of
the moment frames following the ASCE 41-17 LDP requirements. We created CSI ETABS model, as
shown in Figure 3, for the linear dynamic analysis. We explicitly modeled the moment frames and used
52
rigid diaphragms with mass lumped at the center to model each floor. We also included the substructure
and lower-level shear walls in the model.
Figure 3. Overview of the ETABS Model

(Existing Building)
The analysis results indicated potential deficiencies in the moment frames’ panel zones and column splices.
The substructure analysis (podium diaphragm, collector elements, shear wall panels) indicated an
insufficient capacity for some shear wall panels.
Deficiencies confirmed using NSP.
To confirm the deficiencies (panel zones, column splices) in the moment frames identified using LDP, we
performed a nonlinear static (pushover) analysis for the existing building following the ASCE 41-17
nonlinear static procedure (NSP) requirements.
53
We created a three-dimensional model using CSI PERFORM-3D V7.0.1. The PERFORM-3D model, as
shown in Figure 4, consisted of the building’s superstructure and substructure, capturing its geometry, mass
distribution, and elastic-plastic force-deformation behavior.
Figure 4. Overview of the PERFORM-3D Model

(Existing Building)
We modeled all diaphragms above the first-floor level as rigid. We modeled the two floors in the
substructure as semi-rigid to better represent the force transfer from the superstructure to the substructure.
We modeled the concrete panel shear walls below the ground floor using elastic planar shell elements. We
modeled all the other primary and secondary elements below the ground floor using elastic sections with
end moment releases as applicable.
We modeled the panel zones using connection panel zone elements with a rotational spring representing
the panel zone shear mechanism. A typical frame compound element may include an auto rigid end zone
54
(at both ends), a WUF-W moment hinge (at both ends), or a P-M column hinge (at both ends), and an elastic
beam or column section between the beam or column hinges. We also included strength sections in the
compound element to confirm elastic behavior at locations where plastic hinges are not expected, such as
column splice locations.
For deformation-controlled components, such as WUF-W moment hinge, P-M column hinge, and panel
zone elements, we defined their parameters and acceptance criteria at different performance levels
following the ASCE 41-17 requirements.
For pre-Northridge WUF-W moment connections with panel shear to shear strength ratio (Vpz/Vy) greater
than 1.10, ASCE 41-17 introduced an upper bound limit for the panel zone Life Safety (LS) and Collapse
Prevention (CP) performance level acceptance criteria, as shown in ASCE 41-17 Equation 9-19. The upper
bound limit is 50% of the panel zone LS and CP acceptance criteria as listed in Table 9-7-2 of ASCE 41-
17. There was no such limit for the previous versions of ASCE 41. We understand the basis for the change
in ASCE 41-17 is due to tests performed at the University of California, San Diego by Kim, et. al. It is
worth noting that panel zone performance with pre-Northridge WUF-W moment connections meeting the
ASCE 41-13 acceptance criteria may not satisfy the ASCE 41-17 requirements due to this upper bound
limit. The resulting CP plastic rotation for the panel zones in our project was between 0.5% and 1.5%,
which is generally between 3θy and 6θy, where θy is the yield rotation of the panel zone shear mechanism.
In ASCE 41-13 the CP limit is 12θy, which would have allowed two to four times the plastic rotation
relative to ASCE 41-17. A further consequence of the reduction in the plastic rotation is a reduction in
strength. Strain hardening is significant for panel zones. The default panel zone strain hardening slope is
6% in ASCE 41, with higher slopes acceptable based on test data. When the panel zone is defined with
significant ductility and strain hardening, the ultimate strength of the panel zone often becomes large
enough to allow for distributed yielding in the adjacent beams and columns. In our case, almost no beams
yielded, mainly because the panel zones could not develop the yield strength of the beams. Panel zone shear
is an established ductile mechanism. The additional limitations on the panel zone rotation introduced in
ASCE 41-17 are meant to capture the limit state of the excessive panel zone deformation causing column
“kinking,” which fractures the beam-column flange welds. So, it is a moment connection failure tracked by
the panel zone, because once that weld is fractured, the panel zone is no longer engaged. Unintuitively to
the authors, there is no strength loss or residual strength defined for this limit state.
For force-controlled components, we defined their strength limits following applicable standards or
references. We followed the acceptance criteria for strength checks per the ASCE 41-17 requirements.
The column splices consist of PJP welds and erection plates. Typical pre-Northridge PJP column splices
with insufficient toughness may experience brittle fracture before reaching their nominal strength, and
therefore, need to be evaluated based on strength limits. We calculated the critical column flange tensile
stresses at splice locations following the guidelines as described in Appendix A of NIST GCR 17-917-46v2
(2017). The column splice P-M-M strength component parameters are defined based on the calculated
critical flange tensile stresses.
We evaluated the existing building for BSE-2E hazard level with CP objective. We considered a loading
pattern proportional to the fundamental modal shape in each horizontal direction for the pushover analysis
lateral load following the ASCE 41-17 requirements. We calculated the target roof displacement in each
direction per the ASCE 41-17 requirements using 5% damping response spectra for the BSE-2E hazard
level. Figure 5 shows the structural capacity curves in both directions with P-delta effects included from
our preliminary pushover analysis.
55
Figure 5. Structural Capacity Curves (in both directions) for the Existing Building
We calculated deformation and force demands at the target roof displacement for selected deformation-
controlled and force-controlled components using the PERFORM-3D analysis model. We compared these
demands with the CP performance level acceptance criteria for the selected components. We concluded
that the structural performance of the existing building does not meet the CP objective under BSE-2E hazard
conditions primarily due to unsatisfactory performance of the panel zone components (deformation-
controlled) and column splice components (force-controlled), indicating seismic retrofit is required for the
building to meet the selected BPOE.
Selected seismic retrofit.
The preferred seismic retrofit approach involves installing piston-type fluid viscous dampers (FVDs) to
protect the existing moment connections and reduce demands on existing columns, column splices, and
column bases. Limited foundation strengthening of parking garage precast shear walls was also required.
Our NDP analysis results using the upper bound damper properties indicated high axial stress in the
columns on the north and south moment frame lines requiring strengthening of some columns. We designed
¼” to ½” thick steel faceplates extending between flanges on the inside face of these columns to address
this issue.
We considered an approach that included enhancing existing beam-column moment connection by

replacing the CJP welds or adding a double plate to the panel zone to achieve better ductility particularly
for the limit state where column kinking may cause premature fracture of the beam to column welds. We
quickly found enhancing the moment connections to not be financially feasible for this project.
FVD design and Nonlinear Time History Analysis
FVD design
Our seismic retrofit uses FVDs provided by Taylor Devices which behave according to the follow equation:
𝐹 = 𝐶𝑁 𝑉 ∝ (Equation 1)
56
where, F = damper force, CN = nonlinear damping coefficient, V = velocity, α = damping constant of 0.3.
Our preliminary FVD design objective was to select a damping coefficient CN to be used in the initial
response history analysis, based on the damper configuration shown in Figures 6 and 7.
Figure 6: FVD Configuration (East and West Elevation)
Figure 7: FVD Configuration (North Elevation on the left, South Elevation on the right)
We made the following assumptions in our preliminary FVD design:
57
1. The fundamental modal shape from the linear dynamic analysis is reasonable for the initial sizing
of dampers used in the nonlinear response history analysis
2. The viscous damping ratio of the structure equals the viscous damping ratio contributed from the
fundamental mode
3. The 1st mode viscous damping is equal to the target effective damping of the structure (βeff) minus
2% of damping in the structural frame
The estimated target roof displacements for the existing building with 5% effective damping at the BSE-
2E hazard level are:
• E-W direction: δ ≈ 55 inches (Δ ≈ 2.25%) with an initial fundamental period of 5.32 seconds
• N-S direction: δ ≈ 37 inches (Δ ≈ 1.51%) with an initial fundamental period of 3.66 seconds
To resolve the panel zone performance issues entirely, we estimated the target drift in the E-W direction
needs to be reduced to 0.9% (equivalent to 22.2 inches of roof displacement), and the target drift in the N-
S direction needs to be reduced to 0.75% (equivalent to 18.5 inches of roof displacement). Theoretically,
these target drifts can be achieved by continuously increasing the effective damping ratios in the system. In
reality, it would not be practical to achieve this goal since any effective damping ratio greater than 30%
would not be effective to further reduce the story drift for this particular structure. As a result, we limit the
effective damping ratio in the system to 25% (including 2% modal damping) in the preliminary design.
We followed the procedures described in MCEER 00-0010 for the preliminary damper design. With 25%
effective damping (23% viscous damping), and assuming the nonlinear damper coefficient, CN, is a constant
for all the devices in each direction, we selected the following damping coefficient for the preliminary
design based on our calculations:
• All the dampers in the E-W direction, CN= 150 (kips-sec/inch)

• All the dampers in the N-S direction, CN= 250 (kips-sec/inch)
Note that the actual viscous damping ratio from nonlinear response history analysis based on the
preliminary damper design may not match the desired viscous damping ratio of 23% as described above.
This is anticipated since the preliminary design was based on the assumptions described in this section
(based on fundamental modal shape from linear analysis).
Nonlinear time history analysis
We used our NSP PERFORM-3D model for our nonlinear time history analysis, with the difference being
that we included PERFORM-3D viscous bar elements as a representation of the FVD and applied the
geotechnical consultant’s site-specific ground motions in accordance with ASCE 41-17. The viscous bar
elements resist only axial load and comprise a viscous damper element and an elastic bar element in series.
ASCE 41-17 Chapter 15 Section 15.3.2 states that the nominal design properties for dampers used as energy
dissipation devices shall be factored by 1.2 for an upper bound analysis and 0.85 for a lower bound analysis
to account for device property variations. Therefore, we performed three sets of analyses (upper bound,
lower bound, and nominal), each comprised of eleven pairs of acceleration-time histories for the BSE-2E
condition.
Our final damper design resulted from an iterative process of checking the overall system response and
component demands for compliance with the BPOE objective. We checked beam moment hinges
58
(deformation check), column P-M hinges (deformation check), panel zones (deformation check), column
compression (strength check), column splices (strength check), and bean moment demand (strength check)
for the collapse prevention limit state considering the BSE-2E time histories. We finalized our FVDs into
two damper sizes split across floors. For all dampers our damping constant is α = 0.3. Figure 8 shows the
dissipated energy in the system for one sample acceleration-time history with maximum total energy of
189,800 kip-in. The energy in fluid viscous dampers accounts for about 70% of the maximum total energy
for this particular time history. Final damper configurations are shown in Figures 9, 10, and 11.
Figure 8. Dissipated Energy in the System for One Sample Time History
59
Figure 9. Damper Design (East and West Elevations)
Figure 10. Damper Design (South Elevation)
60
Figure 11. Damper Design (North Elevation)
Analysis results
The nonlinear time history analysis shows improved structural system response and better compliance with
the BPOE requirements overall. Even so, a very small number of panel zone components still do not meet
the CP performance level acceptance criteria for the BSE-2E condition. In addition, the added dampers
apply high demands to some columns, necessitating some further retrofitting of a subset of those elements.
Across all analyses and all-time histories, all the following components meet the CP performance level
acceptance criteria for the BSE-2E condition:
• Beam moment hinges (deformation check)

• Column P-M hinges (deformation check)
• Panel Zones (deformation check)
• Column splices (strength check)
The panel zone performance is controlled by the lower bound damper property model that has larger story
drifts. Controlling locations are primarily at lower levels in the interior bays of the moment frame lines.
The addition of dampers reduced drift but increased the axial stress ratio in the columns. In ASCE 41-17
Eq. 9-19, the ductility of the panel zone is a function of the axial stress ratio. However, using the maximum
axial stress ratio to determine the panel zone acceptance criteria is too conservative because the maximum
panel zone deformation demands and maximum axial stress ratios are out of phase. Because the dampers
have nonlinear damping properties, they are not perfectly out of phase with the story drift. We determined
the acceptance criteria for critical panel zones at every step of every time history to determine the DCR at
every step of every time history, which was very computationally time-consuming but produced the most
accurate execution of the panel zone evaluation. Although some individual time history analyses show high
demand-to-capacity ratios (DCRs), when looking at the average demand across all time histories, we noted
61
eight (8) panel zones in the North and South elevations that did not quite meet the CP acceptance criteria.
Despite this, we concluded, and the peer reviewer agreed, that the expected global building performance
meets the intent of BPOE.
The upper bound damper properties control the column compression performance. Controlling locations
are the outermost columns of the North and South moment frames, between the 3rd and 10th floors. We
designed faceplates (0.5-in to 0.25-in thick) extending between flanges on the inside face of the column
only to add compression capacity, which reduced the maximum DCR from 1.12 to 0.87.
In addition to accounting for the upper bound and lower bound damper properties in the nonlinear time
history analyses, ASCE 41-17 section 15.3.2 states that the damper component is required to sustain a force
and displacement associated with 130% of the maximum calculated velocity for devices analyzed in the
BSE-2E time history analyses. Therefore, we determined the maximum damper axial forces by averaging
the demand in each damper at each time history across either the upper bound, lower bound or nominal
bound analyses. Using Equation 1, we extracted the corresponding velocity at the maximum demand,
amplified it by 1.3, and recalculated the damper forces (maximum stroke remains the same). The final
dampers supplied by Taylor Devices are sized to accommodate these amplified forces.
For the analysis model, including dampers with nominal properties, we also performed the nonlinear time
history analysis using the eleven pairs of acceleration-time histories for the BSE-1E condition. The results
indicate that all the components meet the LS performance level acceptance criteria.
We chose to ignore the stiffness contribution of the gravity frames as well as the contribution of the slab to
the moment connection capacity. Given that our flexural hinges in the beams representing the moment
connection weren’t yielding, there was no need to consider the contribution of the slab. Had the building in
its pre-retrofitted condition been sufficiently close to meeting the performance objective, we would have
considered the stiffness and strength contributions of the gravity frames.
Advanced pre and post processing techniques were employed on this project similar to the techniques
described in the 2019 SEAOC Convention Paper “Using Computer Automation to Simplify Seismic
Evaluation of Existing Buildings’ by Moore, et. al. We developed programming tools to read modelling
parameter and acceptance criteria information from spreadsheets and rapidly create definitions in the
nonlinear analysis model. Users familiar with PERFORM 3D know that its solver allows for better
computing performance relative to other available nonlinear analysis software, however, tedious inputs can
make executing a large model unwieldy, leading to both time and quality assurance challenges. We
employed programming techniques on this project that effectively eliminated these data inputting
challenges. For post processing, we developed tools to rapidly extract the data from the binary output files
of PERFORM 3D with excel functions. NLTHA inherently have massive amounts of data as calculations
are performed by the software in a model with hundreds of nonlinear elements, over thousands of time
steps, for multiple time history runs. This analysis for this building would be expected to take many months
to complete. Because of the programming tools employed on this project we were able to drastically reduce
the analysis time in the project and meet the owner’s aggressive timeline needs.
62
Photo 2. FVD in Longitudinal Direction of Building
Photo 3. FVD in Transverse Direction of Building
63
Improved Resiliency
Moment frame buildings are often susceptible to substantial damage due to the large peak inelastic story
drift relative to other structures. The high inelastic story drifts trigger damage states with high cost and
repair time consequences. Pre-Northridge Welded Steel Moment Frames often have significant collapse
and excessive residual story drift potential, further increasing their damageability and downtime of the
building on top of safety concerns. In meeting the requirements of the ASCE 41-17 BPOE, we found that
we were drastically reducing the inelastic story drift and seismic base shear with our fluid viscous damper
retrofit. Additionally, we expect the damper’s ability to reduce floor accelerations to further mitigate
damage, particularly to nonstructural components.
To understand how we are improving the resiliency of the building by reducing damageability and
downtime, we performed studies using Haselton Baker Risk Group’s Seismic Performance Prediction
Platform (SP3). We used the Advanced SP3-RiskModel, which allows the user to input Engineering
Demand Parameter (EDPs) directly from the NLTHA used to determine the damage state and associated
consequences for damageable components in the building. The structural response from the NLTHA
inputted into the SP3 model is shown in Figures 12, 13, 14, and 15. The three EDPs used in our
damageability are: (1) Median Peak Floor Acceleration (PFA), (2) Peak Story Drift Ratio (SDR), and (3)
Median Residual Drift (RD). Categories of damageable components include Structural, Exterior Finishes,
Partition Walls, Other Nonstructural, Ceilings, Lighting, Elevators, Piping, and HVAC. We studied the
improved resiliency of the building using the hazards included in our performance objective, events with a
return period of 224 years (BSE-1E) and 975 years (BSE-2E), each with 11 earthquake simulations.
Figure 12. Retrofitted Building Median Story Drift- Longitudinal Direction (from SP3 report)
64
Figure 13. Retrofitted Building Median Story Drift- Transverse Direction (from SP3 report)
Figure 14. Retrofitted Building Median Peak Floor Acceleration- Longitudinal Direction (from SP3
report)
65
Figure 15. Retrofitted Building Median Peak Floor Acceleration- Transverse Direction (from SP3
report)
In its pre-retrofitted condition, the building had a significant collapse risk for the 975-year seismic event,
making it difficult to compare to the performance of the building in its retrofitted condition as safety is the
primary concern for the building. In its retrofitted condition, not only was the collapse risk mitigated, but
the anticipated damageability and downtime in the 975-year event were found to be favorable. The Scenario
Upper Loss (SUL) is less than 20%, and the building downtime is expected to be as low as six months.
For the more frequent event with a 224-year return period, there is a relatively low collapse risk in the pre-
retrofitted building allowing for a comparison of performance to the retrofitted building not skewed by
large additional costs from the risk of collapse or excessive story drift. As expected, the damageability is
drastically reduced in the retrofitted condition, with the SUL approximately three times less than that of the
pre-retrofitted condition. In the retrofitted condition, functional recovery could conceivably occur within
three months while it would be closer to a year in its pre-retrofitted condition.
It should be understood that downtime is also a function of impeding factors such as inspection, financing,
permitting, engineering mobilization, and contractor mobilization. However, the performance of the
structure allows for expedited recovery. Also, our study is based on conservative assumptions on the
damageable components of the building. Even better performance is possible by identifying and mitigating
the key contributing components to damageability and downtime.
Tall Building Design Criteria Discussion
This building meets the definition of a “Tall Building” with the roof height being 205 ft according to the
2020 version of the “An Alternative Procedure for Seismic Analysis and Design of Tall Buildings Located
in the Los Angeles Region” document by the Los Angeles Tall Buildings Structural Design Council (2020
LATBSDC Guidelines). The 2020 LATBSDC Guidelines is generally used for new buildings, however, it
can be a resource for engineers when evaluating existing buildings. In this section we include a discussion
of key items in our project relevant to the 2020 LATBSDC Guidelines.
Serviceability Performance
We used the ASCE 41-17 BPOE for this project which requires the building to be evaluated at two hazard
levels. The LATBSDC recommends that the building remain essentially elastic during a frequent
earthquake (43-year return period). While we did not study serviceability performance for this hazard
(because it is not required by ASCE 41-17), we believe this to be a very practical level of performance and
relatively easy to perform since the structure remains elastic.
Inherent Damping
We used 2% modal damping and no Rayleigh damping in our non-linear analysis that explicitly accounts
for hysteretic damping of structural components and viscous damping from the fluid viscous dampers. We
used 2% based on our judgement and past project experience. ASCE 7-16 requires inherent damping to be
no more than 3% when there are energy dissipation devices. Using the 2020 LATBSDC Guidelines we
could use as much as 2.5% equivalent viscous damping. However, it is unclear if these recommendations
apply to buildings with energy dissipation devices. The 2020 LATBSDC Guidelines requires an analysis
be performed to determine if equivalent viscous damping should be reduced when SSI is explicitly modeled
and provides commentary suggesting that SSI is a significant source of equivalent viscous damping. It does
66
not provide guidance on this analysis, and it is unclear if a similar analysis is required for buildings with
energy dissipation devices.
Peak Transient Drift
We did not have story drift criteria using ASCE 41-17. The 2020 LATBSDC Guidelines provides peak
transient drift limits for the MCER hazard. While we did not study this performance of the building at this
hazard level, our building has little to no reserve capacity for Collapse Prevention at the BSE-2E hazard
level. Given that Collapse Prevention is the performance level for the MCER hazard in the 2020 LATBSDC
Guidelines, it is interesting to see the mean and max story drifts (provided in Figure 16) for this building
that just meets the Collapse Prevention performance. The 2020 LATBSDC Guidelines requires the mean
peak transient story drift to be less than 3% and the maximum peak transient story drift to be less than 4.5%
in each story. This criterion is more likely to control in buildings whose behavior controlled by highly
ductile components. Our building is controlled by the generally non-ductile panel zones, which is why the
median and maximum peak transient story drifts are well below the Collapse Prevention limits provided in
the 2020 LATBSDC Guidelines
Figure 16. Peak Transient Drift at the BSE-2E Hazard
Residual Drift
Similar to peak transient drift, we did not have criteria for residual drift. The 2020 LATBSDC Guidelines
provides residual drift limits for the MCER hazard. The 2020 LATBSDC Guidelines requires the residual
story drift to be less than 1% and the maximum residual story drift to be less than 1.5% in each story. While
our building easily meets the residual drift requirements for mean residual drift, the residual story drift in
some stories is closer to the maximum residual drift limit in the 2020 LATBSDC Guidelines. Had our
analysis shown the building did not meet the residual requirements of the 2020 LATBSDC Guidelines, we
would have strongly considered altering our strengthening approach to meet the 2020 LATBSDC
Guidelines even though it was not explicitly in our design criteria. Large residual drift can be indication of
67
reduced reliability of the analysis, because of how difficult it is to accurately model the behavior of
buildings with large inelastic drift.
Figure 17. Residual Drift at the BSE-2E Hazard
Closing Remarks
This project included the voluntary seismic retrofit of a seventeen-story steel moment-resisting frame
building with FVDs following ASCE 41-17. Even though the changes to panel zone criteria in ASCE 41-
17 resulted in substantially more supplemental damping than the previous ASCE 41-13 criteria, this case
study shows that fluid viscous dampers are an effective way to protect flexible structures. The additional
damping introduced into the structural system reduces drifts and protects elements susceptible to large
displacements without requiring significant strengthening of foundations, column bases, or diaphragms.
Beyond generally meeting BPOE criteria, the building’s expected damageability and downtime are also
significantly reduced. With a continued reduction in computing cost, it is feasible to effectively process
large amounts of data using automated tasks generated by commercial or proprietary applications to assist
in the successful analysis and design of seismic retrofit projects.
68
Photo 4. Completed Building (Note FVDs Visible in Windows)
Acknowledgments
In addition to the author list, we would like to acknowledge the following SGH team members for their
contributions to the analysis and design effort on the project: Ron Hamburger, Craig Goings, Joseph
Moody, Ayush Singhania, Austin Zhang, Molly Pobiel, and Gali Voss de Bettancourt.
Additionally, we would like to acknowledge the following project partners for their contributions to the
success of the project: Marshall Property & Development, Wolcott Architecture, Wood, Howard Building
Corporation, and Taylor Devices.
We also would like to thank the project’s independent peer-reviewers: FARZAD NAEIM, Inc. and Shannon
& Wilson, who provided thoughtful comments and ideas in a very efficient manner; as well as the City of
El Segundo Building Department.
69
References
American Society of Civil Engineers/Structural Engineering Institute (2017), “Seismic

Rehabilitation of Existing Buildings,” ASCE/SEI 41-17, Reston, VA.
American Society of Civil Engineers/Structural Engineering Institute (2016), “Minimum Design

Loads for Buildings and Other Structures,” ASCE/SEI 7-16, Reston, VA.
American Concrete Institute (2014), “Building Code Requirements for Structural Concrete (ACI
318-14) and Commentary,” ACI 318-14, Farmington Hills, MI.
American Institute of Steel Construction (2016), “Specification for Structural Steel Buildings,”
ANSI/AISC 360-16, Chicago, IL.
Computers and Structures, Inc. (2018), “Components and Elements for PERFORM-3D, Version
7.”
Los Angeles Tall Buildings Structural Design Council (2020), “An Alternative Procedure for
Seismic Analysis and Design of Tall Buildings Located in the Los Angeles Region,” Reston, VA.
Kim, D.W., Blaney, C. and Uang, C.M. (2015), “Panel Zone Deformation Capacity as Affected by
Weld Fracture at Column Kinking Location,” Engineering Journal, AISC, First Quarter, 2015, pp.
27-46.
Moore, K., Ranchel, P., Singhania, A. (2019), “Using Computer Automation to Simplify Seismic
Evaluation of Existing Buildings,” SEAOC 2019 Convention Proceedings.
National Institute of Standards and Technology (2017), “Guidelines for Nonlinear Structural
Analysis for Design of Buildings. Part IIa – Steel Moment Frames,” NIST GCR 17-917-46v2.
Ramirez, O.M., et al, (2000), “Development and Evaluation of Simplified Procedures for Analysis
and Design of Buildings with Passive Energy Dissipation Systems,” Technical Report MCEER-00-
0010, University of New York at Buffalo, Buffalo, NY 14260.
70
71
NEW NONLINEAR MODELING PARAMETERS AND ACCEPTANCE CRITERIA

FOR RC STRUCTURAL WALLS
Saman A Abdullah
University of California, Los Angeles, and University of Sulaimani, Kurdistan, Iraq
John W Wallace
University of California, Los Angeles
Abstract
The nonlinear backbone modeling parameters and acceptance criteria in ASCE/SEI 41-17 and 369-17 for
structural concrete walls were developed based on limited experimental data and knowledge available in
the late 1990s (FEMA 273/274-1997), with only minor revisions since. As a result, the wall provisions tend
to be, in many cases, inaccurate and conservative, and thus can produce uneconomical retrofit schemes.
This study utilizes available experimental data and new information on performance of structural walls to
develop updated modeling parameters and acceptance criteria for flexure-, diagonal shear-, and shear-
friction-controlled walls. To accomplish these objectives, a recently developed comprehensive wall
database was utilized, which currently contains detailed information and test results from more than 1100
wall tests. The proposed updates include a new approach to identify expected wall dominant behavior
(failure mode), cracked and uncracked flexural and shear stiffness values, and updated backbone modeling
parameters and acceptance criteria. The updates, which either have been approved or are currently being
balloted for adoption in ACI 369 (and for eventual inclusion into ASCE 41-23), are expected to significantly
contribute to the practice of seismic evaluation and retrofit of wall buildings.
Introduction
The ASCE/SEI 41-17 standard (and other similar documents, e.g., ACI 369-17) represents a major advance
in structural and earthquake engineering to address the seismic hazards posed by existing buildings and
mitigate those hazards through retrofit. For seismic evaluation of existing buildings, these standards provide
nonlinear backbone modeling parameters (e.g., effective stiffness values, deformation capacities, and
strengths), as well as define acceptance criteria (linear and nonlinear) to determine adequacy for a specified
hazard. The modeling parameters and acceptance criteria for reinforced concrete (RC) structural walls were
developed based on limited experimental data and knowledge available in the late 1990s (FEMA 273-97),
with only minor revisions in Supplement #1 in 2007 (Elwood et al., 2007). As a result, the wall provisions
tend to be, in many cases, inaccurate and conservative, and thus can produce uneconomical retrofit schemes.
This study involves utilizing extensive experimental data and new information to develop modeling
parameters and acceptance criteria for concrete structural walls that produce improved seismic assessments
of wall buildings. To accomplish these objectives, a recently developed comprehensive wall database,
known as UCLA-RCWalls, was utilized, which currently contains detailed information and test results from
more than 1100 wall tests surveyed from more than 260 programs reported in literature (Abdullah, 2019;
Abdullah and Wallace, 2019). The database includes three major clusters of data: 1) information about the
test specimen, test setup, and axial and lateral loading protocols, 2) analytically computed data, e.g.,
moment-curvature relationships (depth of neutral axis, c, nominal moment, Mn, first yield moment, My,
72
curvature at Mn, fn, first yield curvature, fy) and wall shear strengths, and 3) test results, e.g., backbone
relations and failure modes.
The proposed updates include: (a) an approach to identify expected wall dominant behavior (failure mode),
and (b) modeling parameters and acceptance criteria (linear and nonlinear) for flexure-controlled, diagonal-
shear-controlled, and shear-friction-controlled structural walls. The updated backbones are as shown in
Figure 1. Unlike the current modeling parameters in ASCE 41-17, the updated values generally represent
median values of the experimental data, with reported statistics (standard deviation and coefficient of
variation). The updates, which have either been approved or are currently being balloted by ACI Committee
369 for adoption in ACI 369-23 (and for eventual inclusion into ASCE 41-23), are expected to significantly
contribute to the practice of seismic evaluation and retrofit of wall buildings. In this paper, only limited
updates and results are briefly presented.
(a) Flexure-controlled walls (b) Diagonal shear-controlled walls (c) Shear friction-controlled walls
Figure 1. Proposed updated backbone curves for concrete structures walls.
Failure Mode Classification
The shear and flexural behaviors of a structural wall are accounted for in a lumped plasticity model using
shear (translational) and flexural (rotational) springs, respectively. In a nonlinear analysis, these springs
will exhibit either linear or nonlinear behavior depending on the dominant wall behavior mode. Therefore,
it is important to quantitatively distinguish between flexure-controlled (generally slender) and shear- or
shear-friction-controlled (generally low-rise or squat) walls/piers. ASCE 41-17 Tables 10-19 and 10-20
include “components controlled by flexure” and “components controlled by shear” in the table captions,
respectively, but the standard does not provide the user with an approach to determine whether a wall is
controlled by flexure or shear. The commentary of ASCE 41-17 (C10.7.1) defines slender and squat walls
as walls with aspect ratio (hw/lw) ≥ 3.0 and ≤ 1.5, respectively, and walls with intermediate aspect ratios
are defined as flexure-shear-controlled walls. However, results presented by Abdullah (2019) show that
shear span ratio (heff/lw), which is similar to hw/lw, is not a good indicator of the expected wall dominant
behavior and failure mode. Therefore, an approach based on the shear-to-flexure strength ratio (VnE/V@MnE)
is proposed based on an evaluation of failure modes in the database, as shown in the subsequent paragraphs.
The reported failure modes in the database (about 1000 wall tests, excluding walls that failed due to
inadequate lap-splices and walls not tested to failure) are presented in Figure 1, where Vn is the minimum
of the nominal diagonal shear strength (Vn,d) and nominal shear-friction strength (Vn,f) calculated based on
ACI 318-19 using tested material properties, V@Mn is the wall shear demand corresponding to the
development of Mn computed based on the shear-span-ratio used in the test and tested material properties,
and V@test is the peak shear strength obtained during the test. Figure 1 (a) indicates that the vast majority of
flexure- and shear-controlled walls have a shear-to-flexure strength ratio (Vn/V@Mn) > 1.0 and < 1.0,
respectively. Walls with failure modes reported as flexure-shear are mainly scattered between 0.7 < Vn/V@Mn
< 1.3. The flexure-shear-controlled walls with Vn/V@Mn < 1.0 generally have limited flexural nonlinearity
(i.e., barely experiencing first yield of longitudinal reinforcement) and, therefore, could reasonably be
classified as shear-controlled walls. On the other hand, for the flexure-shear-controlled walls with Vn/V@Mn
73
> 1.0, the behavior is initially governed by flexural cracking and yielding similar to flexure-controlled walls
because Vn is initially greater than V@Mn, but the wall shear strength gradually reduces, as the wall is cycled
through large nonlinear displacement excursions, until it drops below V@Mn, and then the wall fails in shear.
Depending on the level of shear and flexural demands, these walls could exhibit deformation capacities
comparable to those of flexure-controlled walls (e.g., Tran and Wallace, 2015). Figure 1 (a) also reveals
that the mean of the peak strengths obtained during the test for the flexure-controlled walls is approximately
1.15 times the shear corresponding to the development of Mn. A similar conclusion can be observed for
shear-controlled walls. An alternative presentation of failure modes is given in Figure 1 (b), where the Y-
axis is the shear friction strength (Vn,f) normalized by the diagonal shear strength (Vn,d). It can be seen that
the data are divided between three regions: 1) blue region: flexure-controlled walls with Vn/V@Mn > 1.0, 2)
red region: diagonal shear-controlled walls (due to failure of diagonal tension or compression strut) with
Vn/V@Mn ≤ 1.0 and Vn,f/Vn,d ≥ 1.0, and 3) yellow region: sliding shear-controlled walls with Vn/V@Mn ≤ 1.0
and Vn,f/Vn,d < 1.0.
Table 1 presents the proposed wall failure mode criteria, where ωv is the dynamic shear amplification due
to higher mode effects per ACI 318-19. Table 2 compares the predicted (using Table 1) failure modes of
the wall tests with observed failure modes of the tests and shows that the proposed approach in Table 1
accurately captures the predominant behavior and failure mode of walls.
Figure 2. Wall failure modes results from a dataset of 1000 wall tests.
Table 1. Criteria for determining the expected wall dominant behavior
74
Table 2. Predicted (using Table 1) versus experimental failure modes.

