Pci Journal - Winter 2012 Vol. 57 No. 1 PDF

Winter 2012
Volume 57, Number 1
Precast/Prestressed Concrete Institute
42 97 124
Earthquake Retrofit of Precast Panels
Reconnaissance Grouted Sleeve for Lateral-Force
Team Reports Connections Resistance
PCI Journal Reviewer Note of Appreciation
Every manuscript submitted to PCI Journal for publication as a peer-reviewed paper goes through a thorough tech-
nical appraisal. Specialists in the paper’s subject area conduct this review process.
Following are the members of this dedicated group of volunteer reviewers who in 2011 appraised prospective PCI
Journal papers. This year it was an especially demanding job as we worked to clear the backlog while also reducing
the length of time from submission to publication. PCI is grateful to each reviewer for the time and effort expended
in performing this valuable service to the engineering profession, the precast/prestressed concrete industry, and
ultimately the public.
Perry Adebar Brian Lee Lawrence

Rashid Ahmed Jonathan Lewis
Corina-Maria Aldea Jason Lien
Sameh Badie Malcolm Lim
Craig Barrett Zhu Liu
Evan Bentz Emily Lorenz
Shri Bhide Mohamed Mahgoub
JoAnn Browning Robert Mast
Larry Church Andrew Maybee
Adrian Ciolko Mike McCool
Ned Cleland Barry McKinley
Dave Darwin Alexander G. Mihaylov
David Deitz Antoine Naaman
Dan Dorgan Frank Nadeau
Sam Fallaha Claude Napier
Marcelo de Araujo Ferreira Charles Nmai
Robert Frosch Rita Oglesby
James Gallt Nabil Al-Omaishi
John Gardner Celik Ozyildirim
Franc Genoese Scott Phelan
James R. Gerloff Basile Rabbat
Jim Guarre Mary Lou Ralls
Simon Harton Sami Rizkalla
Bernie Hertlein Carin Roberts-Wollmann
Susan Hida Joe Rose
John Holt Anton Schindler
Venkatesh Iyer Steve Seguirant
Fouad Jaber Larbi Sennour
Bruce Johnson Halil Sezen
Bijan Khaleghi Johan Silfwerbrand
Yoo Jae Kim Irwin Speyer
Edward S. Knowles Eric Steinberg
Walter Korkosz Brian Swartz
Paul Kourajian Habib Tabatabai
Yahya “Gino” Kurama Teddy Theryo
Fadjar Kusumo-Rahardjo Wafeek Wahby
Bo Kusznir Nur Yazdani
Todd A. Lang Wael Zatar
Karen Laptas Paul Zia J
John Lawler
Table of Contents
Seismic
Preview of PCI’s New Zealand Earthquake Reconnaissance Team Report 42
42 Robert B. Fleischman, José I. Restrepo, Joseph R. Maffei, and Kim Seeber
Preview of PCI’s Japan Earthquake Reconnaissance Team Report 47

Richard Sause, Robert Frosch, S. K. Ghosh, Jason Lien, Clay Naito, Larbi Sennour, and Toshi Yamanishi
Observations from the February 27, 2010, Earthquake in Chile 52

S. K. Ghosh and Ned Cleland
Effects of Loading Conditions on the Behavior of Unbonded Post-tensioning 76

47 Strand-Anchorage Systems
Kevin Q. Walsh and Yahya C. Kurama
Seismic Performance and Retrofit of Precast Concrete Grouted Sleeve Connections 97

Andrea Belleri and Paolo Riva
Structural Behavior of an Innovative All–Precast Concrete Dual System 110

for Residential Buildings
Adrian M. Ioani and Eusebiu Tripa
52 Architectural Precast Concrete Panel Systems Used for Lateral-Force Resistance 124
J. Paul Hobelmann, Macarena Schachter, and Matthew C. Cooper
76
The PCI Journal (ISSN 0887-9672) is published quarterly by the Precast/
EDITORIAL DESIGN & PRODUCTION ADVERTISING SALES Prestressed Concrete Institute, 200 W. Adams St., Suite 2100, Chicago,
IL 60606. Copyright © 2012, Precast/Prestressed Concrete Institute. The
Rachel J. Detwiler, PhD, P.E. Editor-in-Chief Paul Grigonis Director, Publishing and Art Direction Kirstin Osgood Manager, Precast/Prestressed Concrete Institute is not responsible for statements
(rdetwiler@pci.org) (pgrigonis@pci.org) Sales and Member Development made by authors of papers or claims made by advertisers in the PCI Journal.
Michelle Burgess Managing Editor Ed Derwent Associate, Graphic Design (kosgood@pci.org) Original manuscripts and letters on published articles are accepted on
review by the PCI Technical Review Committee. No payment is offered. Direct
(mburgess@pci.org) (ederwent@pci.org) (312) 360-3206 all correspondence to PCI Journal at journal@pci.org or Precast/Prestressed
Carrie Wyrick Technical Editor Concrete Institute, c/o PCI Journal, 200 W. Adams St., Suite 2100, Chicago,
(carrie.wyrick@gmail.com) IL 60606. For information on advertising rates, send an email to adsales@
pci.org. Subscription rates are $70 per year and $175 for three years in the
Heather Green Copy Editor United States and $150 per year and $410 for three years for international.
Nyema Russell Administrative Assistant A single or back issue is $12. International subscriptions are delivered by an
(nrussell@pci.org) international carrier; allow one to three weeks.
This paper is milled from a 3rd certified source
Postmaster: Please send address changes to PCI Journal, 200 W. Adams

St., Suite 2100, Chicago, IL 60606. Periodicals postage rates paid at Chicago
and additional mailing offices.
2 W in t e r 2 0 1 2 | PCI Journal
Winter 2012 • Volume 57, Number 1
Development of a Laser-Speckle Imaging Device 135 Departments

to Determine the Transfer Length in Pretensioned
Concrete Members Chairman’s Message 5
Weixin Zhao, Kyle Larson, Robert J. Peterman, B. Terry Beck, and Chih-Hang J. Wu The Power of Collaboration
Note of Appeciation Inside Front Cover President’s Message 6