Estimated
Estimated Estimated
Shear- SUM
Flexure Shear
Friction
Experimental # of Tests 489 33 2 524
Flexure Percentage 93.3% 6.3% 0.4% 100%
Diagonal Shear Percentage 0.8% 95.3% 4.0% 100%
Shear-Friction Percentage 4.3% 15.9% 79.7% 100%
Flexure-Shear Percentage 37.0% 57.5% 5.5% 100%
SUM 568 400 78 1046
Flexure-Controlled Walls
The UCLA-RCWalls database was filtered to obtain a dataset of 188 “Conforming Walls” that generally
satisfied the detailing requirements of ACI 318-14 §18.10.6.4 for special boundary elements and a dataset
of 256 “Non-Conforming Walls”. Detailed information on the two datasets is reported by Abdullah (2019).
The datasets were first used to evaluate the current modeling parameters of ASCE 41-17 and then to propose
updated modeling parameters and acceptance criteria. Figure 3 compares Modeling Parameter a from
ASCE 41-17 (plastic hinge rotation at 20% lateral strength loss) with test data from the two datasets. Two
primary observations result from a review of Figure 3: 1) the current Modeling Parameter a constitutes a
conservative lower-bound estimate of deformation capacity of Conforming and Non-conforming walls, and
2) the predictor variable does not correlate well with Parameter a and thus
produces large dispersions.
(a) Conforming walls (b)Non-conforming walls

Figure 3. Evaluation of Parameter a given in ASCE 41-17 versus test data.
As noted, the current ASCE 41-17 nonlinear deformation-based modeling parameters (i.e., Parameters a
and b) are given as plastic hinge rotations. Where a lumped plasticity model is used, typically at or near the
base of a wall, the hinge behavior is modeled as a near-rigid spring with effectively no elastic deformation.
However, in the proposed updates, the deformation-based modeling parameters are given as total hinge
rotation capacities (Figure 1a), which include both the elastic and plastic deformations of the hinge region
75
(lw/2). This approach is proposed because: 1) Modeling parameters are not sensitive to approaches (or
assumptions) used to calculate yield rotation, θy, 2) Modeling parameters are consistent with the total drift
ratio or chord rotation used to define modeling parameters for shear-controlled walls and coupling beams,
respectively, and 3) Modeling parameters can be converted to strain limits by dividing these rotation values
by an assumed hinge length. Strain limits are convenient where fiber models are used, which is becoming
increasingly popular in engineering practice.
The two datasets were studied extensively to identify parameters that have moderate to significant influence
on each modeling parameter on the backbone relation of Figure 1(a). Based on the results, two sets of
modeling parameters and acceptance criteria are proposed, one for walls with conforming “or special”
detailing and the other for walls with non-conforming “or ordinary” detailing, as shown in Table 3 and
Table 4, respectively. The proposed modeling parameters produce low dispersions (coefficient of variation
ranging from 0.18 to 0.25), as shown in Table 5 and Table 6.
Table 3. Proposed modeling parameters and numerical acceptance criteria for Conforming
reinforced concrete structural walls and associated components controlled by flexure
Acceptance Criteria
Conditions e
Performance Level
𝑤𝑤 𝑉𝑉 𝑐𝑐 dnl
𝑙𝑙𝑤𝑤 𝑐𝑐𝐷𝐷𝐷𝐷 𝑣𝑣 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 Overlapping
a IO
2
𝑏𝑏𝑠𝑠 ′
𝐴𝐴𝑐𝑐𝑐𝑐 �𝑓𝑓𝑐𝑐𝑐𝑐 hoops used?
≤ 10 ≤4 Yes 0.032
≤ 10 ≥6 Yes 0.026
≥ 70 ≤4 Yes 0.018
≥ 70 ≥6 Yes 0.014
θyE + 0.1(dnl - θyE)
≤ 10 ≤4 No 0.032
≤ 10 ≥6 No 0.026
≥ 70 ≤4 No 0.012
≥ 70 ≥6 No 0.011
Acceptance Criteria
Conditions e
Performance Level
𝑁𝑁𝑈𝑈𝑈𝑈 cnl c'nl d'nl b enl b
𝑙𝑙𝑤𝑤 𝑐𝑐𝐺𝐺𝐺𝐺
′ LS CP
𝑏𝑏𝑠𝑠2 𝐴𝐴𝑔𝑔 𝑓𝑓𝑐𝑐𝑐𝑐
≤ 10 ≤ 0.10 0.5 0.036 0.040
≤ 10 ≥ 0.20 0.1 0.030 0.032
1.15 0.75 enl 0.85 enl
≥ 70 ≤ 0.10 0.0 0.018 0.020
≥ 70 ≥ 0.20 0.0 0.014 0.014
a
Overlapping hoop definition shall be per ACI 318-19
b
Parameters d'nl and e nl shall not be taken smaller than parameter dnl.
c
The shear amplification factor ωv need not be applied if VMCultDE is obtained from nonlinear analyses procedures.
de
Linear interpolation between the values given in the table shall be permitted; however, interpolation between the values
specified for Conforming walls and Non-conforming walls shall not be permitted.
76
Table 4. Proposed modeling parameters and numerical acceptance criteria for Nonconforming
reinforced concrete structural walls and associated components controlled by flexure
Acceptance Criteria
Conditionsd,e
Performance Level
dnl
𝑙𝑙𝑤𝑤 𝑐𝑐𝐷𝐷𝐷𝐷
Detailinga,b,c,g IO
𝑏𝑏𝑠𝑠2
≤ 10 0.024
≤ 10 0.019
θyE+0.1(dnl - θyE)
≥ 60 0.010
≥ 60 0.008
Acceptance Criteria
Conditionsd,e,g
Performance Level
𝑁𝑁𝑈𝑈𝑈𝑈 cnl c'nl d'nlf enlf,h
𝑙𝑙𝑤𝑤 𝑐𝑐𝐷𝐷𝐷𝐷
′ LS CP
𝑏𝑏𝑠𝑠2 𝐴𝐴𝑔𝑔 𝑓𝑓𝑐𝑐𝑐𝑐
≤ 10 ≤ 0.10 0.4 0.032 0.035
≤ 10 ≥ 0.20 0.1 0.020 0.021
1.15 0.75 enl 0.85 enl
≥ 60 ≤ 0.10 0.0 0.015 0.015
≥ 60 ≥ 0.20 0.0 0.010 0.010
a
Ash,required should be as calculated per ASCI 318-19 Chapter 18. In case of boundary elements with transverse reinforcement in
the form spiral or circular hoop, the term Ash,provided/Ash,required should be replaced with ρs,provided/ρs,required, where ρs,required is
calculated per ACI 318-19 Chapter 18.
b
If values of both Ash,provided/Ash,required and s/db fall between the limits given in the table, linear interpolation should independently
be performed for both Ash,provided/Ash,required and s/db, and the lower resulting value of parameter dnl should be taken.
c
Values of Ash,provided/Ash,required and s/db should be provided over a horizontal distance that extends from extreme compression
fiber at least cDE/3.
d 𝑙𝑙 𝑐𝑐
This table applies to walls and wall segments with ρlw ≥ 0.001. For 0.0025 ≥ ρlw ≥ 0.001 and 𝑤𝑤𝑏𝑏2𝐷𝐷𝐷𝐷 ≤ 20, modeling parameters
𝑠𝑠
𝑙𝑙𝑤𝑤𝑐𝑐𝐷𝐷𝐷𝐷
dnl, d’nl and enl should be multiplied by a reduction factor. The reduction factor shall be 0.4 for ρlw = 0.001 and ≤ 10 and
𝑏𝑏𝑠𝑠2
𝑙𝑙𝑤𝑤𝑐𝑐𝐷𝐷𝐷𝐷 𝑙𝑙𝑤𝑤𝑐𝑐𝐷𝐷𝐷𝐷
1.0 for ρlw = 0.0025 and = 20. Linear interpolation of the reduction factor with respect to ρlw and is permitted for
𝑏𝑏𝑠𝑠2 𝑏𝑏𝑠𝑠2
intermediate values.
e
This table applies to walls with one or multiple curtains of web reinforcement.
f
Parameters d'nl and enl should not be taken smaller than parameter dnl.
g
Linear interpolation between the values given in the table is permitted; however, interpolation between the values specified for
Conforming walls and Non-conforming walls is not permitted.
h ′
For walls with no boundary transverse reinforcement and NUD > 0.08 𝐴𝐴𝑔𝑔 𝑓𝑓𝑐𝑐𝑐𝑐 , enl and d’nl should be multiplied by 0.8 but should
not be taken less than dnl.
77
Table 5. Statistics of modeling parameters for Conforming reinforced concrete structural walls and
associated components controlled by flexure
Lognormal
Coefficient of
Parameter Mean Median Standard
Variation, COV
Deviation
M CyGE,calculated /MCyGE,experimental 1.01 1.00 0.12 0.12
c' 1.03 1.02 0.10 0.10
c 1.15 0.84 0.97 0.84
d 0.98 0.95 0.17 0.17
d' 1.01 1.01 0.22 0.21
e 1.03 1.01 0.22 0.21
The statistics are for the ratios of estimated-to-experimental values.
*
Table 6. Statistics of modeling parameters for Non-conforming reinforced concrete structural walls
and associated components controlled by flexure
Lognormal
Coefficient of
Parameter Mean Median Standard
Variation, COV
Deviation
M /M 0.97
CyGE,calculated 0.97 0.14
CyGE,experimental 0.14
c' 1.03 0.97 0.15 0.15
c 1.22 1.00 0.95 0.78
d 0.95 0.93 0.22 0.23
d' 1.01 0.97 0.24 0.24
e 1.01 1.02 0.21 0.21
*
Shear-Controlled Walls
The UCLA-RCWalls database was filtered to obtain a subset of shear- and flexure-shear-controlled walls
(i.e., Vn/V@Mn ≤ 1.0) tested under quasi-static, reversed cyclic loading protocols. It is noted that no detailing
criteria were applied to the dataset because detailing variables such as Ash, s/db, and hx are not typically
relevant for shear-controlled walls, i.e., there are no limits placed on these variables in ACI 318-19. Based
on that, a total of 365 wall tests were identified. The dataset was studied extensively to identify parameters
that have moderate to significant influence on each modeling parameter on the backbone relation (Figure
1b). Based on the results, an updated set of nonlinear modeling parameters and acceptance are proposed, as
shown Table 7. The shear deformation capacity at initiation of strength loss of walls controlled by shear, as
represented by dnl, was found to be governed by the shape of the cross-section and the ratio of wall shear
strength to shear demand corresponding to its flexural strength. As the latter ratio approaches unity, the
wall behavior tends to shift from a shear-controlled behavior to a flexure-shear-controlled behavior, i.e.,
the wall experiences inelastic flexural deformation in addition to inelastic shear deformation prior to
initiation of lateral strength loss. Although the dnl values in Table 7 only represent shear deformation, the
higher dnl values in that table for walls with high ratios of shear strength to shear demand incorporate the
inelastic flexural deformation. While the level of applied axial load was not found to significantly influence
the point at initiation of strength loss (dnl), it was found to be a critical parameter for the post-strength loss
behavior and the rate of strength degradation up to axial collapse, as represented by parameters d’nl and enl.
Figure 4 and Figure 5 compare the existing (ASCE 41-17) and updated modeling parameters and acceptance
criteria (Table 7), respectively.
In addition to updated deformation-based modeling parameters, new shear strength equations are proposed
for strength at shear cracking and yielding. Eq. 1 is proposed to calculate cracking shear strength of
78
structural walls and wall segments, VCcrWall, corresponding to Point F in Figure 1b. Linear interpolation
between Eq. 1 (a) and (b) based on Ig_flange/Ig_rect can be used for the cracking shear strength of walls and
wall segments with 1.0 < Ig_flange/Ig_rect < 1.5, where Ig_flange is the gross moment of inertia of the wall section
bounded by the effective flange width about its centroidal axis, neglecting reinforcement, and Ig_rect is the
gross moment of inertia of the rectangular portion of the wall section about its centroidal axis, neglecting
reinforcement.
′
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝛼𝛼𝑐𝑐 𝜆𝜆�𝑓𝑓𝑐𝑐𝑐𝑐 𝐴𝐴𝑐𝑐𝑐𝑐 for rectangular sections Eq. 1(a)
′
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 2𝛼𝛼𝑐𝑐 𝜆𝜆�𝑓𝑓𝑐𝑐𝑐𝑐 𝐴𝐴𝑐𝑐𝑐𝑐 for flanged sections Eq. 1(b)
Where:
𝛼𝛼𝑐𝑐 = 3 for ℎ𝑤𝑤 ⁄𝑙𝑙𝑤𝑤 ≤ 1.5 and 2 for ℎ𝑤𝑤 ⁄𝑙𝑙𝑤𝑤 ≥ 2.0. 𝛼𝛼𝑐𝑐 varies linearly between 3 and 2 for 1.5< ℎ𝑤𝑤 ⁄𝑙𝑙𝑤𝑤 < 2.0.
The yield shear strength of a structural wall or wall segment, VCydWall, corresponding to Point B in Figure
1b is proposed to be determined using Eq. 2.
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318𝐸𝐸
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = �2.0 − 1.10 𝑉𝑉𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
� 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318 ≤ 1.8 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318 Eq. 2
≥ 0.8 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318
Where
′
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318 = �𝛼𝛼𝑐𝑐 𝜆𝜆�𝑓𝑓𝑐𝑐𝑐𝑐 + 𝜌𝜌𝑡𝑡 𝑓𝑓𝑦𝑦𝑦𝑦𝑦𝑦 �𝐴𝐴𝑐𝑐𝑐𝑐 Eq. 3
VCydWall is multiplied by 0.85 where the web transverse reinforcement, 𝜌𝜌𝑡𝑡 , is less than 0.0015.
79
Table 7. Proposed modeling parameters and numerical acceptance criteria for nonlinear
procedures: reinforced concrete structural walls and associated components controlled by sheara
Condition Acceptance Criteria
Cross-section 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318𝐸𝐸 gnl dnlc
IO
shape b
𝑉𝑉𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
≥ 1.0 0.015
Rectangular
≤ 0.5 0.006
0.004 gnl+0.1(dnl - gnl)
≥ 1.0 0.020
Flanged
≤ 0.5 0.009
Condition Acceptance Criteria

𝑁𝑁𝑈𝑈𝑈𝑈 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318𝐸𝐸 d'nld enld
′ LS CP
𝐴𝐴𝑔𝑔 𝑓𝑓𝑐𝑐𝑐𝑐 𝑉𝑉𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
≥ 1.0 0.025 0.03
≤ 0.075
≤ 0.5 0.015 0.02
0.65 enl 0.80 enl
≥ 1.0 0.015 0.015
≥ 0.150
≤ 0.5 0.010 0.010
Condition
Cross-section 𝑁𝑁𝑈𝑈𝑈𝑈 cnle c'nl
shape b ′
𝐴𝐴𝑔𝑔 𝑓𝑓𝑐𝑐𝑐𝑐
≤ 0.10 0.25
Rectangular
≥ 0.15 0.00
1.10
≤ 0.15 0.40
Flanged
≥ 0.20 0.00
a
Linear interpolation between values listed in the table is permitted.
b
Linear interpolation between values listed in the table based on Ig_flange/Ig_rect is permitted for walls
and wall segments between wall and flanged designations with 1.0 < Ig_flange/Ig_rect < 1.5.
c 𝑁𝑁 𝑉𝑉
dnl is taken as 0.005 when 𝐴𝐴 𝑈𝑈𝑈𝑈 ≥ 0.20 and 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶318𝐸𝐸 ≥ 0.8 or when 𝜌𝜌𝑡𝑡 and 𝜌𝜌𝑙𝑙 are less than 0.0015
𝑓𝑓′𝑔𝑔 𝑐𝑐𝑐𝑐 𝑉𝑉 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
𝑉𝑉𝐶𝐶𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊318𝐸𝐸
and ≤ 0.5.
𝑉𝑉𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
d
d’nl and enl should not be taken less than dnl.
e
cnl should be taken as zero where 𝜌𝜌𝑡𝑡 is less than 0.0015.
Figure 4. Comparison of existing and updated values for modeling parameter d and e.
80
Figure 5. Comparison of proposed and existing nonlinear acceptance criteria for IO, LS, and CP.
Shear-Friction-Controlled Walls
Shear friction-controlled walls/interfaces are currently treated as force-controlled elements/actions in

ASCE 41-17, which is overly conservative because these walls are typically more ductile than diagonal
shear-controlled walls. To address this issue, the UCLA-RCWalls database was filtered and a dataset of 70
shear-friction-controlled walls tested under reversed cyclic loading protocols was obtained and studied to
develop a new shear friction strength expression and nonlinear modeling parameters.
The proposed nonlinear modeling parameters for shear-sliding are presented in inches (Table 8) because
the sliding deformation along an interface is a local behavior and is independent of wall height. The
modeling parameter bnl in Table 8 estimates the slip at which lateral strength is lost at an interface. Walls
and wall piers sustaining sliding at an interface are expected to maintain gravity load carrying capacity
beyond the slip defined by bnl. Wall tests used to derive nonlinear modeling parameters in Table 8 were not
tested to axial collapse and, therefore, could not be used to define the expected slip displacement at which
axial collapse occurs. The values of modeling parameter bnl provided in Table 8 are based on experience
and judgement. Under sustained transverse loads such as earth or fluid loads, a sliding interface may become
unstable. In such case, bnl is limited to a’nl. The proposed shear-friction modeling parameters were calibrated
for interfaces along the wall and wall segment height. They are not intended for use with interfaces having
different boundary conditions such as the vertical slab/wall interfaces.
The test results also indicated that shear-friction behavior at an interface is characterized by almost zero
slip along the interface until the yield shear-friction strength is exceeded. Where a lumped-plasticity
translational element is used to simulate shear sliding along an interface, but does not include the effects of
diagonal shear deformations within the wall or wall segment, the load-deformation relationship of the
element should be defined as presented in Figure 1c. Alternatively, if the elastic shear flexibility of the wall
or wall segment is aggregated into the lumped-plasticity translational element used to simulate the nonlinear
shear-friction behavior, the load-deformation relationship of the element should be defined as presented in
Figure 1b using modeling parameters and stiffness values for shear-controlled walls up to point B, and as
presented in Figure 1c beyond point B.
Cold joints that meet ACI 318-19 roughness definition are treated as 1.67 times stronger than their “smooth”
counterparts, while monolithic interfaces are treated as 2.33 times stronger than their “smooth”
counterparts. However, test results (e.g., Figure 6) indicate that under cyclic loading, the effect of surface
roughness is not as significant as specified in ACI 318-19. The results demonstrated that shear-friction
coefficients (µ) at concrete interfaces transferring cyclic shear and moment demands are not significantly
influenced by the type of interface (monolithic vs roughened or cold-joint) and are close to shear-friction
coefficients on the lower end of values for interfaces not cycled (ACI 318-19 §22.9.4.2). The provisions of
ACI 318-19 for shear-friction strength were developed primarily based on results from “push-off” tests
81
under monotonic loading protocols, which differ from wall loading conditions under earthquake demands.
Additionally, the results showed that the 0.2f’c Acv upper limit on shear-friction strength well envelopes the
test data; thus, it is recommended to be retained. However, the upper limit of 800 psi on strength in ACI
318-19 Table 22.9.4.4(e) is not justified by experimental evidence and is recommended to be removed,
regardless of the interface type.
Table 8. Modeling parameters and numerical acceptance criteria for nonlinear procedures:
reinforced concrete structural walls and associated components controlled by shear-friction
a
Conditions Sliding Displacements Strength Ratios Acceptance Criteria
(in.)
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 b
Interface Type anl a'nl bnl c'nl cnl IO LS CP
𝑤𝑤𝑣𝑣 𝑉𝑉𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
Monolithic or ≥ 1.0 0.65 1.30

roughened to ¼ in. 0.50
amplitude ≤ 0.5 0.20 0.40 0.1 anl 0.75 bnl bnl
4.0 in. 1.10
≥ 1.0 0.80 1.60
Other 0.60
≤ 0.5 0.40 0.80
a
Linear interpolation between values listed in the table shall be permitted.
b
The shear amplification factor ωv need not be applied if VMCyDE is obtained from nonlinear analyses procedures.
Table 9. Statistical valuesa for modeling parameters for reinforced concrete structural walls and
wall segments controlled by shear-friction
Lognormal Standard Coefficient of
Parameter Mean Median
Deviation Variation, COV
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 b 1.01 0.97 0.20 0.20
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 c 1.05 1.09 0.26 0.25
c'nl b 1.03 1.07 0.25 0.24
c'nl c 0.99 0.98 0.20 0.21
cnl 1.03 0.90 0.45 0.44
anl 1.06 1.07 0.35 0.33
a'nl 1.02 1.00 0.41 0.41
a
b
For values predicted by Eq. 4
c
For values predicted by Eq. 5
Similar to diagonal shear-controlled walls, a new shear friction strength equation is proposed for shear-
friction strength at yielding, which corresponds to Point B in Figure 1c. The expression, considering shear
transfer across any given plane along wall height, is given in Eq. 4. Figure 7 compares the shear-friction
yield strength predicted by Eq. 4 with values obtained from the dataset.
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = �2.5 − 2.15 𝜔𝜔 � 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 ≤ 1.8 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 Eq. 4
𝑣𝑣 𝑉𝑉𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
≥ 0.8 𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶
′
𝑉𝑉𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝜇𝜇�𝐴𝐴𝑣𝑣𝑣𝑣 𝑓𝑓𝑦𝑦𝑦𝑦𝑦𝑦 + 𝑁𝑁𝑈𝑈𝑈𝑈 � ≤ 0.2𝑓𝑓𝑐𝑐𝑐𝑐 𝐴𝐴𝑔𝑔 Eq. 5
82
Where
μ is the coefficient of shear friction and is taken as 0.7 for concrete cast monolithically or placed against
hardened concrete that is intentionally roughened to a full amplitude of approximately 1/4 in, and 0.6
for concrete placed against hardened concrete that is not intentionally roughened.
Shear-friction strength should be evaluated at all possible failure planes along a wall or wall segment height,
such as weak interfaces located at the end of dowel bars, at an existing or potential crack, at an interface
between dissimilar materials, or at an interface between two concretes cast at different times. It is possible
that a construction joint at the foundation-wall interface with dowel bars (μ = 0.6) is stronger than a
monolithic interface (μ = 0.7) at the end of the dowel bars.
In Eq (5), fyfE should not be taken greater than 75,000 psi and should be computed considering reductions
with respect to anchorage. ACI 318-19 Table 20.2.2.4(a) limits yield strength of shear-friction bars to
60,000 psi. This limit was introduced in the 1970s primarily due to lack to data for high strength bars.
Higher strength steel is permitted by other codes, e.g., Grade 500 MPa (72,500 psi) in CSA A23.3-04 and
KCI (2012) and Grade 600 MPa (87,000 psi) in Eurocode 2, Eurocode 8, and fib Model Code (2010). Test
results (Beak et al., 2017) show that at peak strength, limited bar yielding occurs in the web and tension
zone, regardless of the bar grade, which contradicts the ACI 318-19 assumption that all shear-friction bars
yield (up to 60,000 psi) when peak load is attained. However, the ACI 318-19 Eq. 22.9.4.2 Equation
indirectly includes the sliding resistance provided by other mechanisms, e.g., dowel action.
Results reported by Baek et al., (2020) suggest flanges and their longitudinal bars significantly increase the
shear-sliding resistance. Therefore, for flanged wall sections, the reinforcing steel crossing the interface,
including the reinforcement within the effective flange width should be included in 𝐴𝐴𝑣𝑣𝑣𝑣 .
Figure 6. Impact of type of interface and distribution of longitudinal reinforcement on shear-