Something to Think About
Editor’s Message 11
The Cover Story
Letters 13
From PCI Headquarters 14
Our Members 29
Project Spotlight 36
Coming Ahead 149
135 Board of Directors 149

and Technical Activities Council
Index of Advertisers 150

2011 Article Index PCI Regional Offices 150
View the index for the 2011 PCI Journal articles on the PCI PCI Staff 151
website at www.pci.org/publications/journal/index.cfm.
Meet Doug Sutton 152
On the
WINTER 2012
VOLUME 57, NUMBER 1
Cover
Precast/Prestressed Concrete Institute
The precast concrete PricewaterhouseCoopers office build-

ing after the earthquake of February 22, 2011, in Christchurch,
New Zealand. Photo courtesy of Frank Kahn. Cover designed
by Paul Grigonis.
42 97 124
Earthquake Retrofit of Precast Panels
Reconnaissance Grouted Sleeve for Lateral-Force
Team Reports Connections Resistance
Journal Advisory Committee Amir Fam George D. Nasser Stephen Seguirant

Chair Millard J. Barney Peter I. Finsen Donald C. Raths C. Douglas Sutton
Douglas M. Mooradian Carin L. Roberts-Wollmann
PCI Journal | Wi n t e r 2012 3

®
THE C O N S U LT I N G ENGINEERS GROUP INC.®
www.cegengineers.com
making all the
PIECES FIT
Architects . Engineers . Parking Consultants
Main Offices: Other Offices:

Mt. Prospect, IL Burleson, TX Horseshoe Bay, TX Apple Valley, MN
Tel: 800-755-5201 Tel: 817-627-4411 Tel: 830-598-6100 Tel: 952-240-1101
San Antonio, TX Albuquerque, NM Lake Mary, FL Ft. Lauderdale, FL

Tel: 800-827-1906 Tel: 505-796-0972 Tel: 321-275-0580 Tel: 954-523-8347
Chairman’s Message
The power of
collaboration
Over the past 26 years that I’ve been involved with PCI, I’ve been genuinely amazed by the
willingness of so many individuals to give of their time and talents to create, nurture, and
promote a system for building structures. I have been privileged to work alongside people
who are indeed titans of the industry and countless others also deserving recognition who
have demonstrated a passion for this work that continues to inspire.
The strength of this institute is most definitely in the experience, dedication, and perspira-
tion of so many volunteers. Its technical knowledge and focus are what make it so much
more than a mere trade organization. We have a strong institute and are financially sound
thanks to prudent management. We have an extremely strong and capable staff to help drive
the industry forward. We must not forget, however, that the heights we might reach and
things we might achieve will be the culmination of the work of the collective whole.
In his latest book, The Third Alternative, Stephen Covey writes about the power of seeing
beyond our selfish interests to the amplified results from synergistic collaboration. This, to
me, sums up PCI at its best. Our opportunities lie in continuing to make our voices heard,
listening to what others have to offer, and working together to keep precast/prestressed con-
crete in the vanguard of construction materials.
I attended a seminar recently about getting people fired up about formulating goals. Here
are some thoughts along those lines for your consideration as related to PCI involvement:
• Those of us on the downhill side of our careers, think about the importance of the
legacy we can leave behind.
• Those who have the majority of their careers ahead, make an impact to benefit the
global society by being on the cutting edge of technology, by pushing the envelope, and
by coming up with the means and methods to design and construct structures that offer
solutions to tough problems.
• Finally, business leaders, get involved and help shape the future of your industry.
We need to emphasize further establishing PCI as the preeminent authority on matters
related to precast/prestressed concrete, and we need to start thinking with a global perspec-
tive. Leveraging our technical brain trust for this and to affect the codes is paramount. We
must challenge ourselves to provide products and services so indisputably reliable that PCI
plant certification precludes consideration of any other alternatives, and we must further
expand our consideration of quality assurance by extending the effort to the field. We
should be able to ensure that the completed structure meets our design intent and standards
through competent inspections.
It is an honor and a privilege to serve in the position of chairman. I don’t take it lightly, and
I know I have big shoes to fill. I am thankful for the opportunity to try to make a positive
Greg Force | 2012 PCI Chairman difference. With the combined talents and efforts of this membership, I am confident of our
Tindall Corp. | Spartanburg, S.C. ability to do so. J

President’s Message
Something to
think about
Our industry has some important potential strategic advantages in the marketplace. But to
turn them into real advantages, we need to understand them fully and act on that under-
standing. I’d like to discuss two of them: high-performance systems and advanced materials.
Potential advantage 1: High-performance systems

Owners, architects, designers, constructors, and the general public increasingly demand
structures that are:
• Sustainable: designed and built in a way that considers regional, global, and long-term
effects, intelligently using materials and energy to ensure the availability of sufficient
resources to future generations. This means, among other things, minimizing nonre-
newable materials, energy use, and the environmental footprint while maximizing the
Please send
correspondence to journal@
service life.
• Resilient: able to resist permanent damage from natural and accidental events (for
pci.org or Precast/Prestressed
example, severe weather, earthquakes, and physical impact) and still remain serviceable.
Concrete Institute, c/o PCI
This benefit is gaining importance, particularly in the wake of the highly destructive
Journal, 200 West Adams Street,
hurricanes and tornadoes of recent years.
Suite 2100, Chicago, IL 60606.
• Protective: more secure and comfortable for users and occupants because of structural
stability, fire resistance, blast resistance, noise isolation, or resistance to mold or rot.
Safety and security are difficult to measure but are increasingly desired as the world
seems more threatening.
• Constructive: designed and built to have positive functional, economic, and aesthetic
effects. This means, among other things, construction with minimal traffic disruption,
material handling, power and water use, dust, fumes, noise, and overall duration. It also
means a practical, beautiful structure that harmonizes with its environment.
• Beautiful: this is self-explanatory (just ask the beholder).
Precast concrete structures are inherently strong in providing these attributes and PCI’s
marketing approach underscores this in many ways, but there’s much more we can do.
James G. Toscas, PCI President
6 W int e r 2 0 1 2 | PCI Journal

Among these, the hot topic has been sustainability. Now a new hot topic is emerging: func-
tional resilience. We are natural leaders in this area. If we take hold of that lead and main-
tain it, we can avoid having to play catch-up later on.
A term that encompasses both sustainability and functional resilience is high performance.
We need to own it. “Designers Notebook: High Performance Precast Insulated Sandwich
Wall Panels” was published in the Summer 2011 issue of PCI’s Ascent magazine. However,
the broadest definition of high-performance is still vague at this point, and if we intend to
own it we need to define it.
To me, a high-performance system is one that has the versatility to meet a variety of needs.
High performance is the most general way to discuss many of the attributes we have been
emphasizing for years, including design flexibility, as touted by our “Discover the Freedom
of Precast” tagline.
“Freedom of Precast” was targeted specifically to architects, who value design flexibility but
often don’t associate it with precast concrete. Today, we need a core message that appeals to
all decision makers and influencers. High-performance systems could be it. If high-performance
systems can meet a wide variety of needs, then high performance can encompass a wide range of
capabilities. It would provide a common theme that suits a variety of times, regions, markets,
or targets and that logically underlies everything we say about precast concrete.
Why do we need to reach a broader range of groups? Today architects are not the only ones
deciding whether to use precast concrete. As sustainability becomes more important, we see
other stakeholders having increasing influence over materials and systems selection. Because
of resource limitations, we may decide, for now, to limit our marketing to one or two deci-
sion-maker or influencer groups, but the core message of our marketing should make sense
to anyone.
When the target audience changes, the focus may need to be changed—but the core message
should not. For instance, resilient and protective are two sides of the same coin: resilience pri-
marily benefits the owner, while protectiveness primarily benefits the user. Different focus,
but the core message—high performance—is the same.
This approach would fit well with PCI’s three-tiered marketing. The core message, which
would be promulgated nationally and even globally, would be the province of PCI itself.
National programs are best positioned to increase general awareness and spur interest.
The offshoots could be developed, customized to fit regional markets, and executed by our
regional affiliates with resources such as information, templates, and collateral materials
from PCI. PCI Producer Members can then be equipped and positioned to close the deal
based more on value and less on price.
Potential advantage 2: Advanced materials

Because we fabricate precast concrete components under relatively controlled conditions,
we have the obvious advantages of higher quality and consistency as well as leaner opera-
tions. Less obviously, we are well positioned to exploit advances in materials technology. For
example, precast concrete producers in North America started producing with self-consoli-
dating concrete years before it caught on with ready-mixed concrete producers. Why? New
materials can be tricky to deal with and can be handled most effectively in a controlled envi-
ronment. Our industry therefore benefits by staying at the forefront of concrete materials
technology, and we should actively seek out, develop, and introduce new materials wherever
possible.

For example, precast concrete producers have worked for years with supplementary cementi-
tious materials, such as fly ash and silica fume. Recent efforts to reduce the carbon footprint
of concrete structures have renewed interest in these materials, but they pose certain produc-
tion challenges. In all likelihood, some practical research in this area could result in these
materials being more extensively used by precasters.
PCI’s research and development efforts focus on practical results that can be immediately
applied, as they should. Who will conduct the research that has potential long-term payback
and would benefit precast concrete more than cast-in-place concrete?
The next material for us to capitalize on might be ultra-high-performance concrete

(UHPC). With elevated compressive and tensile strengths and possibly greater stiffness,
ductility, or durability, UHPC could significantly expand the range of applications for
which precast concrete is technically and economically viable. There have been a few spec-
tacular examples of this, such as the Shawnessy Light Rail Transit Station in Calgary, AB,
Canada. Again, the benefit to the precast concrete industry is potentially greater than to
the cast-in-place concrete industry. Who will research and explore the possible new applica-
tions?
We might have an answer to these questions right beside us. At the last meeting of the PCI
Foundation Board of Trustees in October 2011, Ken Lambla, dean of the School of Arts
and Architecture at the University of North Carolina at Charlotte, gave an exceptional and
thought-provoking presentation. After hearing status reports on the foundation’s highly suc-
cessful Architectural Design Studios, he tossed aside his prepared remarks and spoke off the
cuff. He outlined the huge potential of the Architectural Design Studio program and intro-
duced the idea of design research. Lambla was talking about the design of structures, but the
same concept could be applied to the design of concrete systems as well.
The purpose of the PCI Foundation is to develop people and technology to improve society
using precast concrete structures. Its key concept is to make precast concrete a better way to
build. Better than what? Of course, we’d like to think it is better than steel or better than
cast-in-place concrete, but I think it’s sufficient to say simply “even better than it is today.”
Design research is what we need to identify and develop new ways to build with precast
concrete, to make it even better than it is today. Whether we want to know how we can use
a new material or how we can create one, the foundation may be the right research vehicle,
but finding new solutions won’t do any good if we’re not prepared to implement them. Our
industry needs to be ready to use new types of materials and build new types of structures.
Are we?
Our opportunity is now

Although we can be proud of our positions in the market for bridges and parking structures,
precast concrete is not the default choice for most structures today. Opinions and prefer-
ences are predisposed to other systems, and we need to fight, project by project, to steer
things our way. We’ve come to accept that precast concrete is not always the first choice, but
why should we? What if precast concrete were the solution of choice and the other guys had
to fight for every project?
Such a future is possible for us, but we need to change the way people think. In the broadest
sense, this is the purpose of marketing.

I don’t need to tell you that the current recession is not typical. Relatively speaking, the
economy is in a state of chaos. Out of this chaos will come new priorities, policies, and pref-
erences. To resolve this economic mess, leaders and professionals will see a need, and feel
justifiably entitled, to break with past practice. The marketplace today is like fresh concrete:
fluid and hard to nail down but formable. In a few years, it will begin to set into a new shape,
recognizable yet different from what it has been. We need to help define that shape.
Now is the time to bend minds—starting with our own. J