friction strength (Anoda, 2014).
83
Figure 7. Comparison of estimated and tested shear friction yield strengths.
Summary
This study involves utilizing available experimental data and new information on performance of structural
walls to develop updated modeling parameters and acceptance criteria for seismic evaluation and retrofit
of reinforced concrete structural walls. The proposed updates include: (a) an approach to identify expected
wall dominant behavior (failure mode), and (b) modeling parameters and acceptance criteria (linear and
nonlinear) for flexure-controlled, diagonal-shear-controlled, and shear-friction-controlled structural walls.
Unlike current modeling parameters, the updated values generally represent median values of the
experimental data, with reported statistics (standard deviation and coefficient of variation). The updates,
which have either been approved or are currently being balloted by ACI 369 Committee for adoption in
ACI 369-23 (and for eventual inclusion into ASCE 41-23), are expected to be significant contributions to
the practice of seismic evaluation and retrofit of wall buildings.
Acknowledgments
The work forming the basis for this publication was conducted pursuant to a contract with the Federal
Emergency Management Agency. The substance of such work is dedicated to the public. The author(s) are
solely responsible for the accuracy of statements or interpretations contained in this publication. No
warranty is offered with regard to the results, findings and recommendations contained herein, either by the
Federal Emergency Management Agency, the Applied Technology Council, its directors, members, or
employees. These organizations and individuals 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.
References
Abdullah S. A., 2019, “Reinforced Concrete Structural Walls: Test Database and Modeling
Parameters,” PhD Dissertation, University of California, Los Angeles, CA.
Abdullah S. A. and Wallace, J. W., 2019, “Drift capacity of RC structural walls with special
boundary elements,” ACI Structural Journal, Vol. 116, No. 1, pp. 183–194.
American Concrete Institute (ACI 369-17), 2017. Standard Requirements for Seismic Evaluation
and Retrofit of Existing Concrete Buildings (ACI 369.1-17) and Commentary, Farmington
Hills, MI, 110 pp.
84
Anoda, J., 2014, "Effect of Construction Joint and Arrangement of Vertical Bars on Slip Behavior
of Shear Walls," Master’s Thesis, Nagoya Institute of Technology, Nagoya, Japan. (in
Japanese)
ASCE (American Society of Civil Engineers) (2017): Seismic evaluation and retrofit of existing
buildings. ASCE/SEI 41-17, Reston, VA, 576 pp.
Baek, J., Kim, S., Park, H., and Lee, B., 2020, “Shear-Friction Strength of Low-Rise Walls with
600 MPa Reinforcing Bars,” ACI Structural Journal, V. 117, No. 1, pp. 169-182.
Baek, J., Park, H., and Yim, S., 2017, "Cyclic Loading Test for Walls of Aspect Ratio 1.0 and 0.5
with Grade 550 MPa (80 ksi) Shear Reinforcing Bars," ACI Structural Journal, Vol. 114, No.
4, pp. 969-982.
FEMA (Federal Emergency Management Agency). (1997). “Guidelines to the seismic
rehabilitation of existing buildings.” FEMA 273, Washington, D.C.
Elwood, K. J.; Matamoros, A.; Wallace, J. W.; Lehman, D. E.; Heintz, J. A.; Mitchell, A. D.;
Moore, M. A.; Valley, M. T.; Lowes, L.; Comartin, C.; and Moehle, J. P., “Update of
ASCE/SEI 41 Concrete Provisions,” Earthquake Spectra, V. 23, No. 3, 2007, pp. 493-523. doi:
10.1193/1.2757714.
Tran, T. A., and Wallace, J. W., 2015, "Cyclic Testing of Moderate-Aspect-Ratio Reinforced
Concrete Structural Walls," ACI Structural Journal, Vol. 112, No. 6, pp. 653-665.
85
ATC-145 Update: Draft Guideline for Post-Earthquake Assessment, Repair, and Retrofit of
Buildings
K.J. Elwood
University of Auckland, New Zealand
J.P. Moehle
University of California, Berkeley
Abstract
Guideline for Post-Earthquake Assessment, Repair, and Retrofit of Buildings is currently being developed
in the ATC-145 project. When completed, the Guidelines will provide practical guidance and criteria for
assessing and repairing earthquake-damaged buildings; a critical component in the path to recovery. This
paper provides an overview of the Guidelines structure and identifies some aspects of the assessment
process still under development.
Introduction
The Federal Emergency Management Agency (FEMA) is currently funding the development of Guideline
for Post-Earthquake Assessment, Repair, and Retrofit of Buildings (the Guideline or Guidelines) under
the ATC-145 project, Guide for Repair of Earthquake Damaged Buildings to Achieve Future Resilience,
which is managed by the Applied Technology Council (ATC). The Guidelines will provide practical
guidance and criteria for assessing and repairing earthquake damage to buildings. The intended users of
the Guidelines are primarily practicing engineers with experience in design and construction in seismic
regions. Information in the Guidelines also may be useful to building owners, building officials, insurance
adjusters, and government agencies.
The Guideline will update, extend, and supersede the FEMA 306, 307, and 308 series of documents on
Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings (FEMA, 1999). The FEMA
306 series, developed after the Northridge Earthquake in response to challenges faced in the assessment
of damaged concrete and masonry wall buildings, provided ground-breaking concepts and methodologies
for the assessment and repair of earthquake-damaged buildings. Review of the FEMA 306 series,
however, identified that the series needed significant updating, particularly its emphasis on nonlinear
static procedures and the recommendations linking observed damage and stiffness, strength and
deformation capacity reduction (λ) factors. Furthermore, the Guidelines seek to create a more streamlined
assessment process which can be extended to a wide range of materials and structural systems.
86
Scope
The scope of the Guidelines includes building structures damaged by earthquake ground shaking. The
initial development of the Guidelines will include reinforced concrete frame and wall construction. Later
development may add chapters addressing additional materials and framing systems. The emphasis on
ground shaking is because most damage is due to ground shaking effects. Damage due to surface fault
rupture, foundation settlement, liquefaction and lateral spreading, landsliding, and secondary effects such
as fire following earthquake are not considered.
The Guidelines will include post-earthquake assessment, repair, and retrofit, as identified in the
following:
• Post-earthquake assessment is the process of determining whether an earthquake has negatively
affected the performance capability of a building. It includes the investigation and documentation of
earthquake damage; parallel structural analysis to guide inspection and confirm observations;
classification of damage for building components according to the mode of structural behavior and
severity of damage; and the evaluation of the effects of the damage on the performance of the
building during future earthquakes.
• Post-earthquake repair is the process of evaluating whether repairs are necessary to restoring the
performance capability of the building as well as determining how to implement the repairs. Repairs
are considered in two categories: Safety repairs are those repairs that are primarily intended to restore
safety of a building and that may be mandated by a jurisdiction, while Other repairs are those repairs
that are primarily intended to restore appearance, durability, and serviceability of a building.
• Post-earthquake retrofit increases stiffness, strength, or deformation capacity of a load path, or
changes a load path, and may be done along with repair. The Guidelines will include guidance for
determining the necessity of seismic retrofit of an earthquake-damaged building, but they do not
include procedures for seismic retrofit. Rather, the Guidelines will refer to ASCE/SEI 41, Seismic
Evaluation and Retrofit of Existing Structures, which contains retrofit provisions that can be applied
for buildings evaluated in accordance with the Guidelines.
The Guidelines will contain guidance for assessing the effects of earthquake damage on dynamic response
to future earthquake shaking of varied intensity and, therefore, they may be useful for assessing
performance of building contents and nonstructural components and systems in future earthquakes. The
Guidelines do not, however, provide guidance for assessment, repair, or retrofit of contents or
nonstructural components and systems.
Overview of the Post-Earthquake Assessment, Repair, and Retrofit Process
The post-earthquake assessment, repair, and retrofit process outlined in the Guidelines is triggered by the
occurrence of a damaging earthquake followed by a request or requirement for damage evaluation. The
request or requirement for damage evaluation can be made by an individual owner or by an Authority
having Jurisdiction over the building. See next section for additional discussion on policy considerations
that may affect a damage evaluation. Figure 1 provides an overview of the overall process recommended
by the Guidelines.
87
The request or requirement for a damage evaluation generally will include establishment of Performance
Objectives. A performance objective includes a statement about a target performance level (or levels) for
one or more future earthquake shaking intensities. A Repair Objective refers to the objective to restore
safety, serviceability, and/or other performance capability to pre-earthquake levels without retrofit. A
Repair and Retrofit Objective refers to repair and retrofit to some performance objective exceeding the
pre-earthquake performance capability. These performance objectives may be mandated by public policy
depending on the pre-earthquake safety of the building and the degree of post-earthquake damage.
The next phase of the process is Inspection and Analysis. During this phase, the engineer collects data on
the earthquake and the original building construction, and then conducts an initial visual inspection to
identify earthquake damage. A structural analysis model is constructed to aid in the interpretation of the
earthquake damage and guide more-detailed inspections.
In the next phase, Damage Assessment, the documented earthquake damage is classified as to its severity
and as to whether repair or retrofit is required. Analysis may be required to determine whether the
building is compliant with current or recent building codes and whether the building has sustained
substantial structural damage or disproportionate structural damage; these determinations may then
determine whether it is acceptable to repair the building to restore its pre-earthquake performance
capability or whether repair and retrofit are required.
In the final phase of the process, the required repairs, or repairs plus retrofit, are established. Material-
specific procedures are described to guide how repair and/or retrofit can be implemented.
Figure 1 Overall Guidelines process
Policy Considerations and the IEBC
It is not the intent of the Guidelines to establish policy related to post-earthquake evaluation of buildings.
The Guidelines may, however, provide useful guidance to the responsible engineer in implementing post-
earthquake assessment, repair, and retrofit for a building that is subject to post-earthquake requirements
by an Authority Having Jurisdiction. It is also envisioned that the defined process in the Guidelines will
be helpful in the future development of policy.
Post-earthquake damage assessment and repair can occur as a voluntary mitigation action or a triggered
mitigation action. Voluntary mitigation is mitigation undertaken at the discretion of a building owner or
other stakeholder, and may be driven by concerns over safety and continued building function. A key
point is that the mitigation is voluntary with respect to the Authority Having Jurisdiction, such that there
may be considerable leeway in selecting post-earthquake performance objectives. Triggered mitigation is
mitigation required by a standing regulation when certain triggering conditions, such as the occurrence of
a damaging earthquake, occur. Because triggered mitigation involves compliance with the standing
88
regulation or building code, the regulation or code sets the trigger, the scope of the triggered work, and
the criteria for that work. In this case, the post-earthquake performance objectives may be more narrowly
and precisely defined.
The Guidelines will be developed with the intent that they be applied in tandem with the provisions of the
International Existing Building Code (IEBC), but some modifications to the IEBC may be recommended.
The IEBC contains provisions for post-earthquake repair and retrofit of buildings that may be applicable
in many parts of the United States. The post-earthquake assessment process in the 2021 IEBC is
summarized in Figure 2. A key damage state used in the IEBC provisions is referred to as Substantial
Structural Damage (SSD), defined in most cases as a condition where the vertical elements of the lateral
force-resisting system have suffered damage such that the lateral load-carrying capacity of any story in
any horizontal direction has been reduced by more than 33 percent from its predamage condition.
As shown in Figure 2, in the 2021 IEBC, if the earthquake damage exceeds the definition of SSD, the
requirements are dependent on whether or not the pre-damaged building was complying, where
compliance is defined in Section 304.3.2 as the ability to resist at least 75% of current IBC loads or
ASCE/SEI 41 criteria for reduced seismic forces. Per Section 405.2.3, non-compliant buildings with SSD
require repair and retrofit, but compliant buildings with SSD “shall be permitted to be restored to their
pre-damaged condition”. Engineers have contended the phrase “permitted to be” should not be applied for
a building which has exceeded SSD.
Figure 2 Summary of the post-earthquake assessment process in the 2021 IEBC.
Furthermore, according to analyses conducted as part of the ATC-145 project, the above definition for
SSD (lateral load capacity reduced by more than 33%) may be too permissive. For example, analyses of
Ordinary RC Frame buildings experiencing less than 10% degradation in lateral resistance have been
shown to be susceptible to significant amplifications in drift demand for a repeated earthquake (ATC,
2021). A degradation of 33% in lateral resistance would result in unacceptable amplifications in drift
demand for a repeated earthquake and associated increases in collapse probability. ATC-145 will be
89
exploring alternative definitions for SSD which are consistent with the findings of systems analyses for
concrete frame and wall buildings.
Inspection and Analysis
The Guidelines will provide recommendations for an iterative process of structural analysis and
inspection to form an understanding of the damage condition of the building. The process assumes that
either drawings are available or there is comparable level of knowledge of the building’s construction. As
summarized in Figure 3, the process begins with a preliminary inspection in which data are gathered on
the earthquake and the existing building, and a site visit is conducted to gather initial data on building
damage. Guidance is provided on carrying out a structural analysis of the building under the damaging
earthquake shaking to further focus attention on potential damage areas. A subsequent detailed inspection
guided by the structural analysis is then carried out. A reconciliation of the inspection and analysis results
may be done to understand any discrepancies between observations and calculations and to better refine
understanding about the damage condition of the building. Options are provided for terminating the
inspection and analysis if all indications are that the building did not sustain damage requiring repairs.
Figure 3 Summary of Inspection and Analysis process
A critical step in the Inspection and Analysis process is the identification of possible damage locations.
These locations are identified by comparing the results of an analysis of the building to nearby ground
motion recordings to lower-bound limits for force- and deformation-controlled components. These lower-
bound limits are intentionally conservative to minimize the likelihood of missing critical damage during
Detailed Inspection. The lower-bound limits are selected such that for components with deformation (or
force) demands less than the lower-bound limit, the probability that the component has experienced
degradation of resistance (i.e., deformation demands beyond the capping point) during the damaging
earthquake is less than 10%. This assessment includes uncertainties in the component deformation
capacity, building model, and ground motion. To avoid bias, the Guidelines will also provide median
deformation limits to assess if the observations of damage from the inspection agree with the analysis.
The lower-bound and median deformation limits will be defined as a fraction of the rotation at significant
degradation (parameters a or d) from ASCE 41. The ATC-145 project is collecting databases of
structural component tests to assist in the definition of lower-bound and median limits.
90
Damage Assessment and Repair Type Classification
After the damage condition of the building is established through the inspection and analysis process, the
building damage is classified in terms of its severity and requirements with regard to restoration and/or
retrofit the building are determined. As shown in Figure 4, the process begins with the assignment of
Component Damage Class based on findings of Inspection and Analysis phase.
Visual damage states corresponding to states S1, S2, S3, and S4 in Figure 5 from a representative set of
experiments shall be compared with the observed damage of components in the damaged building to
determine the component damage class (DC). The objective of the visual damage state assessment is to
determine if the component deformation demands in the damaging earthquake likely exceeded the
capping point (S2). Appropriate tests for this comparison are identified from databases and look-up tools
being developed by ATC-145 project. The number of tests required depends upon how closely the test
specimen design parameters and boundary conditions match those of the component in the damaged
building. Key visual damage indicators for RC components include the extent and severity of cracking,
rather than just crack widths, and hence, images of residual damage during tests will be a critical
component to the databases being developed by ATC-145.
If uncertainty exists as to whether the component has exceeded the capping point (S2) after the visual
damage assessment, the deformation demands from analysis can be compared with the median
deformation limits described previously. The priority given to the visual damage assessment over the
analysis results reflects the intent that repair decisions should be driven by critical visible damage in the
components and avoid being influenced by uncertainty in the analysis results.
Bar buckling or low-cycle fatigue (LCF) of longitudinal reinforcement may lead to premature fracture of
reinforcement in earthquakes. Consequently, some consideration is required of the possibility that LCF
has compromised reinforcement in an earthquake-damaged building. For flexure-dominated components,
the Guideline provides a procedure for assessing if the reinforcement may have been impacted by LCF
and requires replacement.
Figure 4 Summary of Damage Assessment process

91
Figure 5 Damage states (S1-S4) and component damage classes (DC0-DC2) in relation to
component backbone. DC2 represents safety-critical damage.
After Component Damage Classes are identified, Figure 6 is used to select the appropriate the Building
Damage Class and associated Repair Type Class. It is proposed that this classification is done based
partly on the severity of the Component Damage Classifications, but also considering whether the
building is Compliant with reduced seismic forces, whether it has sustained Substantial Structural
Damage, and whether it has sustained Disproportionate Structural Damage.
92
Figure 6 Proposed relationship between Component Damage Classes and Building Repair Type
Future work
The Guidelines described in this paper are currently under development. It is anticipated that a final draft
of the Guidelines covering concrete frames and walls will be available for use by engineers by end of
2022. Input on the draft Guidelines will be sought from the engineering community for future
improvements. The Guidelines have been developed such that other materials and systems can be added
to the Guideline relatively easily in future years.
During the current project year, prior to release of the working draft, the Project Technical Committee
will be completing the following tasks:
• Conduct case studies using buildings which have experienced past earthquakes to fine-tune
provisions of the Guidelines;
• Develop improved definition for Substantial Structural Damage and refine damage classification
scheme;
• Develop component deformation limits for shear-controlled walls, beam-column joints, and other
concrete components;
• Develop databases for identifying visual damage states for reinforced concrete components; and
• Develop modeling parameters and acceptance criteria for repaired concrete components.
Acknowledgements and Limitations
The work forming the basis for this publication was conducted pursuant to a contract with the Federal
Emergency Management Agency (FEMA). The substance of such work is dedicated to the public.
ATC-145 is an ongoing project, and conclusions reached to date are subject to change. The authors are
solely responsible for the accuracy of statements or interpretations contained in this publication. No
warranty is offered with regard to the results, findings and recommendations contained herein, either by
the Federal Emergency Management Agency, the Applied Technology Council, its directors, members or
employees. These organizations and individuals do not assume any legal liability or responsibility for the
accuracy, completeness, or usefulness of any of the information, product or processes included in this
publication.
The authors acknowledge the contributions of the entire ATC-145 project team: Mike Mahoney (FEMA
Project Officer), Bill Holmes (FEMA Technical Monitor), Jon Heintz and Chiara McKenney (ATC
Project Managers), Nic Brooke, Abbie Liel, Greg Deierlein, Jim Malley, Bill Tremayne, and John
Wallace (Project Technical Committee), Santiago Pujol and James Malley (Project Review Panel), and
Saman Abdullah, Jay Bhanu, Ryo Kuwabara, Donavan Llanes, Kai Marder, Gonzalo Munoz, Polly
Murray, Eyitayo Opabola, Matias Leon, Amir Safiey, Mehdi Sarrafzadeh, Prateek Shah and Tomomi
Suzuki (Project Working Group).
93
References
ATC, 2021, Resilient Repair Guide Source Report: Post-Earthquake Assessment of Reinforced Concrete
Buildings, ATC-145-2, Applied Technology Council, Redwood City, CA.
FEMA, 1999, Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings, FEMA 306,
prepared by the Applied Technology Council for the Federal Emergency Management Agency,
Washington, D.C.
FEMA, 1999, Repair of Earthquake Damaged Concrete and Masonry Wall Buildings, FEMA 307,
Washington, D.C.
FEMA, 1999, Evaluation of Earthquake Damaged Concrete and Masonry Wall Buildings, FEMA 308,
Washington, D.C.
94
Phenomenological Nonlinear Modeling for Performance Based Design of High-Rise Shear Wall
Buildings
Ian McFarlane, P.E., S.E.
Principal
Magnusson Klemencic Associates
Co-Authors: Joey Piotrowski, P.E., S.E. Senior Design Engineering, Magnusson Klemencic Associates;
Kevin Aswegan, P.E., S.E., Associate, Magnusson Klemencic Associates; Juan D. Pozo, Ph.D., Design
Engineer, Magnusson Klemencic Associates; Kristijan Kolozvari, PhD., P.E., Associate Professor,
California State University, Fullerton; John Hooper, P.E., S.E., Director of Earthquake Engineering,
Magnusson Klemencic Associates.
Abstract
While much research has been done on detailed nonlinear modeling techniques, this study is focused on a
nonlinear layered shell approach that captures adequate response for many performance-based design
high rise buildings. The results demonstrate that a simplified modeling approach using nonlinear layered
shells using CSI ETABS Nonlinear can be adequately calibrated to experimental test data for walls and
coupling beams. Then, the corresponding models are validated as flanged wall and coupled wall
assemblies with additional experimental test data. Finally, analytical results are presented to validate this
modeling approach compared to a traditional fiber-based modeling approach using CSI Perform 3D.
Introduction
Nonlinear response history analysis (NLRHA) for design of tall buildings has become common for tall
buildings in regions of high seismicity (for example, following PEER TBI V2 or Los Angeles Tall
Buildings Structural Design Council 2020). These procedures have been developed to result in building
designs that are capable of achieving the seismic performance objectives intended by ASCE 7. This
evaluation requires the use of nonlinear analysis to subject the building to ground motion time histories to
evaluate that the global and component level responses are within acceptable performance limits.
Many buildings following this approach are concrete core wall structures. They may utilize concrete or
steel floor systems, and there may be a supplemental lateral system of planar blade walls or steel braced
frame systems, however, the primary lateral force resisting system is a concrete core wall with coupling
beams in one or both directions.
The complexity of this analysis has increased based on advancements in nonlinear analytical capabilities.
Models have advanced from lumped plasticity models to distributed plasticity models with varying
complexity. The current industry standard practice is to use fiber-based wall sections with coupling
beams modeled with nonlinear shear hinges (such as with CSI Perform 3D). With advancements in
modeling software, and additional analysis demands (more detailed models, and additional quantities of
time history records), there is a need for efficient modeling recommendations that are compatible with the
modest ranges of nonlinearity anticipated in these structures.
95
The buildings in question incorporate walls with the following typical attributes:
• Walls with aspect ratio greater than 2.0

• Low axial load (typically 0.1Agf’c or less)
• Walls are relatively thick as proportioned for shear, either with shear amplification (ACI 318-19
18.10.3.1) or via NLRHA demands
• Longitudinal reinforcement ratio in range of 1% to 3% at seismic base
The wall behavior is typically as follows:
• Predominate yielding mechanism in the coupling beams

• Secondary yielding of longitudinal reinforcement is predicted in the upper half of the wall due to
higher mode effects
• It is typical that longitudinal reinforcement at the seismic base does not exhibit significant
yielding and does not form a traditional plastic hinge mechanism.
• Where strains are determined with current state-of-the-practice, wall concrete compression strain
is typically less than 0.005, and longitudinal tensile strain is typically less than 0.01.
The goal of this study is to validate a relatively simple and practical modeling approach based on the
ETABS nonlinear layered shell element that adequately captures the response of these types of walls
within the range of nonlinearity that is typically encountered.
Nonlinear Modeling Approach
Wall Modeling
The walls are modeled using a layered shell element. This uses a combination of elastic and inelastic
layers to achieve the combined intended result. The inelastic layers are steel reinforcement and concrete
models. The elastic layers represent the shear behavior and out-of-plane bending behavior.
Definition of shell layers: The layered shell consists of four distinct components:
Longitudinal Rebar Layer: Longitudinal reinforcement is lumped into a single membrane layer
for modeling efficiency. The nonlinear steel material model is used for the axial response and
defined based on the overall longitudinal reinforcement ratio.
In-Plane Concrete Layer: The in-plane concrete behavior is modeled as a membrane layer with
nonlinear confined concrete material model for axial response.
Out-Of-Plane Concrete Layer: The out-of-plane concrete behavior is modeled as a plate element
with only linear out-of-plane bending and thickness equal to the nominal wall dimensions.
In-Plane Shear Layer: The in-plane shear behavior is modeled as a membrane layer with elastic
behavior, with the addition of a property modifier to account for wall cracking.
Concrete Material Model: Properties for confined concrete are determined using a curve fit to the Razvi
model for confined concrete as shown in Figure 1. The effect of an unconfined cover region is included
in the creation of the relationship. The definition of the confined concrete material has zero stiffness and
strength in tension. The confined concrete is modeled such that the sectional dimensions of the shear
walls are based on the full wall thickness, while stress/strain properties are based on a weighted average
of the stress/strain curves for the confined and unconfined portions of the relevant wall thickness. While it
is possible to model separate layers for the confined core versus unconfined cover, it was found that this
96
approach did not result in a better match to the experimental results and added computational complexity
to the analysis.
For elastic and inelastic element definitions, expected concrete strengths are used in the determination of
the modulus of elasticity. The modulus of elasticity is calculated based on ACI 318-19, Section 19.2.2.1
as a function of concrete density.
Figure 1 Sample Nonlinear Concrete Properties (Razvi Model)
While the model identified above has the ability to capture post-peak strength degradation, the reality is
that the typical range of response for tall buildings does not exceed the strain associated with the peak
strength. Where strains exceed 0.0015, it is typical to provide intermediate or full section confinement.
Where full confinement is provided, it is typical that strain would not generally exceed 0.005.
Rebar Material Model: The steel reinforcement model uses an elastic perfectly plastic model (E-P-P).
The model is effective in tension and compression with yield strength based on expected material
properties. Given the low-to-moderate tensile strains that are anticipated, a more accurate material model
including strain hardening is not necessary.
Meshing: Wall meshing proportions are selected to be consistent between the calibration models,
validation models, and meshing used in practice on typical walls. In plan, separate shell elements are
meshed for confined and unconfined concrete regions. Vertically, the walls are meshed with one shell
element per story.
Strain Gauges: Strains are typically measured using full story-height gauges.
97
Wall Calibration
The reinforced concrete shear wall modeling parameters have been verified by comparing to the results
observed in testing of reinforced concrete blade walls (Tran and Wallace, 2012).
The meshing of the calibration model is shown in Figure 2. The model is discretized with shell elements
consistent with the boundary element size, and five shells over the model height. A displacement history
is applied that is consistent with the experimental results.
Figure 2 Model Mesh Geometry

The results are indicated in Figure 3. While the initial stiffness of the model exceeds the experimental
results, the peak strength and energy dissipation are well matched.
98
Figure 3 Reinforced Concrete Blade Wall Test (Tran and Wallace, 2012) vs. Modeled Behavior
Next, strains are evaluated to compare the experimental data to the analysis in Figure 4. The results found
that using a strain gauge height of two elements (roughly length of wall divided by 2) provided a
reasonable comparison to the experimental data. If a single element height gauge length is used, the
strain in the analysis would substantially over predict the experimental results.
99
Figure 4: Blade Wall Experimental Strain (Kolozvari et al, 2014) vs ETABS Analysis Strain
The strain results should also be reviewed with respect to the typical range of modeling as described in
the introduction. This modeling approach is therefore adequate to capture well beyond the typical strain
range anticipated (0.005 in compression and 0.01 in tension). It should also be noted that this modeling
approach has been calibrated to experimental results that represent a single critical section and plastic
hinge at the base which is typically far more concentrated nonlinearity than the distributed yield
anticipated in a typical tall building.
Coupling Beam Modeling
The coupling beams are modeled using the chord-rotation model outlined in Section 10.7.2 of ASCE 41-
17. Each beam is modeled as an elastic beam cross-section with a zero-length displacement-type shear
hinge at mid-span.
Coupling Beam Calibration
The concrete coupling beam modeling parameters have been calibrated to the results observed in testing
of diagonally reinforced coupling beams (Naish and Wallace, 2009). The stiffness of the elastic beam
cross-section, the yield strength and ultimate strength of the beams, their corresponding displacements,
and the energy dissipation factors have all been calibrated to the observed behavior as shown in Figures 5
and 6.
V@Mpr noted in the calibration in Figure 5 is defined as follows: 𝑉𝑉@𝑀𝑀𝑝𝑝𝑝𝑝 = 2𝐴𝐴𝑠𝑠 1.25𝐹𝐹𝑦𝑦 sin 𝛼𝛼
100
Figure 5 Diagonally Reinforced Coupling Beam Test vs. Modeled Behavior
Figure 6 Diagonally Reinforced Coupling Beam Test vs. Modeled Behavior
101
Similar beam calibrations have been successfully performed with conventionally reinforced concrete
beams, and steel reinforced beams. These calibrations show a similar match to experimental data,
although they are not used for this validation study.
Validation with Experimental Results
Using the calibrations previously described, the model is then validated using additional experimental
results. Analytical models were built to replicate experimental tests of flanged walls, planar coupled
walls, and coupled flanged walls as a means to evaluate the adequacy of the overall modeling approach.
Flanged Wall
A U-shaped flanged wall was tested with bi-directional loading protocol (Beyer et al, 2008). The
analytical model was developed to match the experimental test, except that the loading protocol was
simplified to be purely uni-directional rather than simultaneous bi-directional loading. The analysis was
completed for both north-south loading and east-west loading in sequential load cases as shown in Figures
7 and 8.
The north-south analysis matched well with the test in terms of overall strength and shape of the response.
The east-west analysis was not as well matched, which is likely due to the bi-directional deformation
history in the experiment compared to the more simplified analysis. Overall, the modeling showed
reasonably good match to the experimental results for loading in both directions.
Figure 7: U-Shaped Wall Test, North-South Loading (Experiment vs ETABS NL)
102
Figure 8: U-Shaped Wall Test, East-South Loading (Experiment vs ETABS NL)
Coupled Planar Wall