Editor’s Message
The
cover story
Starting with this issue, we’re expanding on the theme by means of a cover story. The cover
story will connect the dots: research, design, construction, and performance. In some cases,
the cover story will highlight one of the peer-reviewed papers by showing how research has
been applied in the field. In other cases, the cover story will emphasize the design, construc-
tion, or performance of precast concrete and point back to the developments that helped
make it possible or show what we can learn from practical experience so that we do things
better in the future.
We can’t promise to include all of the stages in the process from beginning to end in every
issue, but we’ll include as many aspects as we can. We hope that the cover story will make
the academic papers more approachable and emphasize the importance of practical applica-
tions. We also hope to encourage—and maybe even inspire—academics to consider the end
use of their research as they plan their experimental work.
Please send
correspondence to journal@
This is our seismic issue, and the contents reflect the international nature of the topic. Our
cover story summarizes the findings of PCI’s New Zealand reconnaissance team on the per-
formance of precast concrete structures in the February 22, 2011, Christchurch earthquake.
pci.org or Precast/Prestressed
We also have a preliminary report on the performance of precast concrete structures in the
Concrete Institute, c/o PCI
March 11, 2011, Tohoku earthquake from PCI’s Japan reconnaissance team and a report on
Journal, 200 West Adams Street,
precast concrete structures in the earthquake that took place on February 27, 2010, off the
Suite 2100, Chicago, IL 60606.
Chilean coast from PCI’s Chile reconnaissance team. All three of these reports pertain to
structures built in accordance with modern building codes that were properly enforced. The
lessons learned by these and other teams of structural engineers will help to make our own
building codes that much better.
Theme-related peer-reviewed papers include a paper from Romania on precast concrete

systems for residential buildings and one from Italy on cyclic behavior of grouted-sleeve
connections between columns and foundations. From closer to home (for North American
readers), we have a paper on the design of self-supporting lateral-frame-resistant enclosures
and another on simulated seismic loading of unbonded post-tensioning strand anchorages.
This issue also contains a note of appreciation to those who provided peer reviews of manu-
scripts being considered for publication in these pages. Whether the reviews resulted in
acceptance, rejection, or major revision of the manuscripts, this publication is much better
because of them. Thank you all!
Rachel J. Detwiler, PhD, P.E.

P R E S T R E S S S U P P LY I N C .
1804 W. Lake Parker Drive • Lakeland, Florida 33805
Phone (863) 683-4492 • Fax (863) 683-2886
Toll-Free 1-800-328-8036
Email: sales@psilakeland.com • Web: www.prestresssupply.com
Letters
Comments? Naaman sends thanks

for Distinguished Educator Award
The editors welcome discussion of PCI
Journal content. Comments must be Since I was told that there will be no time for acknowledgments and
confined to the scope of the article thanks during the PCI awards presentation in Salt Lake City, I would like
under discussion. All discussion of to present my sincere appreciation and thanks to the members of the PCI
articles in this issue must be received by Student Education Committee and the Precast/Prestressed Concrete Institute
April 1, 2012. for selecting me for this award. When I adopted prestressed concrete as a
main topic for my research and teaching more than four decades ago, I did not
Please address reader comments to PCI expect the satisfaction and technical fulfillment it would provide. It has been a
Journal at journal@pci.org. wonderful subject, still humming in my heart and in my brain. I truly consider
the PCI Distinguished Educator Award to be the crowning for my career,
Letters should include the writer’s name, and I hope that my example will entice the younger generation of educators
title, company, city, and email address to keep working in and advancing our knowledge of this wonderful structural
or phone number and may be sent to material that is prestressed concrete.
the respective authors for comment. Thank you.
All letters become the property of PCI
Journal and may be edited for space, Antoine E. Naaman, PhD, FPCI
clarity, and style. Professor Emeritus, Department of Civil and Environmental Engineering,
College of Engineering, University of Michigan, Ann Arbor, Mich. J

From PCI Headquarters
Current and former PCI staff members attend Dan Jenny’s 90th birthday celebration. From left are George Nasser, Brian
Goodmiller, Sidney Freedman, Dan Jenny, Paul Johal, and Jason Krohn. Courtesy of .Dan Jenny.
Former PCI technical director celebrates 90th birthday

To celebrate his 90th birthday, former PCI technical director Dan Jenny hosted a luncheon for about 50
people on August 20, 2011, at the Summer House of the Lutheran Retirement Home in Arlington Heights, Ill.
Jenny’s guests included his family; friends; and PCI technical activities managing director Jason Krohn;
architectural systems, industrial operations, and safety director Sidney Freedman; former research and develop-
ment director Paul Johal; former marketing director Brian Goodmiller; and former editor of the PCI Journal
George Nasser.
During Jenny’s 25-year career at PCI, he was heavily involved in the development of the first and subse-
quent editions of the PCI Design Handbook: Precast and Prestressed Concrete, made recommendations con-
cerning code provisions affecting the precast/prestressed concrete industry, managed the Technical Activities
Committee, and participated in the beginnings of PCI-sponsored research programs. Jenny was also editor of
the PCI Journal during his early years. Later, he became PCI’s first research and development director and was
acting PCI president in 1987 before he retired. In recognition of his many achievements, he has been honored
as a PCI Titan and Fellow. In addition, the Daniel P. Jenny Research Fellowship Award was named in his
honor.
—George Nasser
Distinguished Educator Award call for nominations

>> The PCI Student Education Committee invites nominations from PCI members for PCI’s 2012 Distinguished Educator Award. The
objective is to recognize distinguished educators in the fields of engineering, architecture, and construction technology who have
made significant contributions to the precast/prestressed concrete industry. Nominations must be received at PCI headquarters by
March 1, 2012. For nomination forms and additional information, contact Alex Morales at amorales@pci.org or (312) 360-3219.

New T. Henry Clark Award
to recognize project quality
The T. Henry Clark Quality Award has been approved by the PCI Board of Directors.
The new award is to recognize a precast concrete producer for a specific project that is a suc-
cess in large part because of high quality demonstrated in drafting, engineering, production,
personnel, and field operations. Winners would require quality levels in excess of the mini-
mum requirements of the PCI Plant Certification Program.
T. Henry Clark, a PCI Titan, died in December 2010. He was president of Ross Bryan
Associates Inc. in Nashville, Tenn., where he worked for 43 years. Clark was responsible
for managing Ross Bryan’s auditing of plants participating in the PCI Plant Certification
Program. Ross Bryan has been the sole audit agency for PCI’s Plant Certification Program
since inception. As a member of the Plant Certification Committee, Clark played a major
role in the evolution and development of PCI’s certification programs and their respective
manuals.
Entries for the first annual award program are due March 15, 2012. Eligible produc-
ers must be PCI certified. Entry criteria and forms will be available on the PCI website in
January 2012. For more information, contact Dean Frank at dfrank@pci.org.
Young Educator Award call for nominations

>> The PCI Student Education Committee invites nominations from PCI members for PCI’s 2012 Young
Educator Award. The objective is to recognize young educators in the fields of engineering, architecture,
and construction technology who have made significant contributions in their early careers to the precast/
prestressed concrete industry. Nominations must be received at PCI headquarters by March 1, 2012. For
nomination forms and additional information, contact Alex Morales at amorales@pci.org or (312) 360-3219.
Engineering Student Design Competition:

Big Beam Contest call for entries
>> The PCI Student Education Committee invites entries from engineering students for PCI’s 2011–2012
Engineering Design Competition. With the help of local PCI Producer Members, students will construct and
test precast, prestressed concrete beams. The awards program, sponsored by Sika Corp., will include cash
prizes for the most efficient design, highest load capacity, best report, and other categories. Applications are
due at PCI headquarters by March 15, 2012, and reports by June 15, 2012. PCI Producer Members are urged
to encourage their local engineering schools to participate in this program. For information on official rules
and application forms, go to www.pci.org/bigbeam or contact Alex Morales at amorales@pci.org or
(312) 360-3219.
PCI Journal | Wi n t e r l 2012 15

The PCI Foundation Legacy Society
>>
Having recently returned from the PCI Convention in Salt Lake City, Utah, I am pleased to
report that the PCI Foundation held the most lively and downright soul-satisfying meeting in
our history. Representatives from all six of our foundation-sponsored universities attended a din-
ner meeting and presented detailed reports of their studio programs. They were complimentary
of foundation support and enthusiastic in their praise of the hands-on involvement of industry
professionals. Graduates of foundation-sponsored programs at several participating schools have
found jobs within the precast/prestressed concrete industry. This is what we have worked for,
Jim Voss
and it is finally happening.
With this good news as a springboard, I would like to introduce you to the PCI Foundation
Legacy Society. The whole idea of starting the PCI Foundation was to thank the industry that has provided my liveli-
hood for the past 30 years. Now you can, too.
Beginning in late 2011, the PCI Foundation Legacy Society will be for those PCI members who wish to give back
to their industry with an estate bequest or other form of charitable donation. Members will be recognized at all PCI
events throughout the year as individuals who actively support a stronger precast/prestressed concrete industry.
Following are some of the ways you can become a society member:
• A multiyear pledge. This is the quickest, easiest way to show your support for the work of the PCI
Foundation.
• Gifts by will or living trust. A bequest in your will or trust is completely free of estate tax and can be made
simply by including the following words in your will: “I give, devise, and bequeath to the PCI Foundation
[insert here a dollar amount, percentage of estate, or describe property, such as stocks or real estate] to be used
for its educational purposes.”
• A gift to reduce your tax burden. You or your financial planner can determine the appropriate gift to maxi-
mize your tax savings at the end of the year.
• Gifts of stock. An account has been established at David A. Noyes Co. to receive gifts of stock.
• Retirement accounts or insurance policies. You can designate a specific dollar amount or percentage and
effect an estate tax savings.
• Gifts of real estate. You can donate a house, vacation home, farm, undeveloped land, or other types of real
estate. This will entitle you to an income tax deduction and potentially reduce your taxable estate.
Regardless of what you contribute, you can be assured that your gift will be used to sustain activities that ensure our
industry will thrive in the future. For more information go to www.pci.org/foundation.
New sandwich wall panel Designer’s Notebook released

The High Performance Precast Insulated Sandwich Wall Panels Designer’s Notebook, DN-23, is now available in print. The
printed copy cost is $3.50 per notebook for fewer than 100 and $3 per notebook for more than 100. This Designer’s Notebook dis-
cusses the ways that high-performance buildings satisfy functional, human comfort, environmental, and economic considerations.
It covers energy efficiency, solar reflectance, commissioning, reduced life-cycle costs, design flexibility, life safety/environmental
conditions, indoor air quality, durability, and aesthetics. Visit the PCI Bookstore to purchase this Designer’s Notebook, or view it
online at www.pci.org/view_file.cfm?file=AS-11SU-10.pdf.

PCI-IW continues successful
Technology Tuesdays program in 2012
The Precast/Prestressed Concrete Institute of Illinois and Wisconsin
will continue its popular series of lunchtime events in 2012. The semi-
nars, called Technology Tuesdays, take place the first or second Tuesday
of each month. Each program is taught by an expert in that area of the
precast concrete field.
2012 topics will include:
• Using Emulative Design to Improve Concrete Projects
• Understanding PCI’s In Plant Quality Control Process for
Roger Becker, PCI’s research and development direc- Precast Specifications (February 7)
tor, gives a presentation on connections at Technology • Sustainability Through Durability, Adaptability and
Tuesday, sponsored by the Precast/Prestressed Concrete Deconstructibility
Institute of Illinois and Wisconsin. The connections ses-
• Functional Resilience
sion was the most popular of 2011. Courtesy of Marty
McIntyre.
• Concrete Materials—Improvements and Available Technology
in Concrete Design
• Integrated Design Using Precast
Attendees will earn 1.5 or 2.0 professional development hours. A cer-
tificate will be provided at the conclusion of each seminar.
Technology Tuesday programs are free but require advance registra-
tion. To be placed on the email list for the monthly programs, visit the
sign-in portion of www.pci-iw.org.
Second-edition Level III personnel

certification manual due out in 2012
The second edition of the PCI Quality Control Personnel Level III
Training Manual, TM-103, will be released in 2012. New exams and
electronic grading for Level III personnel certification will launch with
the release of the second edition. For more information, contact Alex
Morales at (312) 360-3219 or amorales@pci.org.
PCI Sustainability Manager

Opening 2011 convention proceedings for sale
>> PCI has an opening for manager, sustainability Peer-reviewed technical papers from the 2011 PCI Convention and
programs, to report to the director of quality National Bridge Conference are now available. To purchase the proceed-
and sustainability programs at PCI headquarters ings, go to the PCI Bookstore at www.pci.org.
in Chicago, Ill. This position will be generally
responsible for strategic and tactical manage-
ment of PCI activities relating to sustainable
development for the precast concrete structures
industry. If you are interested in this position,
please forward a letter of interest and resume to
Christine Bulvan at cbulvan@pci.org.

2011 PCI safety award winners announced
Every year, awards are given to those companies whose plants had “0” ratings for days away and
restricted time, total case incidence, and lost workdays incidence rates during the previous year,
in this case 2010. The awards, divided into categories based on the number of employees in each
plant, are determined by PCI’s Plant Safety and Environmental Committee. This committee col-
lects and analyzes Producer Members’ accident data from the previous year.
Winners in the 1- to 49-employee category are Conewago Enterprises Inc. of Hanover, Pa.;
Coreslab Structures (CONN) Inc. of Thomaston, Conn.; Coreslab Structures (ORLANDO) Inc.
of Orlando, Fla.; David Kucera Inc. of Gardiner, N.Y.; EnCon Northwest LLC of Camas, Wash.;
EnCon Utah LLC of Tooele, Utah; Florida Precast Industries Inc. of Sebring, Fla.; Gate Precast
Co. of Pearland, Tex.; Lombard Architectural Precast Products Co. of Alsip, Ill.; Metromont
Corp. of Bartow, Fla.; Oldcastle Precast of Avon, Conn.; Prestress of the Carolinas, a division
of the Lane Construction Corp., of Charlotte, N.C.; R. B. Johnson Co. of McMinnville, Ore.;
Stonco Inc. of Omaha, Neb.; Stresscon Denver Metro division, of Dacono, Colo.; and Waffle
Crete International Inc. of Hays, Kans.
Winners in the 50- to 99-employee category are Structural Prestressed Industries Inc. of Miami,
Fla., and United Concrete Products Inc. of Yalesville, Conn.
PCI-certified plants reach significant milestones

A number of plants have been recognized for their longstanding participation in the PCI Plant
Certification Program.
Plants that have participated in the PCI Plant Certification Program for 40 years are Nucon
Schokbeton/Stress-Con Industries Inc. of Kalamazoo, Mich., and Tindall Corp. of Fairforest, S.C.
Plants that have participated in the PCI Plant Certification Program for 45 years are Blakeslee
Prestress Inc. of Branford, Conn.; IPC Inc. of Iowa Falls, Iowa; Knife River Corp. of Harrisburg,
Ore.; Metromont Corp. of Charlotte, N.C.; The Shockey Precast Group of Winchester, Va.;
Tindall Corp. Corrections division of Conley, Ga.; Tindall Corp. Prestress division of Conley,
Ga.; and Wells Concrete Products Co. of Wells, Minn.
PCI Convention and National Bridge Conference

call for papers and student posters
>> PCI invites submissions for the peer-reviewed papers sessions and for the Student Research Poster Session at the
2012 PCI Convention and National Bridge Conference September 29 to October 3 at the Gaylord Opryland Resort
and Convention Center in Nashville, Tenn. Abstracts may be submitted for the peer-reviewed paper sessions or for
the Student Research Poster Session. All abstracts, papers, and student posters must relate to the design, analysis,
materials, production, erection, or maintenance and repair of precast concrete. Case studies, research, and project
profiles are welcome. Abstracts should be no more than 200 words for peer-reviewed papers or 500 words for the
Student Research Poster Session and should adequately describe the topic. All abstracts must be submitted elec-
tronically at www.softconf.com/c/PCI2012 by February 6, 2012. For more information, visit www.pci.org
/convention/callforpapers or contact Jennifer Peters at jpeters@pci.org or Brian Miller at bmiller@pci.org.

PCI, PCA meet with New York City officials
regarding proposed cement limits
On October 31, 2011, representatives from PCI and the Portland Cement Association
(PCA) met with legislative staffers and two members of the New York City Council to discuss
two proposed laws intended to make New York a more sustainable city. Intro. 577 would limit
the portland cement content of concrete to a maximum of 400 lb/yd3 (240 kg/m3) as a means
of reducing the carbon footprint of concrete. Intro. 603 would require the use of recycled
materials, such as crushed concrete or glass cullet as aggregate in nonstructural concrete.
PCI and PCA explained the desirability of performance specifications over prescriptive
specifications and discussed potential effects of the proposed laws on the quality and safety
of concrete construction.
At press time no decisions had been made regarding either of the proposed laws.
—Rachel J. Detwiler
New Zealand Engineering Society touts

PRESSS System for all new construction
Support for PCI-sponsored PRESSS (Precast Seismic Structural Systems) earthquake-
resistant technology has been advanced by the New Zealand Society for Earthquake
Engineering (NZSEE).
NZSEE issued a press release suggesting that all new construction in New Zealand adhere
to seismic technologies such as the PRESSS standards. Proposed as a solution for rebuilding
Christchurch, New Zealand, following the 2011 earthquake, NZSEE executive officer Win
Clark says that the PRESSS system should be considered for all new buildings up to nine
stories.
The release noted that the Southern Cross Endoscopy building, which was built using
PRESSS technology, performed well in the recent earthquakes. It also cited several other
New Zealand structures built with PRESSS technology.
The PRESSS system uses unbonded post-tensioning to control rocking of a structure‘s
joints during a seismic event and allows the structure to spring back to upright.
Solicitation for Jenny Research Fellowship proposals

>> PCI’s Research and Development Council is soliciting proposals for the 2012–2013 academic year Daniel P.
Jenny Fellowships. Several $20,000 awards will be offered and are intended primarily for support of candi-
dates for master’s degree–level research related to precast/prestressed concrete. PhD candidates will also
be considered. Multiple proposals are welcome. Proposals must be submitted to Roger Becker at rbecker@
pci.org by February 20, 2012. The PCI Research and Development Council will meet in late March 2012 to
decide on the awards. Results will be available by April 30, 2012. For more information, visit pci.org/cms
/index.cfm/tech/research.

The PCI Foundation raised almost $25,000 during its auction at the 2011 PCI Convention in Salt Lake
City, Utah. Courtesy of Paul Grigonis.
PCI Foundation auction raises nearly $25,000

The PCI Foundation raised almost $25,000 during its auction at the 2011 PCI
Convention. Auction chair Marianne Methven of Hamilton Form and the auction team
gathered almost 70 auction lots from as many donors. Items ranged from Zen Parenting to
a train trip across the Canadian Rockies. Groups of PCI members, including the Titans and
Fellows, PCI Board of Directors, Women in PCI, and alumni of Leadership PCI, put their
funds together to donate special packages for the auction. Producer members, associates, and
individuals also took part in donations. More photos of the auction and information on PCI
Foundation programs are available at the Foundation Auction Blog at pcifoundation
.blogspot.com.
Becker nominated for ACI Board of Direction

The Committee on Nominations for the American Concrete Institute
(ACI) has recommended Roger Becker, PCI’s managing director of
research and development, for the ACI Board of Direction.
Becker is a member of ACI committees 301 Specifications for
Concrete, 318 Structural Concrete Building Code; ACI subcommit-
tees 318-C Safety, Serviceability, and Analysis, 318-G Precast and
Prestressed Concrete; joint ACI-ASCE Committee 423 Prestressed
Roger Becker
Concrete; the Concrete Research Council; and the Technology
Management Committee of the ACI Strategic Development Council.
He is also a member of the Structural Design for Fire Conditions Standards Committee at
ASCE.