A series of coupled planar wall tests (Santhakumar, 1974) were used to evaluate the ability of the
modeling approach to replicate the global response of a coupled wall assembly. The ETABS NL layered
shell model is also compared with a more detailed SFI-MVLEM OpenSees model (Kolozvari et al, 2018).
The results in Figure 9 show that the ETABS NL model is capable of closely matching the ultimate
strength and general hysteretic behavior of both the experimental results and a more detailed analysis
model.
103
Figure 9: Coupled Wall Test – Experimental vs Analytical Results (from Kolozvari 2018), with ETABS
NL results overlaid
Coupled Flanged Wall
A coupled flanged wall is most representative of the configuration typically utilized for tall performance-
based design towers. Testing of a coupled flanged wall (Sugaya et al, 2000) was evaluated as part of the
FEMA P695 analysis for coupled walls (Tauberg et al, 2019). This evaluation included an OpenSEES
MVELM modeling approach to replicate the experimental response. The ETABS NL layered shell
approach was used to replicate this experimental data, as show in Figure 10. The results show the
ETABS NL approach replicates the global response very favorably.
104
Figure 10: Flanged Coupled Wall Test – Experimental vs Analytical Results (from Tauberg 2019) with
ETABS NL results overlaid
Validation with Analytical Results

As an additional validation, a representative building model was compared between a traditional
modeling approach (CSI Perform 3D model) and the layered shell approach (CSI ETABS Nonlinear).
The selected model is a 40-story center-core point tower as shown in Figure 11. The results below
compare general building dynamics and selected results from one ground motion analysis.
105
Figure 11: CSI Perform 3D Model (left), CSI ETABS Nonlinear Model (right)
Mode shapes and participating mass ratios were compared between models to evaluate any
differences in overall building dynamics. Modal periods as shown in Table 1 are very close. Mass
participation ratios (not shown) also demonstrated an acceptable match between models.
Table 1: Comparison of Modes (Perform 3D Model vs ETABS NL Model)
Mode Perform3D Mode (s) Etabs NL Mode (s)
1 3.83 3.81
2 3.72 3.69
3 3.04 3.01
4 0.98 0.97
5 0.94 0.93
6 0.77 0.76
Next, story shear and overturning moment were compared between models based on a single
ground motion analysis. As shown in Figure 12, small differences were observed between models but the
overall the comparison is favorable.
106
Figure 12: Core Shear and Overturning Moment Comparison (Perform3D vs ETABS NL)
Maximum building drift is also compared for a single ground motion in Figure 13. Although small
differences appeared, the general behavior and peak values were nearly identical and therefore the
comparision is favorable.
Figure 13: Building Drift Comparison (Perform 3D vs ETABS NL)
107
Finally, displacement of a single roof joint is compared over the time history in Figure 14. The peak
values also compared favorably. Although the displacement records do become out of sync later in the
time history, given that peak values and overall response is very similar, the overall comparison is still
favorable.
Figure 14: Roof Joint Displacement for Time History (Perform 3D vs ETABS NL)
Summary and Conclusions
The intent of this study was to demonstrate that a modeling approach of nonlinear layered shells can
adequately analyze a performance-based design high rise building within the typical range of response.
The study has shown that experimental data can be closely matched with nonlinear layered shell models
of blade walls and shear hinge modeling of coupling beams in ETABS NL. These calibrations were
further validated with experimental data from flanged walls, coupled planar walls, and coupled flanged
wall tests. Finally, the ETABS NL approach was compared to an industry accepted standard model using
Perform 3D. All of these results justify that the ETABS NL modeling approach is effective for modeling
both global behavior (stiffness, drift, etc) and local behavior (beam rotations and strain) to an appropriate
level of accuracy for industry practice.
References
ACI 318-19 (2019), “Building Code Requirements for Structural Concrete and Commentary”, American
Concrete Institute, Farmington Hills, Michigan
ASCE 7-16 (2016), “Minimum Design Loads and Associated Criteria for Buildings and Other
Structures”, American Society of Civil Engineers, Reston, Virginia
ASCE 41-17 (2017), “Seismic Evaluation and Retrofit of Existing Buildings”, American Society of Civil
Engineers, Reston, Virginia
CSI, “ETABS Building Analysis and Design”, Computers and Structures, Inc. Walnut Creek, California.
CSI, “Perform 3D Performance-Based Design of 3D Structures”, Computers and Structures, Inc. Walnut
Creek, California
Beyer K., Dazio A, Priestley MJN (2008), “Quasi-static cyclic tests of two U-shaped reinforced concrete
walls”, Journal of Earthquake Engineering, 12, 1023–1053.
108
Kolozvari, K., Terzi, V., Miller, R., and Saldana, D. (2018), “Assessment of Dynamic Behavior and
Seismic Performance of a High-Rise RC Coupled Wall Building”, Engineering Structures, 176:606-620.
Kolozavari, K., Tran, T., Orakcal, K., Wallace, J. (2014). “Modeling of Cyclic Shear-Flexure Interaction
in Reinforced Concrete Structural Walls. II: Experimental Validation”, ASCE Journal of Structural
Engineering
Naish, D., Wallace, J. (2009). “Experimental Evaluation and Analytical Modeling of ACI 31-05/08
Reinforced Concrete Coupling Beams Subjected to Reversed Cycling Loading”, University of California
at Los Angeles Structural and Geotechnical Engineering Laboratory.
PEER TBI (2017), “Guidelines for Performance-Based Seismic Design of Tall Buildings”, PEER Report
2017/06, May 2017, Prepared by the TBI Guidelines Working Group, Berkeley, California: Pacific
Earthquake Engineering Research Center, University of California
Razvi, S., Saatcioglu, M. (1999). “Confinement Model for High-Strength Concrete”, ASCE Journal of
Structural Engineering
Santhakumar, A.R. (1974), “Ductility of coupled shear walls”, University of Cantebury, Christchurch,
New Zealand.
Sugaya, K., Teshigawara, M., Kato, M, and Matsushima, Y. (2000). “Experimetnal Study on Carrying
Shear Force Ratio of 12-Storey Coupled Shear Wall”, 12WCEE
Tauberg, N., Kolozvari, K., Wallace, J. (2019). “Ductile Reinforced Coupled Walls: FEMA P695 Study
Final Report”, University of California, Los Angeles.
Tran, T., Wallace, J. (2012). “Experimental Study of Nonlinear Flexural and Shear Deformation of
Reinforced Concrete Structural Walls”, 15WCEE.
109
MULTI-PERIOD RESPONSE SPECTRA
S. Rezaeian1, N. Luco1, and C. Kircher2

1
Research Structural Engineer, U.S. Geological Survey (USGS), Golden, Colorado
2
Kircher and Associates, Palo Alto, California
Abstract
Multi-period response spectra (MPRS) are incorporated in the development of seismic design ground
motions in the 2020 edition of the NEHRP Recommended Seismic Provisions for New Buildings and
Other Structures (2020 NEHRP Provisions) and are approved for adoption in the American Society of
Civil Engineers (ASCE) Standard, Minimum Design Loads and Associated Criteria for Buildings and
Other Structures (ASCE/SEI 7-22). MPRS are incorporated in these design regulations because it was
discovered that the standard spectral shape based on two periods and one reference site class was
substantially understating spectral response in moderately long period structures located on soft soil sites
where ground motion hazard is dominated by large magnitude events. These are the motions that are
relevant to tall buildings in the Los Angeles region and of interest to the Los Angeles Tall Buildings
Seismic Design Council (LATBSDC). The MPRS incorporation updated Chapters 11, 20, 21, and 22 of
the 2020 NEHRP Provisions (a.k.a. FEMA P-2082); changes are described in detail in the commentary of
FEMA P-2082. The MPRS also influenced the development of the 2018 U.S. Geological Survey (USGS)
National Seismic Hazard Model (NSHM) for the conterminous U.S. because valid ground motion models
for all periods and site classes of interest were required. FEMA P-2082 is complemented by the FEMA P-
2078 technical report that provides a procedure for approximating MPRS outside of the conterminous
U.S. This paper presents a condensed version of the relevant sections of FEMA P-2082 and FEMA P-
2078 that would interest the LATBSDC.
Introduction
The U.S. Geological Survey (USGS) computes the design ground motions of the National Earthquake
Hazards Reduction Program (NEHRP) Provisions by combining hazard results from the USGS National
Seismic Hazard Models (NSHMs) with the site-specific design procedures developed by the Building
Seismic Safety Council (BSSC) Provisions Update Committee (PUC). In previous versions of NEHRP
Provisions, these design procedures have consisted of risk-targeted ground motion calculations,
multiplication by maximum-direction factors, deterministic capping, and multiplication by site
amplification coefficients. For the 2020 NEHRP Provisions (BSSC, 2020), also known as Federal
Emergency Management Agency (FEMA) P-2082, the design procedures were updated based on the
recommendations of the BSSC Project 17 committee (BSSC, 2019). One of their recommendations was
to incorporate multi-period response spectra (MPRS), instead of using ground motions from only three
periods (0, 0.2, and 1 s) and one reference site class (BC), and eliminate the site amplification
coefficients, Fa and Fv, from the provisions. As a result, the new USGS hazard models provide ground
motions not only for more periods (i.e., 22 periods ranging from 0 to 10 s), but also for various
hypothetical site classes ranging from hard rock to very soft soil (i.e., eight site classes ranging from A to
E). Values of design parameters and corresponding MPRS are now obtained from the USGS website
(https://doi.org/10.5066/F7NK3C76) for user-specific site locations (given latitude and longitude) and site
class.
Updates to the 2020 NEHRP Provisions design ground motions in the lower 48 states of the conterminous
U.S. (CONUS) come from two main sources: a) recommendations from the BSSC Project 17 committee
to update the design procedures, and b) updates to the modeling of earthquake sources and ground
110
motions in the 2018 USGS NSHM. Major changes to the design procedures recommended by the BSSC
Project 17 committee include: 1) using MPRS, 2) modifying the deterministic capping procedure through
disaggregation of probabilistic hazard, and 3) updating the maximum-direction factors. In this paper, we
focus on the first recommendation to use MPRS and how it influenced the 2018 USGS NSHM
development. Major updates in the 2018 USGS NSHM include: 1) incorporation of new ground motion
models (GMMs) and soil amplification factors in the central and eastern U.S. (CEUS), 2) incorporation of
basin depths from local seismic velocity models in the western U.S. (WUS), which improves estimates of
long-period ground motion amplification in deep sedimentary basins in the Los Angeles, San Francisco
Bay, Salt Lake City, and Seattle regions, 3) relatively minor modifications to WUS and subduction zone
GMMs to maintain consistency across all periods and site classes and to consider the aforementioned
basin depths, and 4) updates to the background seismicity model to include recent earthquakes. The first
three updates were necessary in order to provide MPRS in the CEUS and to provide acceptable long-
period estimates of ground motions in basin locations in the WUS.
For locations in the U.S. states and territories outside of the conterminous U.S. (OCONUS), updated
USGS hazard models that could provide MPRS ground motions were not available for the 2020 NEHRP
Provisions. These locations include Alaska, Hawaii, Guam and the Northern Mariana Islands, Puerto Rico
and the U.S. Virgin Islands, and American Samoa, for which current USGS hazard models only provide
ground motions at three periods (0, 0.2 and 1 s) and one reference site class (BC). For the 2020 NEHRP
Provisions, MPRS design ground motions in these locations were estimated by the USGS based on a
procedure recommended in the FEMA P-2078 technical report (Kircher et al., 2020). This procedure is
based on using site-specific response spectrum shapes that are developed as functions of ground motion
values at 0.2 and 1 s at a reference site class BC, commonly known as parameters SS and S1. FEMA P-
2078 also provides background information on the necessity of incorporating MPRS in design procedures
and summarizes the relevant changes made to Chapters 11, 20, 21, and 22 of the 2020 NEHRP
Provisions. Relevant sections of that report are summarized in this paper, and the reader is encouraged to
refer to the full FEMA P-2078 report as well as FEMA P-2082 and Kircher et al. (2019) for more in-depth
explanations.
Definition of Multi-Period Response Spectra (MPRS)
For a given site, maximum considered earthquake (MCER) multi-period response spectra (MPRS) are
defined as the values of MCER ground motion at 22 response periods (i.e., 0.0 s, 0.01 s, 0.02 s, 0.03 s,
0.05 s, 0.075 s, 0.1 s, 0.15 s, 0.2 s, 0.25 s, 0.3 s, 0.4 s, 0.5 s, 0.75 s, 1.0 s, 1.5 s, 2.0 s, 3.0 s, 4.0 s, 5.0 s, 7.5
s, and 10 s) for each of eight site classes defined in the 2020 NEHRP Provisions (i.e., A, B, BC, C, CD,
D, DE, and E). At every period and site class, MCER is defined by the 5-percent damped, maximum-
direction, risk-targeted, deterministically capped, horizontal response spectral acceleration. The eight site
classes and the associated values of shear-wave velocity, VS30, including those used by the USGS to
develop MPRS, are provided in Table 1.
Table 1. Site classes and associated values of shear wave velocity.

Site Class Shear Wave Velocity, VS30 (fps) USGS2
Name Description Lower Upper Center VS30
Bound1 Bound1 (mps)
A Hard rock 5,000 1,500
B Medium hard rock 3,000 5,000 3,536 1,080
BC Soft rock 2,100 3,000 2,500 760
C Very dense sand or hard clay 1,450 2,100 1,732 530
CD Dense sand or very stiff clay 1,000 1,450 1,200 365
D Medium dense sand or stiff clay 700 1,000 849 260
DE Loose sand or medium stiff clay 500 700 600 185
E Very loose sand or soft clay 500 150
1
Upper and lower bounds, as defined in Table 20.2-1 of 2020 NEHRP Provisions.
111
2
Center of range (rounded) values used by USGS to develop MPRS.
The 22 response periods and eight site classes represent a matrix of 176 (22 × 8) values of MPRS, as
illustrated by the example matrix of deterministic lower limits in Table 2. The response spectrum of the
site of interest would be selected from the MPRS matrix for the site class of interest based on the site-
specific value of VS30 or taken as the maximum of the MPRS of site classes C, CD, and D defined as the
Default site conditions.
The 176 (22 × 8) combinations of response period and site class define the same set of MPRS for all sites
in the U.S. and its territories, whether the site of interest is located in the conterminous WUS or CEUS
where spectral accelerations for all response periods and site classes are currently available, or located in
OCONUS regions (i.e., Alaska, Hawaii, Guam, Puerto Rico, or American Samoa) where available
spectral accelerations are more limited (e.g., only values of SS and S1 in the MPRS matrix are available),
pending future USGS updates. The intent is that earthquake ground motions as characterized by MPRS be
defined in a consistent manner for all U.S. and territorial sites, such that, for example, an engineer in
Alaska or Hawaii would use the same type of seismic criteria and design methods as an engineer in
California or South Carolina.
Table 2. Example matrix of MPRS at 22 response periods and eight hypothetical site classes. The example values of
response spectral accelerations are from Table 21.2-1 of 2020 NEHRP Provisions for the deterministic lower limit.
Period 5%-Damped Response Spectral Acceleration (g)
T (s) A B BC C CD D DE E
0.00 0.50 0.57 0.66 0.73 0.74 0.69 0.61 0.55
0.01 0.50 0.57 0.66 0.73 0.75 0.70 0.62 0.55
0.02 0.52 0.58 0.68 0.74 0.75 0.70 0.62 0.55
0.03 0.60 0.66 0.75 0.79 0.78 0.70 0.62 0.55
0.05 0.81 0.89 0.95 0.96 0.89 0.76 0.62 0.55
0.075 1.04 1.14 1.21 1.19 1.08 0.90 0.71 0.62
0.10 1.12 1.25 1.37 1.37 1.24 1.04 0.82 0.72
0.15 1.12 1.29 1.53 1.61 1.50 1.27 1.00 0.87
0.20 1.01 1.19 1.50 1.71 1.66 1.44 1.15 1.01
0.25 0.90 1.07 1.40 1.71 1.77 1.58 1.30 1.15
0.30 0.81 0.98 1.30 1.66 1.83 1.71 1.44 1.30
0.40 0.69 0.83 1.14 1.53 1.82 1.80 1.61 1.48
0.50 0.60 0.72 1.01 1.38 1.73 1.80 1.68 1.60
0.75 0.46 0.54 0.76 1.07 1.41 1.57 1.60 1.59
1.0 0.37 0.42 0.60 0.86 1.17 1.39 1.51 1.58
1.5 0.26 0.29 0.41 0.60 0.84 1.09 1.35 1.54
2.0 0.21 0.23 0.31 0.45 0.64 0.88 1.19 1.46
3.0 0.15 0.17 0.21 0.31 0.45 0.63 0.89 1.11
4.0 0.12 0.13 0.16 0.24 0.34 0.47 0.66 0.81
5.0 0.10 0.11 0.13 0.19 0.26 0.36 0.49 0.61
7.5 0.063 0.068 0.080 0.11 0.15 0.19 0.26 0.31
10 0.042 0.045 0.052 0.069 0.089 0.11 0.14 0.17
MPRS in the 2020 NEHRP Provisions
During the closing months of the 2015 NEHRP Provisions cycle, a study was undertaken on behalf of the
BSSC PUC to investigate the compatibility of site amplification coefficients, Fa and Fv, with the GMMs
used by the USGS to produce design maps (Kircher & Associates, 2015). This study discovered that the
standard three-domain (two-period) spectral shape defined by the short-period spectral response
acceleration parameter, SDS, the 1-second spectral response acceleration parameter, SD1, and long-period
transition period, TL, is not appropriate for soft soil sites (site class D or softer), in particular where
ground motion hazard is dominated by large magnitude events. Specifically, on such sites, the standard
spectral shape substantially underestimates spectral response for moderate to long period structures. This
112
conclusion necessitated requiring site-specific ground motion analysis for softer soil sites in ASCE/SEI 7-
16 (ASCE, 2016).
At the beginning of the 2020 NEHRP Provisions cycle, Project 17 (BSSC, 2019), a joint committee of
BSSC volunteers and USGS representatives, was charged with formulating rules by which the next-
generation seismic design value maps would be developed for consideration by the PUC. This included
considering the use of response spectral values for a range of periods and site classes, developed directly
from hazard models and by the USGS. These were to replace the three-domain (two-period) spectral
definition. A four-part MPRS proposal, separately addressing MPRS-related changes to Chapters 11, 20,
21, and 22, was then formulated and adopted by the 2020 NEHRP Provisions (BSSC, 2020) and the
ASCE/SEI 7-22 (to be released in 2022). Key changes to each of these four chapters are described in-
depth in Chapter 2 of the complementary FEMA P-2078 (Kircher et al., 2020) report and are summarized
below.
Key Changes to the Seismic Design Criteria of Chapter 11
The seismic design criteria of Chapter 11 now incorporate values of seismic design parameters for a
specific site class, SMS and SM1 (and SDS and SD1 defined as two-thirds of SMS and SM1), derived directly from
site-specific MPRS that include site amplification, spectrum shape, and other site (and source) effects.
Users would obtain values of these and other ground motion data from a USGS web service for user-
specific values of the location (i.e., latitude and longitude) and site conditions (i.e., site class) of the site of
interest. Accordingly, the tables of site amplification coefficients, Fa and Fv, are eliminated and values of
“reference site” (site class BC) parameters, SS and S1, are replaced with appropriate values of SDS and SD1,
wherever SS or S1 appear in the requirements (e.g., to define Seismic Design Category).
As an alternative to the traditional two-period design spectrum, the updated Chapter 11 incorporates site-
specific MPRS in the definition of earthquake ground motions. Like parameters SMS and SM1 (and SDS and
SD1), users would obtain values of site-specific MPRS from a USGS web service for specific values of the
location and site conditions of the site of interest. Site-specific MPRS provide a more refined description
of the frequency content of the ground motions that would be suitable for multi-mode response spectrum
analysis and for developing ground motions for nonlinear response history analysis.
Values of seismic design parameters SDS and SD1 (and SMS and SM1) are calculated by the USGS from the
multi-period design spectrum for the site class of interest in accordance with the requirements of Section
21.4 of the 2020 NEHRP Provisions as illustrated in Figure 1. Figure 1 illustrates the requirements of
Section 21.4 for a hypothetical high seismicity site with soft soil site class DE site conditions (VS30 = 600
fps). In this example, the value of SDS is about 1.03 g (i.e., 0.9 × 1.14 g), and the value of SD1 is about 1.58
g (i.e., (3 s/1 s) × 0.53 g) with a corresponding transition period, TS, of about 1.54 s. Because the site is
soft, the transition period, TS, is greater than 1.0 s. The calculated value of SD1 exceeds SDS, and the design
spectrum is truncated at the design short period spectral response acceleration. The frequency content of
the design spectrum (i.e., two-thirds of the MCER spectrum) of this example reflects the combined effects
of site amplification and spectral shape, both of which contribute significantly to the long-period
frequency content for this soft soil site. Spectrum shape effects were not included in the site amplification
coefficients, Fa and Fv, which necessitated requiring site-specific ground motion analysis for softer soil
sites in ASCE/SEI 7-16. The incorporation of MPRS in 2020 NEHRP Provisions and ASCE/SEI 7-22
eliminates the need for such analyses.
Key Changes to the Site Classification Criteria of Chapter 20
Changes to Chapter 11 required addition of three new site classes (BC, CD and DE) to more accurately
define the frequency content of earthquake ground motions (Table 1). New site classes, including revised
ranges of VS30 values and related site classification criteria, are defined in Chapter 20. A new “Default”
113
site condition is defined as the more critical spectral response of site classes C, CD, and D, for design
where soil properties are not known in sufficient detail to reliably determine the site class.
2.4
MCEr Multi-Period Response Spectrum - Site Class BC
2.2 MCEr Multi-Period Response Spectrum - Site Class DE
Design Multi-Period Response Spectrum - Site Class DE
2.0
Two-Period Design Spectrum - Site Class DE
1.8
Response Spectral Acceleration (g)
1.6
SD1/T = max(Sa1,T x 0.9 x Sa[1s ≤ T ≤ 2s])/T vS30 > 1,200 fps
1.4
max(Sa1, T x 0.9 x Sa[1s ≤ T ≤ 5s])/T vS30 ≤ 1,200 fps
1.2
1.0
0.8
0.6
SDS = Max(0.9 x Sa[0.2s ≤ T ≤ 5s])
0.4
0.2
0.0
0.1 1.0 10.0
Period (seconds)
Figure 1. Example derivation of SDS and SD1 and the new definition of two-period design spectrum from a site-specific
multi-period design spectrum for a hypothetical high seismicity site with soft soil site conditions. For more in-depth
description see FEMA P-2078 (Figure 2.2-2 of FEMA P-2078, first appeared in Kircher et al., 2019).
Key Changes to the Site-Specific Ground Motion Procedures of Chapter 21
Changes to Chapter 21 incorporate the MPRS available from the USGS web service into the site-specific
requirements of Chapter 21 by: 1) permitting their use for design in lieu of those determined by a
traditional site-specific ground motion analysis, and 2) requiring that site-specific ground motions not be
less than those obtained from the USGS web service without peer review (i.e., to provide a lower-bound
safety net for ground motions developed by a site-specific analysis).
Other changes to Chapter 21 are explained in-depth in Chapter 2 of FEMA P-2078 and include: 1)
eliminating the risk coefficient method for determining probabilistic (risk-targeted) MCER ground
motions from uniform-hazard (2% probability of exceedance in 50-year) ground motions, 2) revising the
period-dependent factors required for conversion of geometric mean (RotD50) ground motions to
maximum direction (RotD100) ground motions, and 3) revising the deterministic MCER ground motion
requirements. Furthermore, as a result of incorporating MPRS and eliminating the site amplification
coefficients, the lower limit on the deterministic MCER response spectrum was replaced with a table of
deterministic lower limit MPRS (Table 2). These updates are consistent with the methods used by the
USGS to develop updated values of seismic design parameters and MPRS provided through USGS web
services.
Key Changes to the Seismic Ground Motion Maps of Chapter 22
Three main changes are made to Chapter 22: 1) figures of mapped values of obsolete parameters SS, S1,
and peak ground acceleration (PGA) for site class BC are replaced with figures of mapped values of
parameters SMS, SM1, and PGAM for Default site conditions, 2) a USGS web service for obtaining values of
114
design parameters SMS, SM1, and PGAM for any site class of interest is referenced, and 3) figures of mapped
values of the obsolete risk coefficients, CRS and CR1, are deleted.
All values of design parameters (and corresponding MPRS) would be obtained from the USGS web
service for user-specific values of the site location (latitude and longitude) and site class (including
Default site conditions). The MCER values are developed in accordance with the site-specific
requirements of Chapter 21 of the 2020 NEHRP Provisions, based on the 2018 update of the USGS
NSHM for sites in the CONUS, and following the methods developed in FEMA P-2078 for sites outside
of the CONUS.
MPRS Ground Motions in the Conterminous U.S.
The 2014 USGS NSHM (Petersen et al., 2015) was used to calculate ground motion parameters for the
2015 NEHRP Provisions and ASCE/SEI 7-16 in the CONUS. The USGS updated this model in 2018-
2019 and refers to it as the 2018 USGS NSHM (Petersen et al., 2020; Rezaeian et al., 2021; Powers et al.,
2021). Whereas the 2014 USGS NSHM provided ground motion parameters at three spectral periods and
one reference site class, the 2018 USGS NSHM provides ground motion parameters for all spectral
periods and site classes needed to develop the MPRS matrix (Table 2). The main four updates in the 2018
USGS NSHM for the conterminous United States are summarized as follows:
1) For the conterminous CEUS, 31 new GMMs are used in the 2018 USGS NSHM including the NGA-
East models with a 2/3 weight, and the updated seed models with a 1/3 weight (see Rezaeian et al., 2021,
for definitions and details). NGA-East was a project that gathered data and simulations applicable to the
CEUS and developed models of median ground motions, uncertainties, and site effects. These models
made it possible to produce ground motions in the CEUS for periods from 0 to 10 s and for site classes A
through E. The GMMs used for the 2014 USGS NSHM had a narrow range of periods and site classes.
2) In the Los Angeles, Seattle, San Francisco Bay, and Salt Lake City regions, where published models of
basin depths are deemed applicable (see Powers et al., 2021, for details), basin depths were incorporated
into the 2018 USGS NSHM. The depths are input parameters to the GMMs used in the WUS, although in
the Seattle region, this entailed modification of the GMMs themselves. At sites where the basin depths are
larger than a default that is estimated from site class, ground motions are amplified. The amount of
amplification increases with basin depth and depends on spectral response period and site class. At other
sites where the depths are smaller, or outside of the four regions, the default site-class based basin depth is
assumed, and ground motions are unaffected. The effects of deep basins were not included in the 2014
USGS NSHM or the site amplification coefficients, Fa and Fv, of the 2015 NEHRP Provisions.
3) In the WUS, two of the GMMs that were used in the 2014 USGS NSHM were excluded in the 2018
update due to incorporation of MPRS because one of the two models could not be used for softer site
classes, and the other could not be used for spectral response periods longer that 3.0 s.
4) Outside of California (because the Uniform California Earthquake Rupture Forecast, UCERF3, has not
been modified), the catalog of past earthquakes was updated for the 2018 USGS NSHM. Seismicity
catalogs are used to calculate spatially smoothed rates of occurrence of future earthquakes on unmodeled
(or unknown) faults. In addition to appending earthquakes that occurred in 2013 through 2017, other
relatively minor updates were made to the catalog and the smoothed earthquake rates.
Examples of MPRS at 11 WUS Southern California Sites
Examples of MCER MPRS at all periods and site classes of interest are provided for 34 locations (same
locations as those used in the commentaries of the NEHRP Provisions) in the CONUS in Chapter 6 of the
FEMA P-2078 report. These values were provisional and based on an early version of the 2018 USGS
115
NSHM, which did not include deep basin effects and a few other minor updates to both probabilistic and
deterministic calculations of ground motions. These early versions of MPRS values were used in FEMA
P-2078 to validate the MPRS approximation procedures for OCONUS locations. The validation exercises
still hold regardless of later updates to the 2018 NSHM because they show that the approximate models
are capable of closely estimating any “true” MPRS by matching to their SS, S1, and TL values. The updated
MPRS values (based on the final 2018 USGS NSHM) and figures for 11 of the 34 example sites that are
located in the southern California region are provided in the appendix to this paper. These are the same
values adopted by the 2020 NEHRP Provisions and are the lesser of the probabilistic and the
deterministic (after applying the deterministic lower limit) MCER at every period and site class.
Approximation of MPRS Ground Motions Outside of the Conterminous U.S.
For states and U.S. territories outside of the CONUS, the USGS hazard models have not been updated
with respect to the previous versions that only provided ground motions for three periods and one
reference site class. The FEMA P-2078 report presents a procedure to approximate MPRS from only the
three currently available ground motion parameters SS, S1, and TL for all OCONUS locations. This
procedure provides a more accurate approximation of the frequency content of the response spectrum
compared to the “two-period design spectrum” of the 2015 NEHRP Provisions, which used site
amplification coefficients and, in many cases, has been found to significantly underestimate MPRS at
long periods.
Key Hazard Parameters Influencing Spectral Shape
The key to approximation of MPRS is estimating the frequency content (shape of the MCER response
spectrum) for all possible site conditions only knowing values of SS, S1, and TL. The primary factors that
influence the shape (frequency content) of earthquake ground motion response spectra are: 1) site
conditions, which are characterized by site class based on VS30 and are presumed to be known for the site
of interest, or a Default site condition is assumed, 2) ground motion level, which is characterized by site-
specific values of SS and S1 and influences the frequency content of ground motions due to soil
nonlinearity (i.e., amplitude-dependent response of soil), particularly at shorter periods and softer site
conditions, 3) the earthquake magnitude(s) governing seismic hazard at the site of interest, which
influences the relative strength of short-period and long-period ground motions (i.e., long-period ground
motions are relatively stronger for sites governed by large magnitude earthquakes) and can be estimated
for short and 1-second periods from SS and S1 hazard disaggregations, MS and M1, and 4) spectral response
ratio, RS/1 = SS/S1.
In general, governing earthquake magnitude(s), MS and M1, are not known prior to performing
probabilistic seismic hazard analysis and disaggregation of hazard. Therefore, the effects of earthquake
magnitude on the frequency content of probabilistic ground motions were implicitly taken into account in
this study by defining probabilistic MPRS as a function of the spectral response ratio, RS/1, and the
mapped value of TL (which is related to earthquake magnitude and an indicator of M1). Although similar
to TL, RS/1 is used as a proxy to represent the controlling magnitude at long periods, it also contains
information about the relative amplitude of the ground motion at long compared to short periods. For
development of deterministic MPRS, the governing earthquake magnitude is similarly inferred from the
spectral response ratio, RS/1. Other site and source parameters could also influence the frequency content
of earthquake ground motions, including earthquake mechanism (e.g., shallow crustal events, deep
subduction zone events, stable continental events), fault type (e.g., strike-slip, reverse, normal),
directivity, and basin effects, but, in general, are less influential compared to the parameters mentioned
above. In addition, information from disaggregation of site hazard (e.g., values of ε0) can also be
influential but are generally not available at all response periods and site classes of interest and therefore
were not considered as input parameters.
116
Approximating MCER MPRS Ground Motions from SS, S1, and TL
To describe generic shapes of MPRS, this study defines response spectrum shape parameters (RSSPs) as
the set of MPRS normalized by SS. This normalization by the SS amplitude allows the study to better
isolate and observe the effects of frequency content on the shape of the response spectrum at each site
condition. RSSPs were developed to characterize the frequency content of probabilistic and deterministic
MPRS separately and as functions of the 1) site class, 2) level of ground motion (SS and S1), and 3)
earthquake magnitude (as inferred from RS/1 and TL). Thousands of example sites in California, Oregon,
Washington, Idaho, and Nevada (various TL values) were used to calculate MPRS and develop RSSPs,
resulting in 41 sets of probabilistic RSSPs (Chapter 4 of FEMA P-2078) and 15 sets of deterministic
RSSPs (Chapter 5 of FEMA P-2078) with distinct frequency contents.
Given values of SS, S1, and TL for a site of interest, probabilistic and deterministic MPRS can be separately
approximated by following four basic steps, each explained in detail in Chapter 3 of FEMA P-2078:
Step 1 – Select representative sets of short-period and long-period RSSPs, based on SS, S1, and TL values.
Step 2 – Determine short-period and long-period response domains (i.e., period range).
Step 3 – Calculate (by scaling RSSPs) short-period and long-period MPRS.
Step 4 – Calculate mid-period MPRS (by blending short-period and long-period response).
Once probabilistic MPRS and deterministic MPRS are approximated, the MCER MPRS ground motions
can be derived by first taking the greater of the deterministic MPRS and the deterministic lower limit
(Table 2), and then taking the lesser of those values and the probabilistic MPRS at each response period
and site class.
This procedure is validated in Chapter 6 of FEMA P-2078 for example locations in the CONUS for which
“true” MPRS values were available. Because the RSSPs were developed using data and GMMs from
WUS, approximated MPRS compare very well with “true” MPRS for these locations. This is not the case
for CEUS locations. For a better fit to CEUS MPRS, RSSPs could be developed in a similar way but
using data and GMMs from the CEUS; however, because OCONUS locations are more compatible with
WUS tectonics (crustal or subduction earthquakes), this was not desirable in FEMA P-2078. Similarly,
RSSPs were developed for site-class based default basin depths because it is not desirable to account for
local basin effects when deriving MPRS for a general site in the U.S. and its territories with unknown
basin conditions.
Summary
Multi-period response spectra (MPRS) are incorporated in the development of seismic design ground
motions in the 2020 NEHRP Provisions and ASCE/SEI 7-22. This was done because the old two-period
response spectrum and site amplification factors can substantially underestimate spectral response in
moderately long period structures located on soft soil sites where ground motion hazard is dominated by
large magnitude events. This paper summarized the relevant changes to Chapters 11, 20, 21, and 22 of the
2020 NEHRP Provisions, including elimination of site amplification factors, Fa and Fv, and new
definitions of ground motion parameters and site classes. The incorporation of MPRS influenced the
development of the 2018 USGS NSHM for the CONUS. The MPRS for OCONUS locations were
approximated using the procedures of FEMA P-2078. MPRS values for 11 example sites in southern
California are presented in the appendix.
Acknowledgment
We thank the USGS summer interns, Stephen Waldvogel and Katrina Peralta, for updating the MPRS
tables of the appendix to match the values that were adopted in the 2020 NEHRP Provisions.
117
References
American Society of Civil Engineers (ASCE/SEI 7-16), 2016, “Minimum design loads and associated criteria for
buildings and other structures,” ASCE/SEI 7-16, American Society of Civil Engineers, Reston, Virginia.
Building Seismic Safety Council (BSSC), 2015, “NEHRP recommended seismic provisions for new
buildings and other structures,” Federal Emergency Management Agency (FEMA) Report P-1050-1.
Available at: https://www.fema.gov/sites/default/files/2020-07/fema_nehrp-seismic-provisions-new-
buildings_p-1050-1_2015.pdf (accessed 18 October 2021).
Building Seismic Safety Council (BSSC), 2019, “BSSC Project 17 final report: Development of next
generation of seismic design value maps for the 2020 NEHRP provisions,” National Institute of Building
Sciences, Washington, DC. Available at:
https://www.nibs.org/files/pdfs/NIBS_BSSC_Project17FinalReport_2020.pdf (accessed 18 October 2021).
Building Seismic Safety Council (BSSC), 2020, “NEHRP recommended seismic provisions for new
buildings and other structures,” Federal Emergency Management Agency (FEMA) Report P-2082-1.
Available at: https://www.fema.gov/sites/default/files/2020-10/fema_2020-nehrp-provisions_part-1-and-
part-2.pdf (accessed 18 October 2021).
Kircher & Associates, 2015, “Investigation of an identified shortcoming in the seismic design procedures of ASCE
7-10 and development of recommended improvements for ASCE 7-16,” prepared for the Building Seismic Safety
Council of the National Institute of Building Sciences, Washington, D.C.
Kircher C, Rezaeian S, and Luco N, 2019, “Proposed multi-period response spectra and ground motion requirements
of the 2020 recommended provisions and ASCE 7-22,” SEAOC 2019 convention proceedings, Structural Engineers
Association of California 2019 convention, Squaw Creek, CA, 28–31 August, pp. 5. Sacramento, CA.
Kircher C, Rezaeian S, and Luco N, 2020, “Procedures for developing multi-period response spectra of non-
conterminous United States sites,” Prepared by Applied Technology Council for Federal Emergency Management
Agency (FEMA) Report P-2078, Washington, D.C. Available at: https://www.fema.gov/sites/default/files/2020-
11/fema_p-2078_multi-period-response-spectra_08-01-2020.pdf (accessed 18 October 2021).
Petersen MD, Moschetti MP, Powers PM, Mueller CS, Haller KM, Frankel AD, Zeng Y, Rezaeian S, Harmsen SC,
Boyd OS, Field N, Chen R, Rukstales KS, Luco N, Wheeler RL, Williams RA, and Olsen AH, 2015, “The 2014
United States national seismic hazard model,” Earthquake Spectra 31(S1): S1–S30.
Petersen MD, Shumway AM, Powers PM, Mueller CS, Moschetti MP, Frankel AD, Rezaeian S, McNamara DE,
Luco N, Boyd OS, Rukstales KS, Jaiswal KS, Thompson EM, Hoover SM, Clayton BS, Field EH, and Zeng Y,
2020, “The 2018 update of the US national seismic hazard model: Overview of model and implications,”
Earthquake Spectra 36(1): 5–41.
Powers PM, Rezaeian S, Shumway AM, Petersen MD, Luco N, Boyd OS, Moschetti MP, Frankel AD, and
Thompson EM, 2021, “The 2018 Update of the U.S. National Seismic Hazard Model: Ground motion models in the
western U.S.,” Earthquake Spectra first published online 14 May 2021. doi:10.1177/87552930211011200
Rezaeian S, Powers PM, Shumway AM, Petersen MD, Luco N, Frankel AD, Moschetti MP, Thompson EM, and
McNamara DE, 2021, “The 2018 update of the US National Seismic Hazard Model: Ground motion models in the
central and eastern U.S.,” Earthquake Spectra 37(S1): 1354-1390.
Appendix: Multi-Period Response Spectra at 11 WUS Southern California Sites
This appendix provides tables and plots of MCER MPRS ground motions for 11 U.S. city sites in southern
California calculated for the 2020 NEHRP Provisions and ASCE/SEI 7-22. These values are based on the
lesser of probabilistic and deterministic ground motions, respectively calculated in accordance with the
requirements of Section 21.2.1 and Section 21.2.2 of 2020 NEHRP Provisions, where deterministic
ground motions include the deterministic lower limit of Table 2 in accordance with Section 21.2.3. The
values provided in this appendix are different from those of the FEMA P-2078 report Appendix D
because a preliminary version of the 2018 USGS NSHM was used at the time of writing that report for
demonstration purposes only.
118
Table A1. Multi-period response spectra (MPRS) MCER ground motions at 22 response periods and 8 hypothetical
site classes (and Default site condition) in the 2020 NEHRP Provisions, Los Angeles site (latitude 34.05, longitude -
118.25).
Period 5% -Damped Response Spectral Acceleration (g)