PCI Foundation holds dinner meeting
The PCI Foundation held a dinner meeting during the 2011 PCI Convention in Salt
Lake City, Utah, in October. The meeting provided an opportunity for PCI Foundation
trustees and academics representing the PCI Foundation–sponsored studio programs to
share information about their programs.
The PCI Foundation studio programs, though diverse, are based on a common concept:
providing students with practical experience in designing precast concrete structures. Each
school reported on its approach and the success of its studio.
In some cases, the high level of energy and enthusiasm within these programs spread to
other departments within the university. One result of the PCI Foundation studio program
has been cooperation between architectural and engineering departments and students.
The gathering itself fostered personal interactions among the participants based on a
common interest in educating and developing the next generation of design professionals.
The guest speaker, Ken Lambla, dean of the University of North Carolina at Charlotte,
complimented the high quality of the studio programs and the camaraderie of the evening
and urged the PCI Foundation to direct a share of its resources toward design research.
—Jim Voss
PCI manuals listed as referenced standards

within 2012 International Building Code
PCI has two manuals that were referenced within the International Building Code (IBC)
since the 2003 editions. For the 2012 IBC, the PCI Recommended Practice for Glass Fiber
Reinforced Concrete Panels, MNL 128-01, is referenced in its entirety within code section
1903.2.
Also for the 2012 IBC, the PCI Design for Fire Resistance of Precast Prestressed Concrete,
MNL 124-89, is referenced within code section 722.2.3.1. The procedures for slab cover of
precast/prestressed concrete within MNL 124-89 are the relevant reference within the IBC.
This PCI manual has ICC Legacy Evaluation Report ER-3264. With this ICC Evaluation
Report, the PCI manual is recognized as complying with the code provisions of the 1997
Uniform Building Code.
Leadership PCI call for nominations

>>
The Leadership PCI (LPCI) Selection Committee is seeking nominations for the 2012/2013 class. LPCI is an
industry leadership development program created to cultivate strong leaders in the precast/prestressed
concrete industry. Any PCI member organization can submit a candidate for LPCI. Nominations must be
received at PCI headquarters by April 15, 2012. Visit pci.org/cms/content/members/leadership/index.cfm
and log in for the LPCI brochure and application. For more information, contact Alex Morales at amorales@
pci.org or (312) 360-3219.

Architecture, engineering students participate in first poster session
During the 2011 PCI Convention in Salt Lake City, Utah, a new program was launched on the show floor. Engineering stu-
dents from eleven universities and architecture students from five Precast Studios delivered poster sessions on their precast/pre-
stressed concrete–related work. A poster session is the presentation of research information by an individual or representatives of
research teams with an academic or professional focus.
The program, sponsored by the PCI Student Education Committee (chaired by Richard Miller) and by the PCI Foundation,
allowed students to share with the precast concrete industry some of the innovative and important work being done around the
United States.
Poster sessions can be seen at leisure by a number of attendees over the course of several days. “The poster session accomplished
everything we hoped it would, and more,” Miller says. “The work by the students was outstanding. I thought the students did an
excellent job of presenting their work, and those attending the convention were very interested. There was a steady stream of visi-
tors interacting with the students. This is the advantage to the poster session.”
During PCI Convention exhibit hours, students were able to stand with posters that represented their research or work in stu-
dios to answer questions. Many of the students’ work reflected the work that is being funded by PCI Research and Development
dollars. “The poster session gave PCI members a chance to see how the money PCI spends on research projects is being put to good
use,” Miller says. “It also showed donors to the PCI Foundation the excellent work of foundation-supported studios.”
In addition to displaying their posters and answering questions, the students were able to attend PCI programs. “When I was
not presenting my poster, I attended great educational sessions. I came back with a wider vision on my project and on the advances
in the precast concrete industry,” says Pinar Okumus, a doctoral candidate at the University of Wisconsin–Madison.
For the studio students involved with the poster session, it represented a continued involvement in an industry where, at least
academically, there used to be very little attention.
“The continued expansion and diversification of the Foundation’s educational initiatives are a benefit to all PCI members and
participating institutions,” says Tom Brock, director of the Precast Studio at the Illinois Institute of Technology. “The payback
is, of course, not immediate. But at IIT, for instance, we can now say that there are 65 graduates, soon to be or already practicing
architects, who have been immersed in the design and technology of precast concrete buildings for an entire semester, a topic that
would have otherwise taken up about 50 minutes of discussion in a technology survey course.”
For more information on the PCI Foundation, visit the website at www.pci.org/foundation.
—Marty McIntyre
The Student Research Poster Session

allowed students to present their
research in an informal setting and
PCI Convention attendees to learn
about current research. Courtesy of
Paul Grigonis.

Kassian Dyck winner of PCI video contest
Kassian Dyck Associates and Kassian Dyck International took first place this year’s How
Do You Shape Your Future? video contest. BASF took second place in the contest, and
Econstruct received an honorable mention. To view their videos, visit www
.youtube.com/2011PCIConvention and view the How Do You Shape Your Future video
comments.
Brown added to PCI’s

shipping, receiving staff
Kim Brown was hired in October 2011 as PCI’s shipping and receiv-
ing associate. In this position, she is responsible for PCI’s shipping,
Kim Brown
receiving, physical order fulfillment, and general storeroom operations
and also supports the Quality Control department. She is available
to take book orders and answer questions concerning publications and QC certification at
kbrown@pci.org and (312) 428-4946.
Brown previously worked at PCI from 2005 to 2008. She has an associate’s degree in
applied science from Robert Morris University in Chicago, Ill.
PCI seeks interested parties for

Precast Erectors Workshop on Lean Principles
PCI has proposed a Precast Erectors Workshop on Lean Principles that would take
place on Thursday, March 29, in conjunction with the 2012 PCI Committee Days and
Membership Conference.
The workshop would describe the use of lean manufacturing methods and best practices
to achieve improvements in productivity, quality, and turnaround times. Participants would
learn how to apply leading-edge productivity best practices to erection processes. The course
would include a presentation on lean and productivity best practices for erection processes,
case studies, hands-on exercises in using lean methods, and open-forum discussions. The
material from this course would provide erectors the knowledge to use modern lean prin-
ciples and to accelerate their lean and continuous improvement efforts.
If you would like to send employees to the workshop, contact Sid Freedman at
sfreedman@pci.org. J
PCI Codes and Standards Manager Opening

>> PCI has an opening for manager, codes and standards, to report to the managing director of technical activi-
ties at PCI headquarters in Chicago, Ill. This position will be generally responsible for strategic and tactical
management of PCI activities relating to the development and improvement of codes and standards that
affect the precast concrete structures industry. If you are interested in this position, please forward a letter
of interest and resume to Christine Bulvan at cbulvan@pci.org.