T (s) A B BC C CD D DE E Default
0.00 0.720 0.800 0.930 1.020 1.050 1.000 0.860 0.740 1.050
0.01 0.720 0.810 0.940 1.030 1.060 1.000 0.870 0.750 1.060
0.02 0.750 0.840 0.960 1.040 1.060 1.000 0.850 0.730 1.060
0.03 0.890 0.970 1.080 1.120 1.110 1.020 0.820 0.680 1.120
0.05 1.280 1.370 1.420 1.410 1.310 1.150 0.860 0.710 1.410
0.08 1.650 1.780 1.840 1.800 1.650 1.380 1.070 0.960 1.800
0.10 1.760 1.950 2.110 2.100 1.930 1.590 1.260 1.160 2.100
0.15 1.730 1.970 2.320 2.420 2.290 1.880 1.490 1.360 2.420
0.20 1.530 1.770 2.250 2.550 2.490 2.100 1.580 1.400 2.550
0.25 1.340 1.560 2.040 2.510 2.600 2.320 1.690 1.430 2.600
0.30 1.180 1.390 1.850 2.360 2.630 2.510 1.840 1.540 2.630
0.40 0.960 1.130 1.540 2.070 2.500 2.530 2.050 1.670 2.530
0.50 0.800 0.950 1.320 1.800 2.290 2.420 2.140 1.840 2.420
0.75 0.590 0.680 0.950 1.330 1.750 1.970 2.040 1.910 1.970
1.0 0.450 0.500 0.720 1.020 1.390 1.660 1.840 1.930 1.660
1.5 0.290 0.320 0.440 0.640 0.900 1.160 1.440 1.650 1.160
2.0 0.220 0.230 0.310 0.450 0.640 0.860 1.170 1.430 0.860
3.0 0.140 0.150 0.190 0.270 0.390 0.540 0.750 0.930 0.540
4.0 0.110 0.110 0.130 0.190 0.270 0.370 0.500 0.620 0.370
5.0 0.087 0.091 0.100 0.140 0.200 0.270 0.370 0.450 0.270
7.5 0.059 0.060 0.065 0.087 0.120 0.150 0.200 0.240 0.150
10 0.042 0.043 0.046 0.059 0.077 0.098 0.130 0.150 0.098
Figure A1. Plots of multi-period response spectra (MPRS) MCER ground motions in the 2020 NEHRP Provisions,
Los Angeles site.
119
site classes (and Default site condition) in the 2020 NEHRP Provisions, Century City site (latitude 34.05, longitude -
118.40).

0.00 0.760 0.850 0.990 1.090 1.110 1.040 0.870 0.750 1.110
0.01 0.770 0.860 1.000 1.090 1.120 1.050 0.880 0.760 1.120
0.02 0.800 0.890 1.020 1.110 1.120 1.050 0.860 0.730 1.120
0.03 0.960 1.040 1.150 1.200 1.170 1.070 0.830 0.680 1.200
0.05 1.380 1.470 1.520 1.500 1.380 1.170 0.870 0.730 1.500
0.08 1.770 1.910 1.970 1.910 1.720 1.400 1.090 1.000 1.910
0.10 1.890 2.090 2.250 2.220 2.020 1.610 1.290 1.200 2.220
0.15 1.850 2.100 2.470 2.560 2.390 1.900 1.520 1.410 2.560
0.20 1.630 1.890 2.370 2.700 2.610 2.120 1.600 1.450 2.700
0.25 1.430 1.660 2.170 2.660 2.720 2.330 1.690 1.430 2.720
0.30 1.260 1.480 1.970 2.510 2.770 2.530 1.840 1.530 2.770
0.40 1.020 1.200 1.640 2.200 2.630 2.610 2.080 1.670 2.630
0.50 0.860 1.000 1.400 1.920 2.400 2.520 2.230 1.890 2.520
0.75 0.710 0.810 1.100 1.490 1.920 2.130 2.170 2.120 2.130
1.0 0.640 0.700 0.920 1.220 1.590 1.840 2.010 2.080 1.840
1.5 0.460 0.470 0.600 0.790 1.050 1.310 1.610 1.800 1.310
2.0 0.360 0.360 0.430 0.580 0.750 0.990 1.320 1.590 0.990
3.0 0.250 0.250 0.280 0.370 0.470 0.620 0.860 1.050 0.620
4.0 0.190 0.190 0.200 0.260 0.320 0.420 0.580 0.700 0.420
5.0 0.160 0.160 0.170 0.200 0.240 0.310 0.410 0.500 0.310
7.5 0.110 0.110 0.100 0.120 0.140 0.170 0.220 0.260 0.170
10 0.069 0.069 0.068 0.077 0.088 0.110 0.140 0.160 0.110
Century City site.
120
site classes (and Default site condition) in the 2020 NEHRP Provisions, Northridge site (latitude 34.20, longitude -
118.55).

0.00 0.680 0.740 0.860 0.940 0.960 0.870 0.740 0.670 0.960
0.01 0.680 0.740 0.860 0.950 0.960 0.870 0.750 0.690 0.960
0.02 0.710 0.770 0.880 0.960 0.960 0.870 0.740 0.660 0.960
0.03 0.840 0.900 0.990 1.020 1.000 0.880 0.720 0.620 1.020
0.05 1.250 1.280 1.290 1.260 1.150 0.980 0.800 0.710 1.260
0.08 1.560 1.630 1.660 1.600 1.420 1.240 1.070 0.990 1.600
0.10 1.650 1.770 1.880 1.850 1.670 1.470 1.290 1.180 1.850
0.15 1.630 1.790 2.130 2.170 1.970 1.730 1.500 1.390 2.170
0.20 1.480 1.660 2.090 2.410 2.210 1.820 1.520 1.400 2.410
0.25 1.320 1.480 1.930 2.420 2.420 1.980 1.540 1.380 2.420
0.30 1.180 1.310 1.750 2.250 2.510 2.170 1.640 1.450 2.510
0.40 0.960 1.090 1.490 1.970 2.400 2.300 1.850 1.530 2.400
0.50 0.800 0.930 1.300 1.760 2.190 2.250 1.960 1.720 2.250
0.75 0.580 0.670 0.930 1.300 1.740 1.950 1.930 1.860 1.950
1.0 0.450 0.500 0.700 0.990 1.360 1.650 1.830 1.880 1.650
1.5 0.290 0.310 0.440 0.620 0.880 1.150 1.440 1.650 1.150
2.0 0.220 0.230 0.310 0.440 0.630 0.850 1.170 1.410 0.850
3.0 0.140 0.150 0.190 0.270 0.390 0.540 0.750 0.930 0.540
4.0 0.110 0.120 0.140 0.190 0.270 0.370 0.510 0.620 0.370
5.0 0.090 0.094 0.110 0.150 0.200 0.280 0.370 0.450 0.280
7.5 0.061 0.063 0.068 0.090 0.120 0.160 0.210 0.250 0.160
10 0.044 0.045 0.047 0.061 0.079 0.100 0.130 0.150 0.100
Northridge site.
121
site classes (and Default site condition) in the 2020 NEHRP Provisions, Long Beach site (latitude 33.80, longitude -
118.20).

0.00 0.610 0.680 0.800 0.880 0.910 0.870 0.800 0.760 0.910
0.01 0.620 0.690 0.800 0.880 0.910 0.880 0.810 0.770 0.910
0.02 0.640 0.710 0.820 0.900 0.920 0.880 0.800 0.750 0.920
0.03 0.750 0.820 0.930 0.970 0.970 0.890 0.790 0.700 0.970
0.05 1.070 1.150 1.210 1.200 1.130 1.010 0.870 0.710 1.200
0.08 1.380 1.490 1.560 1.540 1.420 1.250 1.030 0.910 1.540
0.10 1.480 1.640 1.780 1.790 1.660 1.480 1.210 1.090 1.790
0.15 1.460 1.660 1.980 2.080 1.980 1.790 1.440 1.300 2.080
0.20 1.300 1.500 1.900 2.180 2.160 1.970 1.580 1.380 2.180
0.25 1.130 1.320 1.720 2.140 2.240 2.100 1.730 1.430 2.240
0.30 0.990 1.170 1.560 2.010 2.250 2.190 1.920 1.560 2.250
0.40 0.810 0.950 1.310 1.750 2.150 2.190 2.120 1.750 2.190
0.50 0.680 0.800 1.120 1.530 1.960 2.100 2.090 1.950 2.100
0.75 0.570 0.660 0.890 1.200 1.570 1.770 1.860 1.890 1.770
1.0 0.530 0.580 0.750 1.010 1.320 1.550 1.730 1.820 1.550
1.5 0.390 0.400 0.500 0.680 0.900 1.130 1.400 1.590 1.130
2.0 0.310 0.310 0.370 0.500 0.670 0.880 1.170 1.410 0.880
3.0 0.220 0.220 0.250 0.330 0.430 0.580 0.800 0.980 0.580
4.0 0.170 0.170 0.180 0.250 0.310 0.410 0.560 0.670 0.410
5.0 0.140 0.140 0.150 0.200 0.240 0.300 0.410 0.490 0.300
7.5 0.097 0.097 0.099 0.120 0.140 0.170 0.220 0.260 0.170
10 0.063 0.063 0.065 0.076 0.087 0.110 0.140 0.160 0.110
Long Beach site.
122
site classes (and Default site condition) in the 2020 NEHRP Provisions, Irvine site (latitude 33.65, longitude -
117.80).

0.00 0.460 0.520 0.610 0.680 0.730 0.730 0.700 0.680 0.730
0.01 0.470 0.520 0.610 0.690 0.740 0.730 0.700 0.690 0.740
0.02 0.480 0.540 0.620 0.700 0.740 0.730 0.700 0.680 0.740
0.03 0.560 0.620 0.700 0.750 0.770 0.750 0.700 0.670 0.770
0.05 0.800 0.860 0.920 0.940 0.920 0.860 0.780 0.720 0.940
0.08 1.030 1.120 1.190 1.210 1.170 1.080 0.980 0.890 1.210
0.10 1.110 1.230 1.370 1.420 1.380 1.290 1.190 1.060 1.420
0.15 1.100 1.250 1.490 1.630 1.650 1.570 1.430 1.270 1.650
0.20 0.970 1.130 1.430 1.680 1.770 1.730 1.620 1.370 1.770
0.25 0.840 0.990 1.290 1.630 1.810 1.820 1.760 1.450 1.820
0.30 0.740 0.880 1.170 1.520 1.800 1.870 1.870 1.590 1.870
0.40 0.600 0.710 0.980 1.320 1.670 1.820 1.890 1.820 1.820
0.50 0.500 0.600 0.830 1.150 1.510 1.720 1.850 1.950 1.720
0.75 0.370 0.430 0.600 0.840 1.150 1.370 1.540 1.650 1.370
1.0 0.280 0.320 0.450 0.640 0.890 1.120 1.320 1.470 1.120
1.5 0.190 0.200 0.280 0.410 0.590 0.780 1.000 1.180 0.780
2.0 0.150 0.150 0.200 0.290 0.420 0.580 0.790 0.970 0.580
3.0 0.100 0.110 0.130 0.180 0.270 0.370 0.510 0.640 0.370
4.0 0.076 0.079 0.093 0.130 0.190 0.260 0.360 0.440 0.260
5.0 0.063 0.065 0.074 0.100 0.140 0.200 0.270 0.320 0.200
7.5 0.045 0.046 0.050 0.067 0.091 0.120 0.160 0.190 0.120
10 0.034 0.034 0.037 0.048 0.063 0.080 0.100 0.120 0.080
Irvine site.
123
site classes (and Default site condition) in the 2020 NEHRP Provisions, Riverside site (latitude 33.95, longitude -
117.40).

0.00 0.500 0.570 0.660 0.730 0.740 0.690 0.610 0.550 0.740
0.01 0.500 0.570 0.660 0.730 0.750 0.700 0.620 0.550 0.750
0.02 0.520 0.580 0.680 0.740 0.750 0.700 0.620 0.550 0.750
0.03 0.600 0.660 0.750 0.790 0.780 0.700 0.620 0.550 0.790
0.05 0.820 0.890 0.950 0.960 0.890 0.760 0.630 0.580 0.960
0.08 1.040 1.140 1.210 1.190 1.080 0.900 0.800 0.780 1.190
0.10 1.120 1.250 1.370 1.370 1.240 1.040 0.930 0.910 1.370
0.15 1.120 1.290 1.530 1.610 1.500 1.270 1.050 1.020 1.610
0.20 1.010 1.190 1.500 1.710 1.660 1.440 1.150 1.010 1.710
0.25 0.910 1.070 1.400 1.710 1.800 1.580 1.300 1.150 1.800
0.30 0.810 0.980 1.300 1.660 1.860 1.730 1.440 1.300 1.860
0.40 0.690 0.830 1.140 1.530 1.820 1.830 1.610 1.480 1.830
0.50 0.600 0.720 1.010 1.380 1.730 1.800 1.680 1.600 1.800
0.75 0.460 0.540 0.760 1.070 1.410 1.570 1.600 1.590 1.570
1.0 0.360 0.400 0.580 0.830 1.160 1.390 1.510 1.580 1.390
1.5 0.250 0.260 0.370 0.540 0.780 1.030 1.310 1.540 1.030
2.0 0.190 0.210 0.280 0.400 0.580 0.790 1.080 1.330 0.790
3.0 0.140 0.150 0.180 0.270 0.390 0.540 0.750 0.950 0.540
4.0 0.110 0.120 0.140 0.200 0.290 0.400 0.550 0.690 0.400
5.0 0.098 0.100 0.120 0.170 0.230 0.310 0.420 0.530 0.310
7.5 0.071 0.073 0.078 0.110 0.140 0.190 0.250 0.300 0.190
10 0.052 0.052 0.055 0.072 0.094 0.120 0.150 0.180 0.120
Riverside site.
124
site classes (and Default site condition) in the 2020 NEHRP Provisions, San Bernardino site (latitude 34.10,
longitude -117.30).

0.00 0.900 0.990 1.140 1.250 1.250 1.110 0.900 0.770 1.250
0.01 0.910 0.990 1.150 1.250 1.260 1.120 0.910 0.790 1.260
0.02 0.950 1.040 1.180 1.270 1.260 1.120 0.890 0.750 1.270
0.03 1.150 1.240 1.340 1.360 1.310 1.130 0.860 0.710 1.360
0.05 1.680 1.790 1.770 1.690 1.500 1.220 0.920 0.780 1.690
0.08 2.150 2.260 2.230 2.100 1.810 1.440 1.150 1.050 2.100
0.10 2.190 2.360 2.470 2.390 2.070 1.640 1.330 1.230 2.390
0.15 2.160 2.370 2.800 2.820 2.460 1.930 1.530 1.400 2.820
0.20 1.990 2.210 2.780 3.170 2.820 2.170 1.610 1.420 3.170
0.25 1.760 2.010 2.610 3.230 3.090 2.410 1.690 1.420 3.230
0.30 1.570 1.820 2.420 3.050 3.280 2.660 1.860 1.530 3.280
0.40 1.310 1.530 2.090 2.770 3.300 2.940 2.140 1.690 3.300
0.50 1.130 1.310 1.830 2.490 3.100 2.990 2.360 1.980 3.100
0.75 0.850 0.960 1.370 1.910 2.520 2.700 2.490 2.340 2.700
1.0 0.670 0.750 1.070 1.530 2.070 2.390 2.510 2.540 2.390
1.5 0.460 0.490 0.710 1.030 1.450 1.820 2.190 2.500 1.820
2.0 0.360 0.390 0.520 0.760 1.090 1.470 1.970 2.450 1.470
3.0 0.260 0.280 0.360 0.530 0.760 1.050 1.460 1.860 1.050
4.0 0.220 0.230 0.270 0.400 0.570 0.780 1.070 1.340 0.780
5.0 0.180 0.190 0.220 0.320 0.440 0.600 0.810 1.010 0.600
7.5 0.120 0.120 0.140 0.180 0.250 0.320 0.420 0.510 0.320
10 0.084 0.085 0.091 0.120 0.150 0.190 0.240 0.290 0.190
San Bernardino site.
125
site classes (and Default site condition) in the 2020 NEHRP Provisions, San Luis Obispo site (latitude 35.30,
longitude -120.65).

0.00 0.400 0.450 0.530 0.590 0.630 0.630 0.600 0.590 0.630
0.01 0.410 0.450 0.530 0.600 0.640 0.630 0.600 0.590 0.640
0.02 0.420 0.470 0.540 0.600 0.640 0.630 0.600 0.590 0.640
0.03 0.490 0.540 0.610 0.650 0.670 0.650 0.600 0.570 0.670
0.05 0.690 0.740 0.790 0.810 0.790 0.730 0.650 0.620 0.810
0.08 0.890 0.960 1.020 1.040 1.000 0.910 0.820 0.780 1.040
0.10 0.950 1.050 1.170 1.210 1.170 1.090 0.990 0.950 1.210
0.15 0.940 1.070 1.290 1.400 1.400 1.330 1.230 1.150 1.400
0.20 0.840 0.970 1.230 1.440 1.520 1.470 1.390 1.280 1.520
0.25 0.730 0.860 1.120 1.410 1.560 1.570 1.510 1.370 1.570
0.30 0.650 0.760 1.020 1.330 1.550 1.610 1.600 1.480 1.610
0.40 0.530 0.620 0.860 1.160 1.460 1.580 1.640 1.670 1.580
0.50 0.440 0.530 0.740 1.020 1.340 1.520 1.610 1.690 1.520
0.75 0.330 0.380 0.530 0.750 1.030 1.220 1.360 1.470 1.220
1.0 0.250 0.280 0.400 0.580 0.800 1.000 1.180 1.320 1.000
1.5 0.170 0.180 0.250 0.370 0.530 0.710 0.910 1.080 0.710
2.0 0.130 0.140 0.190 0.270 0.390 0.530 0.730 0.900 0.530
3.0 0.092 0.097 0.120 0.170 0.250 0.350 0.480 0.610 0.350
4.0 0.071 0.075 0.088 0.130 0.180 0.250 0.340 0.430 0.250
5.0 0.060 0.062 0.071 0.099 0.140 0.190 0.260 0.320 0.190
7.5 0.044 0.045 0.049 0.066 0.090 0.120 0.160 0.190 0.120
10 0.034 0.035 0.037 0.048 0.064 0.081 0.100 0.120 0.081
San Luis Obispo site.
126
site classes (and Default site condition) in the 2020 NEHRP Provisions, San Diego site (latitude 32.70, longitude -
117.15).

0.00 0.560 0.630 0.730 0.800 0.820 0.760 0.680 0.630 0.820
0.01 0.570 0.640 0.740 0.810 0.820 0.770 0.680 0.640 0.820
0.02 0.590 0.660 0.760 0.820 0.830 0.770 0.680 0.630 0.830
0.03 0.700 0.770 0.860 0.890 0.870 0.780 0.670 0.610 0.890
0.05 1.010 1.080 1.130 1.110 1.020 0.870 0.720 0.650 1.110
0.08 1.290 1.390 1.450 1.400 1.260 1.060 0.880 0.810 1.400
0.10 1.370 1.510 1.630 1.610 1.450 1.240 1.050 0.970 1.610
0.15 1.340 1.540 1.820 1.880 1.740 1.510 1.280 1.180 1.880
0.20 1.190 1.380 1.740 1.980 1.900 1.670 1.440 1.310 1.980
0.25 1.030 1.210 1.570 1.940 1.980 1.790 1.560 1.360 1.980
0.30 0.910 1.070 1.420 1.820 2.000 1.870 1.680 1.460 2.000
0.40 0.730 0.860 1.180 1.560 1.880 1.870 1.740 1.640 1.880
0.50 0.610 0.710 0.990 1.360 1.710 1.800 1.740 1.740 1.800
0.75 0.440 0.500 0.710 1.000 1.330 1.490 1.540 1.600 1.490
1.0 0.330 0.370 0.530 0.760 1.040 1.230 1.380 1.480 1.230
1.5 0.200 0.220 0.310 0.460 0.650 0.840 1.050 1.230 0.840
2.0 0.150 0.160 0.220 0.310 0.450 0.620 0.840 1.040 0.620
3.0 0.098 0.100 0.130 0.190 0.270 0.380 0.530 0.670 0.380
4.0 0.071 0.074 0.088 0.130 0.180 0.250 0.350 0.440 0.250
5.0 0.056 0.058 0.067 0.094 0.130 0.180 0.250 0.300 0.180
7.5 0.037 0.037 0.041 0.055 0.075 0.099 0.130 0.160 0.099
10 0.027 0.027 0.029 0.038 0.050 0.065 0.084 0.099 0.065
San Diego site.
127
site classes (and Default site condition) in the 2020 NEHRP Provisions, Santa Barbara site (latitude 34.45, longitude
-119.70).

0.00 0.760 0.850 0.990 1.070 1.070 0.990 0.870 0.770 1.070
0.01 0.770 0.860 0.990 1.070 1.080 1.000 0.880 0.780 1.080
0.02 0.800 0.880 1.020 1.090 1.090 1.000 0.870 0.750 1.090
0.03 0.940 1.030 1.140 1.180 1.130 1.010 0.850 0.700 1.180
0.05 1.370 1.460 1.500 1.470 1.330 1.120 0.870 0.710 1.470
0.08 1.770 1.900 1.950 1.860 1.640 1.380 1.050 0.910 1.860
0.10 1.880 2.080 2.200 2.160 1.910 1.600 1.220 1.090 2.160
0.15 1.840 2.090 2.450 2.500 2.290 1.900 1.450 1.300 2.500
0.20 1.630 1.890 2.370 2.660 2.500 2.150 1.590 1.380 2.660
0.25 1.420 1.660 2.160 2.640 2.640 2.340 1.720 1.440 2.640
0.30 1.270 1.480 1.980 2.510 2.720 2.470 1.900 1.560 2.720
0.40 1.040 1.220 1.670 2.220 2.630 2.540 2.150 1.720 2.630
0.50 0.880 1.020 1.430 1.960 2.420 2.470 2.300 1.940 2.470
0.75 0.680 0.790 1.100 1.530 1.990 2.160 2.130 2.080 2.160
1.0 0.540 0.620 0.870 1.210 1.640 1.890 2.020 2.060 1.890
1.5 0.350 0.390 0.540 0.760 1.060 1.350 1.640 1.820 1.350
2.0 0.260 0.290 0.380 0.540 0.760 1.030 1.370 1.630 1.030
3.0 0.170 0.180 0.230 0.320 0.460 0.630 0.880 1.080 0.630
4.0 0.120 0.130 0.160 0.220 0.310 0.420 0.580 0.710 0.420
5.0 0.097 0.100 0.120 0.170 0.230 0.310 0.420 0.500 0.310
7.5 0.063 0.066 0.074 0.098 0.130 0.170 0.220 0.260 0.170
10 0.044 0.046 0.050 0.065 0.084 0.110 0.140 0.160 0.110
Santa Barbara site.
128
site classes (and Default site condition) in the 2020 NEHRP Provisions, Ventura site (latitude 34.30, longitude -
119.30).

0.00 0.730 0.810 0.940 1.030 1.040 0.970 0.860 0.780 1.040
0.01 0.730 0.820 0.950 1.030 1.050 0.980 0.870 0.790 1.050
0.02 0.760 0.840 0.970 1.040 1.050 0.970 0.860 0.760 1.050
0.03 0.900 0.980 1.090 1.120 1.090 0.990 0.840 0.690 1.120
0.05 1.300 1.380 1.430 1.390 1.270 1.100 0.880 0.690 1.390
0.08 1.660 1.790 1.840 1.760 1.580 1.340 1.010 0.870 1.760
0.10 1.770 1.950 2.090 2.050 1.840 1.570 1.160 1.030 2.050
0.15 1.740 1.980 2.330 2.380 2.210 1.880 1.370 1.220 2.380
0.20 1.550 1.790 2.250 2.550 2.430 2.130 1.530 1.310 2.550
0.25 1.370 1.590 2.080 2.540 2.570 2.310 1.690 1.390 2.570
0.30 1.220 1.430 1.900 2.410 2.640 2.460 1.890 1.530 2.640
0.40 1.000 1.170 1.600 2.150 2.570 2.510 2.200 1.730 2.570
0.50 0.850 0.990 1.390 1.900 2.360 2.440 2.310 1.960 2.440
0.75 0.730 0.840 1.140 1.550 2.010 2.190 2.190 2.160 2.190
1.0 0.690 0.760 1.010 1.350 1.760 2.020 2.170 2.220 2.020
1.5 0.490 0.520 0.670 0.910 1.200 1.500 1.810 2.020 1.500
2.0 0.390 0.410 0.490 0.670 0.890 1.170 1.560 1.850 1.170
3.0 0.270 0.280 0.320 0.430 0.570 0.760 1.040 1.270 0.760
4.0 0.200 0.210 0.230 0.310 0.400 0.520 0.700 0.860 0.520
5.0 0.170 0.180 0.190 0.240 0.300 0.380 0.510 0.610 0.380
7.5 0.110 0.120 0.120 0.140 0.170 0.210 0.270 0.320 0.210
10 0.073 0.074 0.077 0.090 0.100 0.130 0.160 0.190 0.130
Ventura site.
129
THE PROPOSED ASCE 7-22 MULTI-PERIOD RESPONSE SPECTRA –

IMPACT ON THE LOS ANGELES METROPOLITAN AREA
Marshall Lew and Kenneth S. Hudson

Wood Environment and Infrastructure Solutions, Inc.
Introduction
A new multi-period response spectrum will be a major change to be introduced in the forthcoming ASCE
7-22 design criteria. This major change to the seismic provisions brings more soil site class definitions
and defines a more complex design spectrum that is intended to reflect better the spectral response
demands on structures when compared with the current ASCE 7-16 design criteria. This study reports on
the expected impact of the new ASCE 7-22 Multi-Period Response Spectra (MPRS) design criteria on
design of structures in the Los Angeles metropolitan area. This study identifies 30 sites with well-defined
average shear wave velocity in the upper 30 meters or 100 feet (denoted as VS30) throughout the Los
Angeles metropolitan area and compares the design spectra based on the methodologies of the current
ASCE 7-16 and the proposed ASCE 7-22.
In addition, this study also compares the design spectra developed using the proposed ASCE 7-22
methodology with the results of design spectra that would be developed using a site-specific ground
motion analysis following the requirements of Chapter 21 of ASCE 7; specifically, the design spectra are
determined using a probabilistic seismic hazard analysis (PSHA) to determine the maximum considered
earthquake response spectra (MCER).
Changes in the New ASCE/SEI 7-22 Seismic Provisions for the Design Earthquake Ground Motions
The proposed changes to the ASCE 7-22 seismic provisions have been presented in the 2020 edition of
the NEHRP Recommended Seismic Provisions for New Buildings and Other Structures (Building
Seismic Safety Council, 2020). These changes are intended to be improvements to the ASCE/SEI 7-16
Standard (ASCE, 2017). Among these changes in ASCE 7-22 are the new definitions of the site classes
and the process in which the MPRS design spectral accelerations are derived.
Changes in Site Class Definition
In ASCE 7-16 Chapter 20, there are six site classes (A, B, C, D, E and F) which are mostly defined in
terms of VS30. ASCE 7-22 will add three additional site classes (BC, CD and DE) that are intended to
provide better resolution of the site average shear wave velocity and associated site effects for common
site conditions. As stated in the commentary of NEHRP report, the new site classes allow for more
accurate derivation of the amplitude and frequency content of earthquake ground motions, and their
variation with shaking intensity (nonlinear effects); in addition, the new site classes are of particular
importance to the characterization of long period ground motions for softer sites. The definitions by
average shear wave velocity of the ASCE 7-16 and ASCE 7-22 site classes are presented in Table 1.
130
Table 1. Definitions of ASCE 7-16 and ASCE 7-22 Site Classes
ASCE 7-16 ASCE 7-22

Site Class VS30 (ft/sec) Site Class VS30 (ft/sec)
A Hard rock >5,000 A Hard rock >5,000
B Rock 2,500 to 5,000 B Medium hard rock >3,000 to 5,000
BC Soft rock >2,100 to 3,000
C Very dense soil and soft 1,200 to 2,500
rock C Very dense sand or hard clay >1,450 to 2,100
CD Dense sand or very stiff clay >1,000 to 1,450
D Stiff Soil 600 to 1,200 D Medium dense sand or stiff clay >700 to 1,000
DE Loose sand or medium stiff >500 to 700
E Soft Soil <600 clay
E Very loose sand or soft clay <500
Soils requiring site - Soils requiring site -
F F
response analysis response analysis
Changes in the Design Response Spectrum
As described by Razaeian, Luco and Kircher (2021), the Buildings Seismic Safety Council commissioned
a study by Kircher & Associates (2015) to investigate the compatibility of site amplifications, Fq and Fv
with the ground motion models used by the USGS to produce design maps for building code use. It was
found that the standard three-domain (two-period) spectral shape used in ASCE 7-16 and earlier editions
is not appropriate for soft soil sites (ASCE 7-16 Site Class D or softer) where the ground motion hazard is
dominated by large magnitude events. It was found that the standard spectral shape underestimates the
spectral response for structures having moderate to long periods.
The ASCE 7-16 design response spectrum is defined using a two-period response spectrum unless the
design is based on site-specific ground motions. The ASCE 7-22 design response spectrum is defined on a
multi-period response spectrum and is intended to provide a more accurate representation of the
frequency content of design ground motions than the two-period response spectrum and is described by
Razaeian, Luco and Kircher (2021). The design spectral acceleration parameters SDS and SD1 for
determining the multi-period response spectrum are to be provided online by the USGS Seismic Design
Web Service for a user-specified site location (i.e., in terms of latitude and longitude) and site class
(USGS, 2021). Use of the two-period response spectrum is retained in ASCE 7-22 as an alternative where
the multi-period spectrum is not available.
Sets of multi-period MCER response spectra (5% damping) at 22 response periods (i.e., 0.0 sec, 0.01 sec,
0.02 sec, 0.03 sec, 0.05 sec, 0.075 sec, 0.1 sec, 0.15 sec, 0.2 sec, 0.25 sec, 0.3 sec, 0.4 sec, 0.5 sec, 0.75
sec, 1.0 sec, 1.5 sec, 2.0 sec, 3.0 sec, 4.0 sec, 5.0 sec, 7.5 sec and 10 sec) are archived in the USGS
Seismic Design Geodatabase at gridded locations across regions of interest within the United States. The
USGS Seismic Design Web Service, for the site location and site class of interest, spatially interpolates
between the gridded sets of multi-period MCER response spectra based on site location (i.e., latitude and
longitude). The multi-period design response spectrum is constructed from two-thirds of these values by
linear interpolation for response periods less than 10 sec and by extrapolation for response periods greater
than 10 sec.
131
Los Angeles Metropolitan Area Sites Selected for Study
A total of 30 sites in different site soil class conditions with known average shear wave velocities (VS30) in
the Los Angeles Basin were selected for this study. Sixteen sites were selected from the ROSRINE
project (Nigbor and Swift, 2001) and an additional fourteen sites were selected from project files of Wood
Environment and Infrastructures Solutions, Inc. The locations of the 30 sites are identified in Table 2 and
the sites are shown in Figure 1; in addition, Table 2 provides the source of the average shear wave
velocity data, the location given in terms of the site coordinates in latitude and longitude, the VS30 at the
site and the corresponding ASCE 7-16 and ASCE 7-22 soil site classes based on VS30.
Table 2. Study Sites

SITE CLASS
SITE SOURCE LATITUDE LONGITUDE VS30
ID SITE NAME ASCE ASCE
NO. (degrees) (degrees) (ft/sec)
7-16 7-22
1 PAC Pacoima Downstream ROSRINE 34.335 -118.398 6724 A A
2 GPK Griffith Observatory ROSRINE 34.118 -118.300 3214 B B
3 LA0 Stone Canyon Reservoir ROSRINE 34.106 -118.455 2230 C BC
4 CEN Century City WOOD 34.057 -118.416 1590 C C
5 ING Inglewood WOOD 33.954 -118.339 1473 C C
6 SYL Olive View Hospital, Sylmar ROSRINE 34.328 -118.444 1410 C CD
7 BVA Brentwood VA Hospital ROSRINE 34.063 -118.462 1348 C CD
8 EXP Exposition Park WOOD 34.016 -118.287 1365 C CD
9 LA2 Downtown LA - South Park WOOD 34.041 -118.271 1306 C CD
10 SPV2 Sepulveda VA #5 B-2 ROSRINE 34.249 -118.478 1246 C CD
11 SAM Santa Monica WOOD 34.025 -118.510 1245 C CD
12 LA1 Downtown LA - Bunker Hill WOOD 34.055 -118.249 1240 C CD
13 WST Westwood WOOD 34.059 -118.446 1235 C CD
14 LAX LA International Airport WOOD 33.944 -118.400 1148 D CD
15 MAC McArthur Park WOOD 34.059 -118.280 1104 D CD
16 116 Willowbrook Park/Gardena ROSRINE 33.923 -118.256 1040 D CD
17 MRC Miracle Mile WOOD 34.062 -118.347 1039 D CD
18 LBH Long Beach Civic Center WOOD 33.768 -118.197 1020 D CD
19 CUL Culver City/Los Angeles WOOD 34.026 -118.376 1015 D CD
20 ARL Arleta ROSRINE 34.236 -118.440 991 D D
21 BLD Baldwin Hills ROSRINE 34.008 -118.362 961 D D
22 WHW West Hollywood WOOD 34.081 -118.389 914 D D
23 WOC White Oak Church ROSRINE 34.208 -118.517 909 D D
24 CNP Canoga Park ROSRINE 34.212 -118.605 886 D D
25 NWH Newhall Fire Station ROSRINE 34.387 -118.534 882 D D
26 SOP Sherman Oaks Park ROSRINE 34.161 -118.439 859 D D
27 ROS Rosemead WOOD 34.052 -118.083 842 D D
28 BIR Downey South Middle School ROSRINE 33.922 -118.138 820 D D
29 PR2 Pico Rivera #2 ROSRINE 33.972 -118.088 794 D D
30 WAT Dolphin Park/Del Amo ROSRINE 33.837 -118.241 623 D DE
132
Because of the geologic materials that constitute the basins in the Los Angeles metropolitan area, the
majority of the study sites are classified as ASCE 7-16 Site Classes C and D. The ASCE 7-16 Site Class
C sites fall into both ASCE 7-22 Site Classes C and CD and the ASCE 7-16 Class D sites fall into both
ASCE 7-22 Site Classes CD and D. Only six sites are classified as ASCE 7-22 Site Classes A, B, BC, C
or DE; there were no sites identified as being ASCE 7-22 Site Classes AB or E. It is likely that greater
than 95 percent of all construction in the Los Angeles Basin would be located at sites being classified as
either ASCE 7-22 Site Class CD or D.
Figure 1. Location of the 30 study sites in the Los Angeles Basin.
Comparisons of ASCE 7-16 and ASCE 7-22 Design Spectra at the Study Sites
To compare the prescriptive design earthquake spectra derived by the ASCE 7-16 and ASCE 7-22
provisions, the procedures in each document were followed using the location and respective site class for
each site and the calculations were performed using the Beta version of the Seismic Design Web Service
133
produced by the U.S. Geological Survey (USGS, 2021). Also shown is the site-specific design spectrum
which was determined by performing a probabilistic seismic hazard analysis (PSHA). A PSHA was
performed for each of the 30 sites using the computer program OpenSHA (Field et al., 2003) to develop
the site-specific uniform hazard response spectra for maximum considered earthquake (MCE) ground
motions having a 2% probability of being exceeded in 50 years. The site-specific probabilistic response
spectra were developed using the fault model in the OpenSHA program. An average of Fault Models 3.1
and 3.2 from the Third Uniform California Earthquake Rupture Forecast (UCERF3) was used as the
earthquake rupture forecast model for the PSHA. In addition to known fault sources, background
seismicity was also included in the PSHA. The computed 5%-damped probabilistic geomean ground
motions were converted to maximum response direction ground motions using the scaling factors
recommended by Shahi and Baker (2014). In accordance with Section 21.2.1 of Chapter 21 of ASCE 7-
16, the probabilistic risk-targeted response spectrum (MCER) was taken as the maximum direction
response spectrum with a 2% probability of being exceeded in 50 years multiplied by the risk coefficients
CRS and CR1 to achieve a 1% probability of collapse within a 50-year period. The deterministic MCER
ground motions were not determined for this study as experience indicates that the deterministic ground
motions are usually higher than the probabilistic ground motions in the Los Angeles Basin and would not
generally govern the MCER ground motions for design purposes. The design response spectrum was
determined by taking two-thirds of the MCER spectrum in accordance with Section 21.3 of ASCE 7-16.
The design earthquake spectra determined by ASCE 7-16, ASCE 7-22 and site-specific analysis for each
of the 30 sites are presented in Figures 2 through 31.
Figure 2. Pacoima Downstream (PAC) Figure 3. Griffith Observatory (GPK)

Site Class A Site Class B
Figure 4. Stone Canyon Reservoir (LA0) Figure 5. Century City (CEN)

Site Class BC Site Class C
134
Figure 6. Inglewood (ING) Figure 7. Olive View Hospital Sylmar (SYL)

Site Class C Site Class CD
Figure 8. Brentwood VA Hospital (BVA) Figure 9. Exposition Park (EXP)

Site Class CD Site Class CD
Figure 10. Downtown L.A. South Park (LA2) Figure 11. Sepulveda VA #5 B-2 (SPV2)
135
Figure 12. Santa Monica (SAM) Figure 13. Downtown L.A. Bunker Hill (LA1)
Figure 14. Westwood (WST) Figure 15. L.A. International Airport (LAX)
Figure 16. MacArthur Park (MAC) Figure 17. Willowbrook Park/Gardena (116)
136
Figure 18. Miracle Mile (MRC) Figure 19. Long Beach Civic Center (LBH)
Figure 20. Culver City/Los Angeles (CUL) Figure 21. Arleta (ARL)
Site Class CD Site Class D
Figure 22. Baldwin Hills (BLD) Figure 23. West Hollywood (WHW)
Site Class D Site Class D
137
Figure 24. White Oak Church (WOC) Figure 25. Canoga Park (CNP)
Figure 26. Newhall Fire Station (NWH) Figure 27. Sherman Oaks Park (SOP)
Figure 28. Rosemead (ROS) Figure 29. Downey Middle School (BIR)
138
Figure 30. Pico Rivera #2 (PV2) Figure 31. Dolphin Park/Del Amo
Site Class D (WAT) Site Class DE
It should be noted that the USGS website indicates that the MPRS output at certain latitudes and
longitudes within the Los Angeles Basin need correction and the resulting output may change; we do not
expect that the MPRS design spectra from the Beta version will vary significantly from the final corrected
output when the corrections are made.
Comparisons of ASCE 7-22 Design Spectra with Design Spectra determined by Site-Specific
Analyses
Figure 32 illustrates the ratio of the average of the ASCE 7-22 design spectra to the site-specific design
spectra for each of the ASCE 7-22 site classes. The comparisons for ASCE 7-22 Site Classes A, B, BC, C
and DE should be ignored as number of sites is just one or two and are not statistically significant.
However, there are 14 Site Class CD sites and 10 Site Class D sites in the study area. For Site Class CD,
the spectral ratio indicates that the ASCE 7-22 design spectra is between 3 to 10% greater than the site-
specific design spectra. This indicates reasonable agreement with this sample for the Los Angeles Basin.
For Site Class D, the ASCE 7-22 design spectra ranges between 93 to 115% of the site-specific design
spectra. ASCE 7-22 spectral values are lower than the site-specific values for periods less than about 0.3
seconds and increase to about 115% in the longer periods. This would indicate that the ASCE 7-22 design
levels will be conservative for longer structural periods, especially for Site Class D. For taller buildings
and buildings with longer fundamental periods, the ASCE 7-22 design levels for Site Classes CD and D
will be more conservative if site-specific ground motion analyses are not performed. This was the concern
for ASCE 7-16 Site Classes D and E where the Kircher & Associates study indicated that the design
spectrum was lower in longer periods than what would be expected on soft soil sites and the multi period
response spectra approach was a remedy to that situation (BSSC, 2019).
139
Figure 32.
Average Spectral Ratio Between ASCE 7-22 MPRS Design Earthquake and Site-Specific Design
Earthquake Spectra
A comparison of the ratio of the ASCE 7-16 and ASCE 7-22 design earthquake spectral accelerations to
the site-specific design earthquake spectral accelerations at selected periods for the study sites that were
within ASCE 7-22 Site Classes C, CD and D according to the average shear wave velocity (VS30) of each
site and is shown in Figure 33. For these selected periods, it is observed that the ASCE 7-22 design
spectra appear to be generally closer in value to the design spectra that are derived from a site-specific
ground motion analysis; the ASCE 7-22 spectral values would be somewhat more conservative than the
site-specific design spectral values. The ASCE 7-16 design spectra diverges more from the site-specific
design spectra and the ratio is consistently lower for periods between periods of about 0.2 to between 2.0
and 5.0 seconds. At periods greater than 5 seconds, ASCE 7-16 design spectra becomes higher. At the
transition between ASCE 7-16 Site Classes C and D at 1,200 ft/sec, it appears that there is some
discontinuity between the spectral ratios.
140
Figure 33. Ratio of ASCE 7-22 and ASCE 7-16 design spectra to Site-Specific design response spectra as
a function of VS30 for the sites in ASCE 7-22 Site Classes C, CD and D in this study.
Implications for Design of Tall Buildings in the Los Angeles Metropolitan Area
Based on the results of this study, it appears that the proposed ASCE 7-22 MPRS design spectra would be
somewhat more conservative than the design spectra from a site-specific ground motion analysis done by
the current practice using PSHA analysis. The ASCE 7-22 design spectra appear to be more consistent
with the PSHA-derived site-specific results than the current ASCE 7-16 design spectra. These conclusions
would also apply to the risk-targeted maximum considered earthquake (MCER) ground motions as the
design spectra would be multiplied by a factor of 1.5 and maintain the same ratios.
141
Conclusions
This study has examined the possible implications of the future ASCE 7-22 provisions with regard to the
changes in the definition and computation of the design earthquake ground motions as found in the ASCE
7-22 recommended seismic provisions which bring forth new and expanded definitions of nine soil site
classes and a new 22-point design response spectrum to supersede the current ASCE 7-16 six soil site
classes and two-point design response spectrum. The study was limited to the geographical area known as
the Los Angeles metropolitan area and some 30 sites with known average shear wave velocities were
evaluated using ASCE 7-16, ASCE 7-22, and site-specific ground motion analyses.
Because of the changes in the shear wave velocity definition of the site classes, ASCE 7-16 Site Class D
sites in this study were either ASCE 7-22 Site Class CD (where the Vs30 was greater than 1,000 ft/sec) or
ASCE 7-22 Site Class D (for VS30 less than or equal to 1,000 ft/sec). Similarly, the ASCE 7-16 Site Class
C sites in this study were either NEHRP Site Class CD or C depending on whether the VS30 was less than
or greater than 1,450 ft/sec. Because of the geologic makeup of the Los Angeles Basin, most of the study
sites were classified as ASCE 7-22 Site Class CD or D. Therefore, it is difficult to make any conclusions
regarding how the changes affects sites that are other than Site Class CD or D.
From this study, it appears that the ASCE 7-22 design spectra for site class CD compare favorably with
design spectra that are derived from site-specific PSHA analyses; the ASCE 7-22 MPRS design spectra
were between about 3 to 10% greater than the site-specific design spectra over the entire period range
from 0 to 10 seconds. For NEHRP site class D, the comparison shows more variation with the spectra
ratio of ASCE 7-22 to site-specific design spectra being between about 7% lower to 15% greater over the
period range studied. The results would suggest that structures located at sites within Site Classes CD and
D designed using the ASCE 7-22 MPRS design earthquake spectra would be slightly more conservative
than using site-specific ground motions for periods greater than 3 seconds.
Some caveats on this study should be stated. First of all, this study was limited to the Los Angeles
metropolitan area and its applicability to other localities or regions may not be reliable. Secondly, the
ASCE 7-22 values for the Los Angeles metropolitan area obtained from the USGS Beta website may
change because of needed corrections; however, it is the opinion of the authors that any changes will not
result in large changes in the values obtained and that the conclusions presented in this paper would not
materially be altered. Thirdly, this is not a statistically significant study for ASCE 7-22 Site Classes A,
AB, C, DE and E based on the limited number of sites analyzed in these site classes; also caution should
be applied to extrapolating these findings to locations where few sites were selected from (such as the
South Bay region in the southwestern portion of the Los Angeles metropolitan area) where basin effects
may vary from these results.
Acknowledgement
Much appreciation is extended to Dr. Robert L. Nigbor for providing data from the ROSRINE project to
facilitate this study.
References
American Society of Civil Engineers, 2017. “Minimum Design Loads and Associated Criteria for
Buildings and Other Structures,” ASCE/SEI Standard 7-16, Reston, Virginia.
Building Seismic Safety Council, 2020. “NEHRP Recommended Seismic Provisions for New
Buildings and Other Structures, Volume I,” FEMA P-2082-1, September.
142
Field, E.H., Biasi, G.P., Bird, P., Dawson, T.E., Felzer, K.R., Jackson, D.D., Johnson, K.M., Jordan, T.H.,
Madden, C., Michael, A.J., Milner, K.R., Page, M.T., Parsons, T., Powers, P.M., Shaw, B.E., Thatcher,
W.R., Weldon, R.J., II, and Zeng, Y., 2013. “Uniform California Earthquake Rupture Forecast, version 3
(UCERF3)—The time-independent model,” U.S. Geological Survey Open-File Report 2013–1165, 97 p.,
California Geological Survey Special Report 228, and Southern California Earthquake Center Publication
1792, http://pubs.usgs.gov/of/2013/1165/.
Field, E. H., Jordan T. H., and Cornell C. A., 2003. “OpenSHA: a developing community-modeling
environment for seismic hazard analysis,” Seismological Research Letters, 2003; 74 (4): 406-419.
Building Seismic Safety Council, “BSSC Project 17 Final Report – Development of Next Generation of
Seismic Design Value Maps for the 2020 NEHRP Provisions,” National Institute of Building Sciences,
December.
Kircher & Associates, 2015. “Investigation of an identified shortcoming in the seismic design
procedures of ASCE 7-10 and development of recommended improvements for ASCE 7-16,”
Building Seismic Safety Council of the National Institute of Building Sciences, Washington, D.C.
Nigbor, R.L. and Swift, J.N., 2001. “Resolution of Site Response Issues in the Northridge
Earthquake (ROSRINE),” Report ROSRINE/USC-01-Rev.0.5, October.
Rezaeian, S., Luco, N. and Kircher, C., 2021. “Multi-Period Response Spectra,” Proceedings of
the 2021 Los Angeles Tall Buildings Conference, Los Angeles Tall Buildings Structural Design
Council, November 12.
Shahi, S.K., and Baker, J.W., 2014. “NGA-West2 Models for Ground Motion Directionality,”
Earthquake Spectra, 30, 1285-1300.
U.S. Geological Survey, 2021. “Seismic Design Web Service Documentation (Beta),” URL
downloaded from https://earthquake.usgs.gov/ws/designmaps/, accessed September 2021.
143
LATBSDC
Annual Conference – November 12, 2021
Update from the LATBSDC Resiliency

Committee + Resiliency Session Introduction
Atila Zekioglu
Principal | Arup
Member | LATBSDC
Presented on behalf of the LATBSDC Resiliency Committee:

Farzad Naeim, Saiful Islam, Kristjian Kolozvari, John Wallace, Farzin Zareian, Atila Zekioglu
Context from the LATBSDC 10/2020 Conference
Mission: Help improve the
Resiliency of Los Angeles
• Engagement/feedback from LA
City and other Stakeholders
• Look for incentives, continue
with case studies and messaging
• Articulate a vision in a Position
Paper
• “LATBSDC: Resilient Design
Guidelines ver. 1” within a 2 to
3 year timeline.
Initial Meeting with LA City 4/15/2021
Objective & Highlights:
To introduce LATBSDC and
share our mission to Improve the
Resiliency of LA
• Focused the initial discussion
on Building-Codes, Seismic
Performance and Resiliency
• Described the Cars+Accidents
vs Buildings+Earthquakes
Comparison/Contrast to drive
the point that as a Society we
can’t look at Buildings as
Cars
• Articulated the need for
Incentives (LA City and
Others) towards motivating
All Stakeholders to strive for
achieving Improved
Resiliency for their Buildings,
in design, to be designed or
existing.
• Expressed our desire to
collaborate….
Today’s Resiliency Session + A Special Ask….
We thank the distinguished presenters for agreeing to
participate in our session today….
Special Ask from LATBSDC Resiliency Committee….

We would greatly appreciate if you can share any experiences from your projects related to
your Resiliency efforts:
• What are/was the Incentive(s) to drive for Improved Resiliency?
• What Obstacles did you, or are trying to, navigate through?
You can email us: atila.zekioglu@arup.com farzad@fnaeim.com Saiful@SaifulBouquet.com
SEISMIC STRUCTURAL HEALTH MONITORING AS AN ESSENTIAL

INGREDIENT OF RESILIENT DESIGN
Farzad Naeim
President, Farzad Naeim, Inc.
CEO, Mehrain Naeim International Inc.
Adjunct Professor, University of California, Irvine
Abstract
Design of resilient buildings and achieving resilient communities are hot topics of discussion by public
officials and research and development by engineers and researchers. Achieving resiliency goals,
however, is impossible without effective use of seismic instrumentation and structural health monitoring
for timely assessment of status of buildings and other civil infrastructure immediately following an
earthquake. Other countries with much less sophisticated instrumentation schemes and distributions are
far ahead of the United States and California in this regard. It is time for California to leverage the
advantage of its impressive collection of instrumented civil structures and put it to use for seismic
structural health monitoring (S2HM) to realize its desire to move towards resilient buildings and
communities.
Introduction
This paper and its accompanying presentation serve as the opening segments for presentations that discuss
and relate S2HM and resilience in this conference. In this paper, the author attempts to illustrate the
critical building status information that can be readily available in California following an earthquake if
the existing instrumentation programs and their database of instrumented buildings such as those
contained in the CSMIP database 1 and buildings instrumented per requirements of LATBSDC guidelines2
are utilized for S2HM purposes. The Chilean and Japanese S2HM experiences will be presented
respectively by Professors Boroschek and Nakashima in the papers and presentations that follow.
Use of California Instrumentation Data for S2HM Purposes and Resilience Verification
A few examples of utilizing the information obtained from instrumented buildings during past California
earthquake along with very simple and readily available fragility functions illustrates the enormous utility
of S2HM for assessing the post-earthquake status of a building and its components and contents and
verify whether it behaved in a resilient manner or not immediately following an earthquake. Several such
examples are introduced during this presentation. Obviously, if the use of simple and elementary
fragilities can produce very useful status information, then much better and more accurate results can be
achieved using more sophisticated techniques such as linking and calibrating analytical models of the
building with the instrumentation data throughout the life of the building (hybrid simulation).
1
https://www.cesmd.org/
2
https://drive.google.com/file/d/1cBv7S4Eh_c8IoPtrch7uxIHP9FbQELO_/view
144
145
146
147
148
149
150
151
152
153
154
155
CHILE S2HM EXPERIENCE
Ruben Boroschek
Professor, Civil Engineering Department
University of Chile
Abstract
Chile is a high seismic country that is starting to apply Structural Health Monitoring methodology in
research areas but also on practical projects. We have used it in tall buildings, ports and industrial
facilities. Results indicated that SHM could help us improve our designs, the rapid evaluation of
infrastructure after an extreme events like earthquakes, structural deterioration identification due to use
and environment, and to provide information accurate and timely for the stakeholders. I show in this
paper some of the applications I have been involved in Chile.
Introduction
Chile is a high seismic country. Very early in our modern history we realized that we needed to design for
these actions. We do have damaging magnitude 7.5 events every five years, a magnitude 8.5 every 15
years and in the South of Chile a magnitude 9 or higher every 300 years approximately. But our
seismicity generates vibrations that in general are non-damaging. If we check the last six years, a
magnitude 6 earthquake occur every other month in the country, Figure 1. In our tectonic setting they
typically produce accelerations around 2 or 3%g. These relative low values are cause by subduction
events on the interplate or medium depth intraplate, so distance to urban areas is generally larger than 30
km. No crustal surface events are considered in this case. It’s like a non controlled shake table that vibrate
our systems, hopefully in the linear range. We can use this seismicity together with normal structural
response to the environment and use to help us understand its characteristic and to evaluate its safety.
Figure 1. Number of earthquakes per year with magnitude 6 or

higher in Chile. Excluding the year of mayor earthquake (2010,
2014, 2015) it is close to one every two months.
156
Engineers in Chile had designed buildings for this type of activity incorporating shear walls since 1906,
after the great Valparaiso earthquake, and has continued to do so until now, (Ruben Boroschek and
Retamales 2016). This has helped to reduce the level of damage in infrastructure. Nevertheless, the recent
magnitude 8.8 earthquake of 2010 has shown that we have a critical weakness. During this earthquake,
the area affected by strong shaking, meaning accelerations around 0.3g was close to 200 by 600
kilometers, (Rubén L. Boroschek et al. 2012; Lagos et al. 2020). The number of buildings higher than
three stories affected were close to 10,000. We had more than 12,000 bridges, overpasses and footbridges,
150 medical facilities, 150 port in the affected areas. Not to mention schools, roads, water and energy
systems. We have more than 230,000 houses destroy after the 2 1/2 minutes of shaking. Early in the
morning, the same event day, we start evaluating damage. At that time in Chile there were nearly 3000
engineers able or qualified to do some level of evaluation. With this number of people, it was impossible
to review all the infrastructure in a short time. Actually, it took us nearly four months to just visit the
damage structures and to do a short summary of the possible damage. This prompted us to develop
automatic systems that will inform the users, the owners and the community of the state health or level of
damage of a given structure. This could help the emergency organizations to know where severe damage
has occurred or if users of a specific infrastructure have to evacuate or need to request a professional to
evaluate. In the industry this information helps, but also sometimes they need to act quickly and
automatically perform emergency procedures. Structural Health Monitoring (SHM) is not something new,
it has been used in aeronautical, electrical and mechanical system for some time now. Its introduction in
the civil engineering structures is new and limited, with few exceptions like Japan, (Kanda et al. 2021).
What is Structure Health Monitoring?
Structure Health Monitoring (SHM) is a methodology that allows to inform users, owners and the
community of the condition of the civil infrastructure and can help manage it. It is associated with
different activities or concepts (some of them redundant), like:
1. State determination
2. Asset-life cycle management,
3. Condition assessment,
4. Performance assessment,
5. Integrity management,
6. Performance management.
In our case, we do it by installing sensors that can communicate, their sensing physical variables
continuously, semi continuously or by event trigger to a local or a remote processing system. The data is
automatically analyzed using different techniques to estimate the state of the system, (Ruben Boroschek
and Santos 2020; Brincker and Ventura 2015; Ebrahimian et al. 2018; Magalhães, Cunha, and Caetano
2009). This state is communicated to stakeholders using protocols that are related to their ability to
understand the information. This information could be very simple for normal users (street like signal,
color tagging, etc.) or detailed information for technical communities. A key aspect is that all this process
is done automatically. Automatization is one of the most important characteristics of SHM. If we have
many sensors and a large number of structures, we cannot expect that the state of health is evaluated by a
human. It has to be automatic, and it has to be reliable and understandable by the user.
There are some groups that also require to automate actions associated with each predefine state of the
system. For example, in industrial facilities, where highly complex or dangerous procedures are needed to
be stop quickly to avoid an environmental damage or an explosion. This is not the typical shutoff valve
that we require in several of our codes, because this activity is taken based on the response of the system
157
and its state of health. The criteria to activate the protocol must be defined prior to the occurrence of the
phenomena.
What is the difference between SHM and SSHM?
SHM has been around for several years in several technical aeras. Seismic SHM (SSHM or S2HM) focus
on the seismic environment. Special techniques and professionals should be involved. The earthquake is a
relative short time, non-repetitive condition, and the response of the system is in general nonlinear. This
situation is very different from other actions like usage and environment where there are some
nonlinearities in the system, but these are generally constrained. Also, in this case the amount of data is
extensive. Nevertheless, S2HM also uses data and procedures from standard SHM to confirm and
strengthen its evaluation.
Base on (Rytter 1993) there are at least four objectives for SHM:
Detection. This is the possibility to detect a change of state. I don't use the word damage because
sometimes this change of state is due to changes in operation or environmental conditions. In general, we
need more information to stablish damage. For buildings under seismic actions the probability of
detection (POD) will depend on the number and type of sensors and if the consequences of the damage
can be observed at local or global level. When it affects the global response, a limited number of sensors
is needed. If this change of state is very local in its effects, then you will need to understand quite well the
structure behavior and you will need to locate the sensors close to the area of damage. In general, for
earthquake conditions and buildings few sensors will give you an information that the structure has
change before and after a known event. This could be done by evaluating changes in modal properties
(periods and mode shape) or in certain response parameters (example tilting or rotations).
To locate the origin of change of state in general you will need several sensors together with numerical
models. Some of these sensors will need to be close to the areas of possible damage. These areas could be
established using calibrated numerical models.
Quantification is to know the severity of the damage. This is in general done using localized sensors and
also with the help by numerical models.
Prognosis is the capacity to predict the remaining life of the structure, based on the damage that has
occurred. In general, this is not possible in civil infrastructure because we don't have enough statistical
information to predict the life span of a structure based on the damage that has occurred on identical o
similar structures.
In more practical terms in the case of buildings and earthquakes we would like that SHM is able to
answer the following questions, (base on discussion of the COSMOS SHM Guidelines working group).
Do I need to work evacuate? Do I need to call for inspection? If I have a stock of infrastructure, how can I
prioritize the inspections considering my limited resources? Can I automatically tag my structure (green,
yellow or red)? Can I help localize the possible areas of damage so, we don't need to remove most of the
architectural components to evaluate the structural system?
Type and number of sensors
The type of sensor, its quantity and location will depend on the objectives and the complexity of the
structure. Several definition and conditions need to be stablished for the monitoring systems. The most
common ones are:
158
Resolution which is the level of vibration, deformation or force that can be detect beyond the noise or
distortion of the monitoring system. For example, if we want to monitor displacements and drifts in a
building using accelerometers, we will usually need a high-resolution of 1 or 10 x 10-6g to obtain
dynamics or residual (final static) displacement. Of course, you can measure displacement directly using
other techniques, but they are not too common in buildings.
Robustness is also very important because you will need to take a decision between having a very robust
system that needs very little maintenance, if your system is critical, have high costs for personnel or it's
very difficult to perform maintenance. On the contrary the if the installation is not so critical, and the
personnel can perform easily and quickly repair and maintenance you could risk installing a not so and
reliable expensive system.
Synchronization between sensor is a common need as well as communications between components to
the outside world. These are generally obtained using cable or wireless system and at the end they are a
major source of system vulnerability. Typically, internal and between sensor synchronization should have
errors lower than 1 ms. and no loss when transmitting data.
Of course, you need to consider the cost, but not only the initial cost, but also the maintenance cost.
Sometimes the initial cost is low compared to maintenance cost, especially if you think that these systems
should last for several years.
Another decision that you will have to do for the system is to know if you're going to have a local, at each
sensor location, energy, processing and communication systems or a central recording unit that could
provide everything and it is in charge of all communications and if it will send all data to the cloud for
calculations or it will do it locally or a mix system.
For example, if the strategy relies on the change of modal properties as indicator of the state of your
structure, then just very few sensors maybe use. One triaxle sensor (I don't recommend this) could be
enough. If you want to track changes of mode shape, then you will need a more extensive setup. But bear
in mind that any, well selected and located sensor give you information. The worst condition is no sensors
at all (Dr. K. Mosalam COSMS SHM Guidelines).
In Chile we have a 24 stories office building where we installed only one triaxle sensor ate the top. It is a
low-cost sensor 16 bits connected to the building server (that has emergency power and other backup
systems), (R. Boroschek, Villalpando, and Peña 2019). With this system we can use the ambient vibration
during business hours and the earthquake events to determine state changes, Figure 2. The sensor low
sensitivity does not allow modal parameter determination during nighttime o weekend periods. The day
information plus earthquakes that are recorded approximately every three months make it possible to
detect a change of global parameters. We can identify global changes but is not necessarily robust, this is
only the first step for valuable information. For this building no alerts are communication to stakeholders,
it is used to validate technology.
As indicate by Dr. K. Kusunoki (COSMOS SHM Guidelines working group) having one sensor at the
ground level to obtain records that feed a structural design model to predict damage should not be called
Structural Health Monitoring, cause actually you're not monitoring any response structural response. But
again, this could generate valuable information.
Several sensors are required if you want to localize and quantify damage and computer models. Base on
the simulation to different scenarios we can define where the accelerometer, strain gages, load cell,
displacement meters or video camaras should be deployed to get the desire information.
159
Figure 2. Frequency determination using a single low-resolution sensor in

an office building. Frequency is determined from ambient vibrations using
for the OMA. Some parameters are identified above noise level during
working hours.
At the University of Chile, we are monitoring the Central Building of the Engineering School since
2009. We continuously monitor ambient and seismic vibrations. So long we have recorded more than
1500 events and some of them quite severe that caused damage. There are 8 sensors in the network, so
modal tracking and some level of information on drift can be obtain (the structure has torsion so only two
level are monitored)(Ruben L. Boroschek and Carreno 2011). We do an online automatic modal
parameter tracking every 15 minutes. Figure 3 show the evolution of nine predominant frequency from
2009-2014, the variation of modal properties due to the change on environmental and use conditions are
strong. It is also quite obvious the strong change of properties due to the damage that occur during the 8.8
Magnitude earthquake in February 2010, (R.L. Boroschek and Bilbao 2019; González, Boroschek, and
Bilbao 2021). Additionally, we have use wave passage techniques to determine the change of state and
possible location of damage, (M.Rahmani, T.-Y.Hao, M.Todorovska 2017). In this case again there is no
active communication to the users of the building only to the operators of the system.
Interstory drift is another common technique used to estimate damage and should be closely related to the
number and quality of sensors, (Kanda et al. 2021; Naeim et al. 2006; Skolnik and Wallace 2010). In
general, we do not obtain drifts directly but by double integration of acceleration records. To do this we
need high quality system and integration procedure. These high-quality data will relate to the sensor
stability and resolution, particularly if the state determination will be complemented by residual
displacements and tilting at the sensor location. The drift derived from the records can be compare with
design parameters, statistical values for this type of structures, code reference or others.
If a computer model or drawings of the structure to be instruments are available, then you could use them
to estimate the best position for the sensors base on available budget and objective requirements. We need
first to know the effect of irregularities and torsion, if they are important on the response, you will be
forced to use a dense horizontal and vertical distribution of sensors to capture their effects. If irregularities
are not critical, maybe we can use a line of sensors that will capture the global properties and adequate
160
resolution of the structure drifts. There are many works that study the minimum number of sensors to
capture drift and to use it as damage index. Most of them conclude that a minimum of four levels of
bidirectional sensors for a typical 15 story regular structure is needed. In general, you will need the
ground and a higher elevation sensor and at least one on the first two floors and then others distribute on
height.
Figure 3. Frequency tracking for Torre Central Building University of

Chile. Nine frequencies are identified most of the time. The wandering of
its values is related to ambient, use and damage.
In another case strong motion monitoring is located only at the interface level in a base isolated data
center, top and bottom of the isolators. The system records earthquakes and determines the response
parameters and amplitude dependent equivalent structural periods, Figure 4. The system is uses to detect
changes of stiffness due to a blockage of the seismic gap. This system automatically informed the
response parameters to predefined stakeholders.
Figure 4. Displacement versus period obtain from seismic records on a base

isolated structure.
161
Another example is a building subjected to blast and seismic actions, (R. Boroschek, Villalpando, and
Peña 2019). It is located in a mine, very close to the blasting area. This building was monitored
continuously for two years. There was a blast every two days and some earthquakes occur during the
monitoring. The objective of the monitoring was to determine the existence and the location of damage 40
minutes after each blast. The blasting generated in the building accelerations that can go up to 2g. The
frequency characteristic of the blasting signals generated higher mode vibrations, associated with slabs,
columns, diagonals and beams local motions. The initial procedures to define the system was to perform
ambient vibration studies for different operational and ambient conditions. This structure is in the high
Andes, so it has high or very low temperature with high snow loads, so the model properties do change
depending on the external conditions. Several models were developed to capture the observed variability.
These numerical models were used to evaluate the response and possible damage scenarios based on the
characteristic of the motions and to determine different performance levels for this structure. Based on
these performance levels, several limits were established together with the sensing system. The system
was highly redundant because we could not fail to respond in 40 minutes. So, everything was automatic
from the recording, data transmission, analysis of the response and the generation and delivery of a report.
The report contained information derived from the data itself but also from the numerical models that
were running in parallel. The report indicated where and which elements to inspect. This process occurred
every two days for nearly two years, and we were always able to the deliver the report accurately and
typically in 20 minutes.
Figure 5. Frequency evolution of ambient vibration. Variations are

associated with environmental and response conditions.
During the construction of a 54 story 200-meter-tall building we installed an automatic system that
captured its response and model properties from construction and seismic vibrations, (Nunez, Boroschek,
and Larrain 2013). Initially, during the first ten stories the monitoring was done by a human a group.
Later we install an automatic system that evaluated the properties of the structure every 15 minutes,
Figure 6. Every week, that corresponds approximately to the finishing of a level, the modal data was
compared with the design model at the given stage. This was carried out during all the construction
process. Just at the beginning, there was some discrepancy between the numerical models and
experimental results. Discussion was carried on; the conclusions were that no action was needed. The
building was monitored till the end of construction, given confidence to the owner, the designer and the
162
contractor that the predicted design model was accurate enough, at least in the linear range. The
monitoring finished in January 2010 and at the end of February the strong 8.8 magnitude earthquake
occurred. The building suffered some damage. There was need for inspection, but the extensive and
expensive architectural finishes already cover most of the structure. Because we have the characteristic of
the building before the earthquake, we went back and recorded ambient vibrations with a high spatial
resolution. The analysis of the data indicated that no strong damage was observed on the structure.
Figure 6. Comparison between experimentally determine first mode frequency and numerical model.
The continuous drop of predominant frequency as the building increased its height is event and also
the excellent agreement between the design numerical model and the experimental data. This is due
to the strong presence of shear wall, limiting the effect of nonstructural components and the stiff soil
at site.
Conclusions
In Chile several systems are and have been monitored, not all of them reach the expected final results of
Structure Health Monitoring, that is automatic alerting and actions. Nevertheless, we have examples at
different stages of development, and we have explored its possibilities. From just continuous monitoring
and human analysis to automatic reporting of damage and its location, SSHM has proved to be a very
valuable tool, but it's still not fully expanded. There are many reasons for that. One of them is the lack of
standards or guidelines, COSMOS and other organizations are already working on them. These
documents could provide the platform to create confidence to the owners and users. The other mayor
problem is the limited number of human resources with the knowledge to apply this technology in a
proper manner. There are others that you can see in (Cawley 2018).
References
Boroschek, R., P. Villalpando, and E. Peña. 2019. “Building Structural Health Monitoring under
Earthquake and Blasting Loading: The Chilean Experience.” In Springer Tracts in Civil
Engineering, Springer, 361–83.
Boroschek, R.L., and J.A. Bilbao. 2019. “Interpretation of Stabilization Diagrams Using Density-Based
Clustering Algorithm.” Engineering Structures 178.
163
Boroschek, Ruben L., and Rodrigo P. Carreno. 2011. “Period Variations in a Shear Wall Building Due to
Earthquake Shaking.” SHMII-5 - 5th International Conference on Structural Health Monitoring of
Intelligent Infrastructure (December).
Boroschek, Rubén L., Víctor Contreras, Dong Youp Kwak, and Jonathan P. Stewart. 2012. “Strong
Ground Motion Attributes of the 2010 Mw 8.8 Maule, Chile, Earthquake.” Earthquake Spectra 28:
19–38.
Boroschek, Ruben, and Rodrigo Retamales. 2016. “Chile Resiliency : A Review of the Housing and
Health Sectors.” In 1st International Workshop on Resilience, ed. G. Cimellaro, G., Caverzan, A.,
Tsionis, G. and Solomos. Turin, Italy: Publications Office of the European Union, 4.
Boroschek, Ruben, and Joao Pedro Santos. 2020. “Civil Structural Testing.” In Handbook of
Experimental Structural Dynamics, Springer New York, 1–95. https://doi.org/10.1007/978-1-4939-
6503-8_29-1 (February 9, 2021).
Brincker, Rune, and Carlos E. Ventura. 2015. Introduction to Operational Modal Analysis. Wiley.
Cawley, Peter. 2018. “Structural Health Monitoring: Closing the Gap between Research and Industrial
Deployment.” Structural Health Monitoring 17(5): 1225–44.
Ebrahimian, Hamed, Rodrigo Astroza, Joel P. Conte, and Costas Papadimitriou. 2018. “Bayesian Optimal
Estimation for Output-Only Nonlinear System and Damage Identification of Civil Structures.”
Structural Control and Health Monitoring 25(4): 1–32.
González, Wladimir, Rubén Boroschek, and Joaquín Bilbao. 2021. “Modal Tracking of an Instrumented
Building Assisted by Temperature Measurements.” Engineering Structures 233(April).
Kanda, Katsuhisa, Masayoshi Nakashima, Yoshitaka Suzuki, and Saori Ogasawara. 2021. “‘Q-NAVI’: A
Case of Market-Based Implementation of Structural Health Monitoring in Japan.” Earthquake
Spectra 37(1): 160–79.
Lagos, R. et al. 2020. “The Quest for Resilience—The Chilean Practice of Seismic Design for Reinforced
Concrete Buildings.” Earthquake Spectra.
M.Rahmani, T.-Y.Hao, M.Todorovska, R.Boroschek. 2017. “Structural Health Monitoring of Torre
Central By the Wave Method.” (January).
Magalhães, Filipe, Álvaro Cunha, and Elsa Caetano. 2009. “Online Automatic Identification of the Modal
Parameters of a Long Span Arch Bridge.” Mechanical Systems and Signal Processing 23(2): 316–
29.
Naeim, F. et al. 2006. “Three-Dimensional Analysis, Real-Time Visualization, and Automated Post-
Earthquake Damage Assessment of Buildings.” Structural Design of Tall and Special Buildings
15(1): 105–38.
Nunez, T.R., R.L. Boroschek, and A. Larrain. 2013. “Validation of a Construction Process Using a
Structural Health Monitoring Network.” Journal of Performance of Constructed Facilities 27(3).
Rytter, Anders. 1993. “Vibrational Based Inspection of Civil Engineering Structures.” Aalborg
University, Denmark.
Skolnik, D. A., and J. W. Wallace. 2010. “A Critical Assessment of Interstory Drift Measurements.” 9th
US National and 10th Canadian Conference on Earthquake Engineering 2010, Including Papers
from the 4th International Tsunami Symposium 5(1388): 3631–40.
164
LESSONS LEARNED FROM SIX-YEAR EXPERIENCE

ON MARKET-BASED IMPLEMENTATION OF SSHM NAMED “q-NAVI”
Masayoshi Nakashima, President

Kobori Research Complex Inc.
Katsuhisa Kanda, Principal Engineer
Saori Ogasawara, Senior Engineering
Yu Fukutomi, Engineer
Kobori Research Complex. Inc.
Abstract
This paper introduced the development and actual implementation of a “market-based” SSHM applied to
Japanese buildings. The SSHM introduced in this paper, named q-NAVI, began its instrumentation in
2015 and has been installed currently to 490 buildings throughout Japan. All expenses associated with q-
NAVI are paid by private budling owners and operated and maintained by a private firm through the
contract between the owner and the firm. First, a summary of the q-NAVI system is introduced. Next,
experience accumulated for the past years is presented in terms of: (a) tangible benefits felt by building
owners (remote service and backup UPS), (b) tangible benefits to earthquake engineering (self-learning
from actual events: updating the threshold values and accumulating data on damage to nonstructural
elements), and (c) possible benefits in the future (verification of actual performance relative to designed
performance and city/community (regional) damage evaluation). Last, the most important lesson learned
through the interaction with building owners and their representatives is identified. They expect that q-
NAVI should deliver the proper diagnosis of Safe (Green), Caution (Yellow), and Danger (Red) most
promptly. They do not expect q-NAVI to locate detailed (member-level) damage location and severity as
they believe that the engineers they contract should conduct such a task in conjunction with the repair
work.
Introduction
Seismic structural health monitoring (named “SSHM” hereinafter) began in Japan in the 1950s mainly
based on “academic insensitive. Its implementation grew with time but in a rather slow speed. A
summary on the history of Japanese SSHM is presented elsewhere (Kanda et al., 2020). The attitude
toward SSHM changed after the 1995 Kobe Earthquake and succeeding earthquakes that hit various parts
of Japan. Those earthquakes highlighted the need for seismic observation to identify the level and
location of actual damage particularly in conjunction with “damage detection” and “judgement for
evacuation.” Yet, the growth in actual implementation of SSHM remained relatively slow.
Finally, a notable change occurred after the 2011 Tohoku Earthquake. In the earthquake, 5.15 million
people in the Tokyo metropolitan region had trouble returning to their homes after the earthquake, and
pedestrians overflowed roads, which created a very serious problem throughout the Tokyo metropolitan
region. The Tokyo Metropolitan Government enforced an ordinance in April 2013 (Tokyo Metropolitan
Government 2013) that, for the sake of safety, people should remain in their buildings, when possible,
rather than leaving them. The ordinance states that, when a large shaking occurs, building managers
should confirm the safety of their buildings and neighborhoods. If safety is confirmed, people in the
165
building are instructed to wait in it. If a large city experiences serious shaking, however, many buildings
will be damaged at the same time, and there will be a serious shortage of engineers who can check the
safety of buildings.
Such difficulties suggest the potential for using SSHM for the prompt assessment of building safety,
particularly in large metropolitan areas. Building owners, particularly those who manage many buildings
and are keen on business continuity planning (BCP), have come to understand this advantage and began
installing SSHM in the buildings that they own and/or manage. This new trend differs notably from the
movement we saw before the 2011 Tohoku earthquake. The installation of SSHM is driven primarily by
business and market forces rather than by public expenditure. Before the earthquake, the number of
instrumented buildings in Japan was in the range of 150. Since that time, the number increased to about
500 in 2016. Although most recent statistical data are not available for instrumented buildings, a
reasonable estimate for the number is about 950 in total, among which about 750 are privately owned.
In this article, we report on an example of the “market-based” SSHM applied to buildings, particularly on
the experience accumulated and lessons learned from the interaction with the building owners and
managers. As anticipated easily, most experience and lessons are non-technical, but such non-technical
matters are likely to be the genuine key for promoting the SSHM. The presented “market-based” SSHM
is a system, designated as q-NAVIGATOR (or q-NAVI in short). It was developed in 2014, and its
installation began in 2015. As of September 2021 (six years and a half after the first installation), a total
of 490 buildings, all privately owned, have been equipped with q-NAVI. The number constitutes about 65
% the privately-owned instrumented buildings in Japan. In what follows, the outline of q-NAVI is
described. Then, a few tangible benefits are noted that the building owners and their representatives felt
and appreciated after actual earthquake events. It is followed by several tangible benefits that the
earthquake engineering community would receive from SSHM. Last, the most important lesson that we
have learned from the building owners and their representatives is presented.
Outline of q-NAVI
The description of this section is an excepted version of the previous publication, entitled: “q-NAVI: A
Case of Market-Based Implementation of Structural Health Monitoring in Japan”, and published in
Earthquake Spectra, Earthquake Engineering Research Institute (Kanda et al., 2020).
System Configuration
The standard SSHM system adopted in q-NAVI is shown in Figure 1, including multiple accelerographs
(simply called sensors hereinafter) that measure the vibration of floors in the building (the number and
location of the sensors will be shown later), a PC for analysis, a monitor screen that displays the results,
and an Uninterruptible Power Supply (UPS) for supplying power at the time of power loss. The PC and
sensors are custom-made, as their performance and quality are supposed very critical. The PC is the one
adopted commonly for operation of industrial machines instead of those used in the office. Other devices
except for the PC and sensors are off-the-shelf and the ones broadly distributed in the market. The HUB
adopted in the system is Power-over-Ethernet compatible, supplying signals and electricity to the sensors
via LAN cabling, and therefore a power supply is not needed for each sensor. The sensors are commonly
installed on the floor slab in an electric pipe shaft (EPS) that runs throughout the floors. In building
structures, the EPS is a very convenient space to install the sensors because it allows handy vertical
cabling and positioning of the sensors at the same location along the height. With this handy EPS
available, it normally takes one or two days to complete the installation of sensors and LAN cabling
The level for triggering the recording is set up for the base floor and at 1.5 in JMA (Japan Metrological
Agency) Intensity (equivalent to about 20 mm/s2 (0.002g) acceleration). The definition of JMA Intensity
is found in JMA (1996) and Campbell and Bozorgnia (2011). Once it is triggered, recording begins, the
166
maximum interstory drift ratio and the maximum floor acceleration are computed by the PC, and one of
the three diagnoses, “Safe (Green)”, “Caution (Yellow)”, or “Danger (Red)” is announced as commonly
done in many emergency risk judgements. “Safe (Green)” means that one can stay in the building, the
building is not damaged, and it can continue to be used. “Caution (Yellow)” means that some damage is
plausible, the damage shall be checked, but there is no risk of building collapse and thus no immediate
evacuation is needed. “Danger (Red)” means that the building occupants should be evacuated
immediately as there is a risk of collapse during aftershocks. Once the SSHM system is triggered, a series
of displays, i.e., detection of shaking, JMS Intensity, and diagnosis, appear in the PC screen. The
diagnosis screen highlights important information such as the date of occurrence, the level of shaking at
the base floor, statuses of interstory drift ratio and maximum floor acceleration for each story/floor, and
most importantly the diagnosis. The on-site building managers can see the screen, quickly understand the
building status, and deliver the message of action to the building occupants. The diagnosis appears in the
screen commonly within one to two minutes after the shaking.
Figure 1. Outline of system configuration adopted in q-NAVI.
Type and Number of Sensors
The type of sensors to install and the number of sensors to deploy are a matter of controversy all the time.
In q-NAVI, we have adopted the following strategy, that is, (1) using (relatively) robust, high-accuracy
sensors (rather than inexpensive sensors like MEMS accelerometers); (2) using “wires” for the transfer of
the sensor data to the PC; (3) deploying a triaxial sensor (that can collect the vibrations in three
directions) in every few stories, instead of placing one per story or placing multiple sensors in a single
story. Strategy (1) was adopted so that the floor acceleration records can reasonably be converted to the
corresponding floor displacements. Strategy (2) was adopted with the conviction based on experience that
the current wireless sensors have difficulties in not a few cases in delivering the data in the (almost) real-
time, which will be fatal for the most important “immediate diagnosis.”
Strategy (3) was adopted to reduce the installation and maintenance cost as much as possible. Torsional
motion was assumed to be very minor, but it is justified, at least in Japan, because the Japanese seismic
code is very stringent against torsional response and therefore Japanese buildings commonly have minor
torsional behavior regardless of the building height. The vertical spacing of sensors significantly affects
the total number of sensors needed for a building. A conventional choice is one tri-axial sensor per story
167
with the assumptions of no-torsion and rigid diaphragm. However, q-NAVI adopted a rule such that the
deployment of four tri-axial sensors is the standard case for mid-rise office buildings of around 10 stories.
As for the interpolation of interstory drift ratios, a simple linear interpolation is used for low- to mid-rise
buildings, and a cubic-spline (Naeim et al. 2006) or a mode-based (Suzuki et al. 2008, Goel 2008)
interpolation is adopted for high-rise buildings.
Characterization of Diagnosis
The boundaries among “Safe (Green)”, “Caution (Yellow)”, and “Danger (Red)” are determined about
the maximum interstory drift ratios, which vary according to the type of structure. The boundary between
“Safe (Green)” and “Caution (Yellow)” is commonly set at: 1/250 – 1/150 for RC and SRC (steel-encased
concrete) frame structures, 1/350 – 1/200 for RC structures with shear walls, 1/220 – 1/100 for steel frame
structures, and 1/300 – 1/180 for steel structures with braces. The boundary between “Caution (Yellow)”
and “Danger (Red)” is commonly set at: 1/150 – 1/60 for RC and SRC frame structures, 1/250 – 1/120 for
RC structures with shear walls, 1/100 – 1/60 for steel frame structures, and 1/150 – 1/80 for steel
structures with braces. Standard values for respective categories have been determined in reference to the
past data and analysis on damaged buildings collected in post-earthquake reconnaissance and risk
evaluations. When setting up the interstory drift ratio limits, design documents of respective buildings are
scrutinized. The drift limits are adjusted for each building in light of various guidelines and specifications
and also considering inherent “uncertainties”. These drift ratio limits are sometimes re-adjusted after
significant shaking, as will be presented in a later section.
As supplemental information, statuses for possible damage to nonstructural components are displayed on
the PC screen. The maximum floor acceleration is used for indirect estimation of such damage, in which
empirical relationships between the maximum floor acceleration and the degree of damage to
representative nonstructural components and building contents are used. The screen highlights a warning
sign (and suggests human inspection) for stories with floor accelerations exceeding the threshold values.
Remote Systems Maintenance and Internet Browsing System Using Cloud Monitoring Service
Because of the nature of the service supplied by q-NAVI to the building owners and their representatives,
its operation must be most reliable during an earthquake. To this end, the SSHM system enforces remote
and continuous maintenance. That is, operation monitoring is performed constantly and automatically
from a remote server. If troubles such as communication errors are found, the related devices are reset. If
a device failure is confirmed, the device is promptly replaced by a human.
When an earthquake occurs, recorded data and analyzed results are automatically uploaded on a remote
monitoring server using a remote maintenance line. These data are also transferred and stored on a cloud
server for backup (Figure 1). Furthermore, a cloud service is designed to be available for the building
owners and their representatives, by which the backup and evaluation data can be viewed collectively via
the Internet. This has been found to be a very convenient service to building owners and their
representatives, and an example will be noted in a later section.
Past Experiences
The installation of q-NAVI began in 2015, and as of September 2021, a total of 490 buildings have been
equipped with q-NAVI. Naturally, many of them have been installed in large cities such as Tokyo (141
buildings) and Osaka (56 buildings). Still, they have spread across the country from Hokkaido (the
northern region) to Kyushu (the southwestern region). Figure 2 shows the plots of those 490 buildings on
a map of Japan. Figure 3 shows the number of buildings equipped with q-NAVI, categorized with respect
to the year of construction, the type of structural material (RC, steel, and SRC (steel-encase RC)), and the
number of stories. It is notable that quite a few old buildings, for instance, 77 buildings that are more than
168
40 years old, have had q-NAVI installed. Notably, the incentive of SSHM is more toward old buildings
constructed to obsolete design codes because the building owners have greater concerns about such older
buildings. As a result of this incentive, a large portion of q-NAVI installation is for existing buildings.
Figure 2. Installation of q-NAVI in buildings in Japan.
150 200
RC RC
SRC
Number of buildings
Number of buildings
150 SRC
100 S S
100
50
50
0 0
Year of construction
(a) (b)
Figure 3. Numbers of buildings equipped with q-NAVI, presented for three structural materials (RC,
SRC, and steel): (a) Year of construction. (b) Number of stories.
Figure 4 shows the earthquakes that had shaken Japan since 2015 when q-NAVI began. The diameter of
the circle that shows the epicenter indicates JMA Magnitude MJ, and the observed earthquakes range for
4.0 ≤ MJ. The figure includes 753 plots for 4.0 ≤ MJ and seven plots 6.0 ≤ MJ. It indicates that Japan has
been shaken frequently and at relatively large magnitudes.
Figure 5 summarizes how often the buildings installed with q-NAVI were shaken by the earthquakes
shown in Figure 4. The curve drawn in blue (plotted in reference to the right vertical axis) shows the
cumulative number of buildings in which q-NAVI was activated and the building response was recorded.
The plot is made for the year (from May 2015, when the q-NAVI operation began to June 2021). Note
that the recording is triggered for a base floor of 1.5 in JMA Intensity (equivalent to about 20 mm/s2
(0.002g) acceleration). A total of about 15,000 cases were recorded over the past six years. In the same
figure, bars plotted about the left vertical axis show the numbers of buildings in which q-NAVI detected
serious shaking whose magnitude at the base floor is 4 or larger in terms of Japan’s JMA Seismic
Intensity (IJMA). JMA Seismic Intensity values of 4, 5-lower, 5-upper, 6-lower, and 6-upper correspond
roughly to Modified Mercalli Intensity (IMM) values of V, VI, VII, VIII, and XI (Campbell and Bozorgnia,
169
2012). The figure indicates that the number of buildings subjected to severe shaking varies significantly
year by year. Still in some years, the number exceeds 20 (not smaller than VI in IMM) even within the
vintage of buildings equipped with q-NAVI.
M7.0
M6.0
M5.0
M4.0
Figure 4. Major earthquakes that have occurred in Japan since 2015.
80 20,000
6+
In which q-NAVI was triggered

70
6
Total number of buildings

Number of buildings
60 5+ 15,000
50 5-
40 4 10,000
total
30
20 5,000
10
0 0
Figure 5. Numbers of buildings in which q-NAVI was triggered.
Tangible Benefits to Building Owners and Their Representatives
Remote Service
As noted earlier, the data recorded by q-NAVI are sequentially uploaded to the remote server. The time
lag from the occurrence of earthquake to the display through the Internet was about five to at most 20
minutes. Three factors contributed to the time delay: first, the post-trigger waiting time to determine the
end of earthquake, along with the data processing time within the PC in q-NAVI; second, the
communication time from the mobile phone router installed in the building to the remote data server; and
third, the data processing time, in which the remote server sets up the display data to the cloud service.
The building managers can access the Internet shortly after the main shock and see those data via the
cloud service. The 2018 Osaka earthquake occurred in the early morning (at 7:58 am on June 18, 2018),
around the time when most people started moving to their workplaces. Public transportation stopped
immediately after the earthquake, so owner representatives and building managers could not go to the
buildings that they manage and maintain. Nonetheless, they were able to collect information on building
170
status very quickly via the cloud service and take appropriate actions without delay. This service was
appreciated very much by the building owners.
Backup UPS Batteries
In the 2018 Hokkaido Eastern Iburi earthquake (at 3:07 am on September 6, 2018), all 25 buildings
equipped with q-NAVI were judged “Safe”. However, most of the Hokkaido region blacked out
immediately due to the shutdown of a thermal power plant at Atsuma town. The q-NAVI system
continued to work for about one hour until the Uninterruptible Power Supply (UPS) batteries ran out.
Although some aftershock records were missing, it was confirmed that the entire shaking during the
mainshock was recorded, the resulting judgment was displayed on screen for all the buildings, and the
data were uploaded to the cloud server without malfunction. Although the monitor vanished after the UPS
battery ran out, the cloud service remained functional and experienced no operational problems.
Redundant cloud servers had been established in two distant places to make the adopted SSHM system
robust. It was also confirmed later that the entire SSHM system restarted without problems after the
power recovery. The importance of UPS has been known widely as a necessary means at times of unusual
events such as earthquakes. However, sensing this benefit remains very seldom. This Iburi event was one
of such very rare cases, but once it was recognized, the building owners and their representative fully
appreciate this equipment.
Tangible Benefits to Earthquake Engineering
Self-Learning from Actual Events and Update of Threshold Values
As might be expected, determining the threshold values among the three categories is rather difficult,
particularly for old buildings. This is because they tend not to be as ductile as those built to recent design
codes, the properties of building materials are more uncertain, the effects of nonstructural elements on the
building’s strength are difficult to quantify, and experimental data on such old buildings is scarce. Typical
examples are few-story old RC or SRC building built in the 1970s (see an example in Figure 6), before
the major overhaul of Japanese seismic design in 1981. In the RC building, all exteriors are filled with RC
walls with openings of various shapes (for windows and doors), which produces spandrel walls, lintel
walls, and wing walls. They are monolithically connected to the surrounding columns and beams, and
susceptible to damage from stress and strain concentrations at corners. The design did not consider such
concentrations of strain seriously, and very little effort was spent on ensuring ductility in the walls. In
cases such as this, where there are many uncertainties, we must be significantly more conservative when
setting the threshold values.
171
Beam Column Wing wall
Spandrel wall
Figure 6. A schematic example of old RC building installed with q-NAVI
For RC structures like the one shown in Figure 6, default threshold values for the maximum interstory
drift ratio are 1/350 and 1/250 for the boundaries among Safe (Green), Caution (Yellow), and Danger
(Red). Those values have been set up to be significantly smaller than those threshold values of 1/200 and
1/120 stipulated modern RC frames with shear walls. The difference is nothing else but the conservatism
associated with uncertainties.
About 110 such RC or SRC buildings designed with obsolete codes have been equipped with q-NAVI,
and they underwent the earthquakes and ground motions shown in Figure 4. In two events, the 2018
Osaka earthquake and the 2021 Fukushima earthquake (at 11:08 pm on February 13, 2021), “Caution
(Yellow)” was announced for two such buildings. After the earthquakes, we visited the buildings in
question, carefully observed their damage, and confirmed that the maximum interstory drift estimated by
q-NAVI was reasonable though the observed damage remained relatively light. It was a result of
conservatism, as noted above. With the consent of the building owners, we decided to relax the threshold
values for those buildings. This experience is clear evidence that we can learn from actual earthquake
responses and adjust the threshold values that separate the damage status to more suitable ones. By
continuously accumulating such data from actual responses and classifying them with respect to the
building type, we will be able to establish more reliable threshold values specific to different building
types.
Accumulation of Data on Damage to Nonstructural Elements
Fortuitously and thanks to the generous assistance of the building owners, we had opportunities to closely
survey nonstructural damage to buildings that were equipped with q-NAVI and shaken in the 2018 Osaka
earthquake. Nonstructural elements were classified into eight types: Exterior walls, Window glasses,
Interior walls and partition walls, Doors, Ceilings, Furniture and building contents, Expansion joints, and
Air conditioners and plumbing elements, and the damage ratio was quantified with respect to the number
of buildings that sustained relevant damage. Since the buildings were equipped with q-NAVI, data on the
interstory drifts and floor accelerations were also available. Correlating the observed damage (by visual
inspection) with corresponding data on external action (by the q-NAVI data), we developed preliminary
“fragility curves” for some nonstructural elements. More details are presented in Kanda et al. (2020).
Accumulation of such data from actual nonstructural responses and classifying them with respect to the
element type will enable us to establish more reliable fragility curves, which are element-type-specific,
for nonstructural elements.
172
Expansion of SSHM Service in the Future
Verification of Actual Performance Relative to Designed Performance
By analyzing the building response observed by SSHM, we can evaluate how the performance predicted
in the phase of seismic design/analysis is different from the actual performance. Quantifying the
difference may also offer a sensible solution in disputes associated with unexpected damage to buildings.
The dispute may be about whether the designer is responsible for the damage by professional misconduct,
or the earthquake force to the building was larger than that stipulated in the design code and therefore the
designer is exempted from responsibility. Objective data acquired by SSHM shall ensure a scientific
basis for the final settling of such disputes. Although the motion is slow, such an evidence-based
approach is moving forward in the structural design profession.
City/Community (regional) Damage Evaluation
If a decent number of buildings are instrumented, and seismographs to record ground motion are densely
installed, we can combine them, interpolate the records, and predict the response (and damage) of all
buildings on the regional basis. Although yet rudimentary, such a method of diagnosis for regional safety
is being investigated. In the Tokyo metropolitan region, quite a few K-NET and KIK-net stations are
available. Furthermore, there is another dense seismograph array named the MeSO-net observation
network, which was established in 2009 (Kasahara et al. 2009). The total number of MeSO-net
seismographs is about three hundred, and they are arranged at intervals of 3 to 10 km. Those
seismographs are embedded 20 m in depth. On the building instrumentation side, over 80 buildings are
instrumented with q-NAVI in Downtown Tokyo (approximately an area of 80 km square). Figure 7
shows the observation points by q-NAVI in triangles and the observation points by MeSO-net in circles
(for those near buildings equipped with a-NAVI) and crosses (for the rest). It is notable that relatively
dense observations are possible in Downtown Tokyo. By interpolating the observation records obtained
by MeSO-net, we can estimate the input ground motion to the building where q-NAVI is installed, and
the relationship among the (estimated) input motion, the (measured) response at the base floor of
building, and the (measured) floor responses of the building is characterized. Meanwhile, fundamental
properties i.e., the height, type of structure, material adopted, year of construction, of individual buildings
(those not instrumented) in the concerned region can be collected and stored in a database. When an
actual earthquake occurs, these pieces of information can be combined, and the responses of un-
instrumented buildings are predicted, hopefully with an accuracy that can be trusted far more than the
accuracy by which we predict without such actual data. An actual application of such a procedure is
seriously argued in Japan (Kusaka, 2019).
173
Figure 7. MeSO-net observation network and neighboring buildings equipped with q-NAVI in Tokyo.
Lessons Learned from Interaction with Building Owners and Their Representatives
Through interactions with owners and owner representatives of the buildings equipped with q-NAVI, we
have reconfirmed that large corporations who manage many buildings most commonly have solid support
systems that can respond promptly to earthquakes and other disaster events. Maintenance people are
stationed permanently in each building. They continuously check the operation of the entire building,
including its mechanical systems, electricity, gas, water, and other utilities such as security and repair.
Furthermore, the building owners secure permanent contracts with construction firms that can come to the
building immediately after any inconvenience and prioritize fixing any problems.
A typical action of theirs immediately when their building sustained severe shaking is as follows. First, q-
NAVI is activated, and the first diagnosis is delivered to the on-site maintenance people who can watch
the q-NAVI screen. The diagnosis is also transferred to the owner-representatives via the cloud service. It
was followed by a quick survey of the building by the on-site maintenance people, who will report their
study to the owner’s representatives. Depending on the seriousness of damage, the owner’s
representatives contact the contracted engineering firm, and the firm dispatches its engineers to the
building. In terms of the timeline, the q-NAVI diagnosis is within a few minutes (the first phase). The
on-site survey is in a range of one hour after the event (the second phase). The off-site engineers arrive at
the building within half a day or one day at the latest (the third phase). This third phase is followed by a
more detailed survey of the damage, repair planning, procurement of human and materials for repair
work, and actual repair (the fourth phase and after), all of which are led by the contracted engineering
firm.
Having such human and material resources in hand, what the building owners and their representatives
expect most from q-NAVI is nothing else but the promptest message about the overall severity of damage
174
in the first phase of post-earthquake action. In this phase, they do not need to know detailed damage
information, or more frankly, they do not want to know the details. It is sufficient for them just to know,
Safe (Green, possible to stay as the damage is minimal), Caution (Yellow, possible to stay although some
damage occurs and must be fixed later), or Danger (Red, dangerous and should be evacuated).
Examinations of damage location and severity are commonly done in the second and third phases,
followed by specific repair. In all honesty, they do not expect q-NAVI for precise (say member-level)
damage detection because they feel that the contract engineering firm should carry out this task in
conjunction with the following repair work.
Conclusions
A summary and major finding of this article are as follows.
(1) A “market-based” SSHM, named q-NAVI, was introduced. It began its service in 2015 and has been
installed currently in 490 buildings throughout Japan. All expenses associated with q-NAVI are paid
by private budling owners, and a private firm performs operation and maintenance through the
contract between the owner and the firm. Unlike many other countries, many SSHM installed in
Japanese buildings are owned and operated by the private sector.
(2) From its first installation, Of all buildings equipped with q-NAVI, 753 cases were observed for the
shaking under earthquakes whose magnitude (Japanese JMA Magnitude) is not smaller than 4.0.
Seven cases were observed for the shaking under earthquake whose magnitude is not smaller than
6.0. Those numbers indicate that Japan has been shaken frequently and at relatively large
magnitudes. The q-NAVI system was triggered about 15,000 times when the shaking was recorded
and stored in the database. The number of severe shakings varies year to year. Still, in some years,
more than 20 cases were observed for the shaking of 5-lower or stronger in the Japanese JMA
Seismic Intensity (equivalent to VI in the Modified Mercalli Intensity).
(3) Notable and tangible benefits that building owners appreciated are the remote service and the backup
UPS. They also appreciated the cloud service as in one earthquake that occurred in the early
morning, all transportations stopped, and building managers could not inspect the buildings of their
management. The “Safe (Green)” signal they could check from the cloud service was a significant
relief when their mobility was impeded. In another earthquake, a massive blackout occurred
immediately after the event, but the backup UPS functioned very properly and q-NAVI could
continue to record the mainshock and aftershocks.
(4) Self-learning of q-NAVI from actual events is a notable benefit for earthquake engineering.
Specifically, the threshold values among Safe (Green), Caution (Yellow), and Danger (Red) could be
updated after examining the actual damage status. It is particularly relevant for old RC and SRC
buildings designed with obsolete codes, where their ductility capacity is hard to quantify.
Accumulation of knowledge on the progress of damage to nonstructural elements is another benefit.
(5) As possible expansions of SSHM to enhance its service, verification of actual performance relative
to designed performance is noted. It will offer a sensible solution in disputes associated with
unexpected damage to buildings. Another possibility is the damage evaluation of structural
performance on a city/community (regional) scale. An ongoing project is introduced, in which q-
NAVI and sensor networks that measure the ground shaking such as K-NET, KIK-net, and MeSO-
net are combined. This combination enables us to evaluate the damage of the building equipped with
q-NAVI and the buildings not installed with any SSHM.
(6) The most important lesson learned through the interaction with building owners and their
representatives is identified. The building owners expect q-NAVI to deliver the proper diagnosis of
175
Safe (Green), Caution (Yellow), and Danger (Red) most promptly. They do not expect q-NAVI to
locate detailed (member-level) damage location and severity. They believe that the engineers they
contract should conduct such a task in conjunction with the repair work.
Acknowledgement
The writers express their sincere gratitude to the building owners and managers who kindly shared the
data recorded by q-NAVI in the buildings that they own and/or manage.
References
Campbell, K. W., and Bozorgnia, Y., 2012. Cumulative absolute velocity (CAV) and seismic
intensity based on the PEER-NGA database, Earthquake Spectra 28, 457-485.
Goel, R. K. 2008. Mode-based procedure to interpolate strong motion records of instrumented

buildings, ISET Journal of Earthquake Technology, Technical Note, 45, No.3-4, September-
December, 2008, 97-113.
Japan Meteorological Agency (JMA), 1996. On Seismic Intensity, Gyosei, Tokyo, 46–224 (in
Japanese).
Kanda, K., Nakashima, M., Suzuki, Y., Ogasawara, S., 2021. “q-NAVI’’: A case of market-based
implementation of structural health monitoring in Japan, Earthquake Spectra, 37(1), 160-179.
Kasahara, K., Sakai, S., Morita Y., Hirata, N., Tsuruoka, H., Nakagawa, S., Nanjo, Z. K., and
Obara, K., 2009. Development of Metropolitan Seismic Observation network (MeSO-net) for
detection of mega-thrust beneath Tokyo metropolitan area, Bulletin of Earthquake Research
Institute, University of Tokyo, Vol.84, pp.71-88.
Kusaka, A., Hasegawa, K., Kanda, K., Kimura, T. and Sakai, S., 2019. Application of earthquake
observations on buildings to disaster response -Part 1 Study of estimation method of input seismic
motions to building with earthquake observations on buildings and MeSO-net, Paper No. 21293,
in Summaries, Technical Papers of Annual Meeting, Architectural Institute of Japan, September
3-6, 2019, Kanazawa, Japan (in Japanese).
Naeim F., Lee H., Hagie S., Bhatia H., Alimoradi A., Miranda E., 2006. Three-dimensional
analysis, real-time visualization, and automated post-earthquake damage assessment of buildings,
The Structural Design of Tall and Special Buildings, 15(1), 105-138.
Suzuki, Y., Koshika, N., Yamada, A., Orui, S., and Shimizu, K., 2008. Real-time building
damage estimation system based on observed building response, Paper No. 05-01-0439, in
Proceedings, Fourteenth World Conference on Earthquake Engineering, October, 2008, Beijing,
China.
Tokyo Metropolitan Government, 2013. The outline leaflet of TMG ordinance covering measures
for stranded persons, https://www.bousai.metro.tokyo.lg.jp/_res/projects/ default_project/_page_/
001/000/939/english_02.pdf, (last accessed 1 November 2019) (in Japanese).
176
NEW ACI PROVISIONS FOR SHEAR WALL DESIGN AND CASE STUDIES
John Wallace, Saiful Islam, Tony Ghodsi, Vladimir Volnyy, Rishabh Singhvi and Akshay Patil
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198

2021 LATBSDC Proceedings

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2021 LATBSDC Proceedings

Uploaded by

Copyright:

Available Formats

Proceedings of the 2021 Conference

THE 2021 LOS ANGELES TALL BUILDINGS

LOS ANGELES, CALIFORNIA

NOVEMBER 12, 2021

THE 2021 LOS ANGELES TALL

NOVEMBER 12, 2021

Tel: (626) 389-1888

Mr. Shahen Akelyan Dr. Ifa Kashefi Mr. Barry Schindler

Dr. Gregg E. Brandow Prof. Kristijan Kolozvari Prof. John W. Wallace

Dr. Lauren D. Carpenter Dr. Marshall Lew Mr. Philip Yin

Dr. Chukwuma Ekwueme Dr. Michael Mehrain Mr. Nabih F. Youssef

Mr. Tony Ghodsi Dr. Farzad Naeim Prof. Farzin Zareian

Dr. Saiful Islam Dr. Thomas A. Sabol Mr. Atila Zekioglu

Mr. John Gavan Mr. Richard Holguin Mr. Donald R. Strand

Dr. Robert E. Englekirk Dr. Sampson C. Huang Mr. Edward J. Teal

Mr. Colin Kumabe

November 12, 2021

Saiful Islam, PH.D., S.E.

545 South Figueroa Street, Suite 1223

CASE STUDY – SoFi STADIUM, INGLEWOOD, CALIFORNIA 13

SEISMIC RETROFIT OF A PRE-NORTHRIDGE STEEL MOMENT FRAME BUILDING 49

ASCE/SEI 41 ASSESSMENT OF REINFORCED CONCRETE BUILDINGS: 71

NEW NONLINEAR MODELING PARAMETERS AND ACCEPTANCE CRITERIA 72

ATC-145 UPDATE: DRAFT GUIDELINE FOR POST-EARTHQUAKE ASSESSMENT, 86

PHENOMENOLOGICAL NONLINEAR MODELING FOR PERFORMANCE-BASED 95

MULTI-PERIOD RESPONSE SPECTRA 110

SEISMIC STRUCTURAL HEALTH MONITORING AS AN ESSENTIAL INGREDIENT OF 144

CHILE S2HM EXPERIENCE 156

LESSONS LEARNED FROM SIX-YEAR EXPERIENCE ON MARKET-BASED 165

FINAL PROGRAM FOR THE 2021 CONFERENCE

CASE STUDY – THE GRAND LA

Patrick Lindblom, P.E., S.E., Scott Erickson, P.E., S.E.

Figure 1 – Architectural Rendering

Building Gravity Framing Overview

Figure 2 – Floor Plan Overlay

Firgure 4 – Podium Layering & Circulation

Figure 5 –Risk Category Diagram

Figure 6 – Coupling Beam S3 Rotational Response

Figure 7 – Wind Tunnel Testing

Building Information Modeling

very difficult to convey on 2D drawings in an efficient manner. As construction proceeded, Visicon

Figure 8 – Architectural Building Model in Visicon

Figure 9 – Embed and PT subcontrator coordination layout

Figure 10 – Sketch of Cinema Re-stack

Figure 11 – Floor Re-label

Figure 12 – Aerial photo of construction

Figure 13 – Cantilevering structural elements at floor transitions

March 19, 2020 October 20, 2021

CASE STUDY: SoFi STADIUM, INGLEWOOD, CALIFORNIA

Rafael Sabelli, Mark Waggoner and Ozgur Atlayan

SEISMIC RETROFIT OF A PRE-NORTHRIDGE STEEL MOMENT

Kevin O’Connell, S.E.

Photo 1. Exterior View of Building During Construction

Objective and Approach

Seismic Ground Motions

Structural Deficiencies and Seismic Retrofit

Deficiencies identified using LDP

Figure 3. Overview of the ETABS Model

Deficiencies confirmed using NSP.

Figure 4. Overview of the PERFORM-3D Model

Selected seismic retrofit.

We considered an approach that included enhancing existing beam-column moment connection by

FVD design and Nonlinear Time History Analysis

Figure 6: FVD Configuration (East and West Elevation)