PCI Calendar
>>
Events
For the most current information on PCI events, visit www.pci.org/events. For industry
events, visit www.pci.org/news/events.
2012 PCI Winter Conference January 19–22, 2012
San Antonio Marriott Rivercenter, San Antonio, Tex.
Zones 1 and 2 Meeting January 24–25, 2012
Hilton Las Vegas Hotel, Las Vegas, Nev.
Committee Days and Membership Conference March 29–April 1, 2012
Wyndham Hotel, Chicago Ill.
Summer Conference June 28–July 1, 2012
Location TBD
PCI Convention and National Bridge Conference September 29–
October 3, 2012
Gaylord Opryland Resort and Convention Center, Nashville, Tenn.
PCI Convention and National Bridge Conference September 21–25,
2013
Gaylord Texan Resort and Convention Center, Grapevine, Tex.
2014
Gaylord National Resort and Convention Center, National Harbor, Md.
2015
Gaylord Opryland Resort and Convention Center, Nashville, Tenn.

>>
PCI Personnel Training and Certification Schools
If you have any questions about the Quality Control School schedule or need help completing
a registration form, please contact PCI’s educational activities director, Alex Morales, at
amorales@pci.org or (312) 360-3219. Registration forms are available at www.pci.org/schools.
Level I/II
January 23–25, 2012 May 2012
World of Concrete, Las Vegas, Nev. Nashville, Tenn.
Level III
May 2012
Nashville, Tenn.
CFA/IES
January 23–25, 2012 May 2012
World of Concrete, Las Vegas, Nev. Nashville, Tenn.
Compiled by K. Michelle Burgess (mburgess@pci.org)

Survey on high- Research & Development Survey
Please list below the four most pressing technical problems that, in your
priority topics for opinion, confront the precast/prestressed concrete industry today. Provide a
research at PCI brief statement of scope and objective(s) of research that would address each
proposed topic. Attach additional sheets as required.
To: PCI Producer, Associate, and 1.

Professional Members
Do you know that each year PCI
allocates 7% of Producer Member
and Associate Member dues for proj- 2.
ects conducted under the auspices of
the PCI Research and Development
Committee? Despite this commitment
by PCI, total industry expenditures 3.
for research conducted on topics of
specific interest to the precast con-
crete industry are still relatively low
compared with those for competitor 4.
industries. It is especially important that
we spend PCI’s research dollars wisely.
Accordingly, we are asking for your help.
Periodically, we solicit input from Please describe your company’s current and recent (past 5 years) nonpropri-
the entire PCI community regarding etary research and/or testing. Attach additional sheets as required.
your opinions on technical research
activities that could produce knowledge
that would benefit your organization
and the industry’s market share. We
then summarize these data in the form
of a prioritized listing titled “Top
10 PCI Research Topics.” This list is
shared with prospective researchers to Additional comments. Attach additional sheets as required.
assist them as they develop research
proposals to be funded as regular PCI
(major) projects and/or PCI (Daniel
P. Jenny) Research Fellowships, as well
as those to be funded entirely or in
part by other research agencies, such Please mail/email/fax this sheet to
as the National Science Foundation, Roger Becker
the National Institute of Standards & Research & Development Director
Technology, and the Federal Highway Precast/Prestressed Concrete Institute
Administration. 200 W. Adams St., Suite 2100 • Chicago, IL 60606
We have designed a survey form to Telephone: (312) 786-0300 • Fax: (312) 621-1114
facilitate your providing us with your Email: rbecker@pci.org
opinions regarding current techncal
research needs. In addition, we are
requesting that you tell us about your Name (please print)
company’s current and recent (nonpropri-
etary) research and development activities. Position
Please send your response to Roger
Becker at PCI as indicated on the survey Company
form.
Harry Gleich Telephone Fax
Chair, PCI Research &
Development Committee Email

Our Members
Thermomass gets four-hour ASTM E119 rating

Thermomass recently successfully completed the ASTM E119 fire test.
The fire test was conducted on a sandwich wall panel consisting of a 2 in. (50 mm) exte-
rior wythe of concrete, a 2 in. insulation core, and a 5 in. (130 mm) interior wythe of con-
crete. The interior wythe was loaded to 4000 lb/ft (6000 kg/m) gravity load and exposed to
fire for four hours, with temperatures reaching 2000 °F (1100 °C).
Darryl Dixon, director of Technical Services at Thermomass, says, “During the four-hour
test, the temperature on the exterior wythe of concrete increased by 130 °F [70 °C], which is
only 50% of the allowed temperature increase per ASTM E119.”
The successful four-hour mark is the first and only ASTM E119 test on an integrally
insulated sandwich wall using an energy-efficient fiber-reinforced polymer wythe connector.
Venkatesh Seshappa, director of Engineering at Thermomass, says, “The ASTM E119 test
expands on the ICC-ESR on our MC/MS Series connectors, and provides an unprecedented
level of testing achievements for the concrete sandwich wall panel industry.”
—Source: Thermomass
Metromont awarded Ga. high school contract

Metromont Corp. was recently awarded a contract for a 388,000 ft2 (36,000 m2) high
school for Newton County in Georgia. The new high school will use a total–precast concrete
building system and will be the largest school in the area.
This will be a state-of-the-art school for the Newton County Board of Education. The
precast concrete walls, floors, and roof system will be produced at Metromont’s Atlanta, Ga.,
manufacturing facility.
The Newton County School Board elected to use the total–precast concrete school
design for several reasons. First, it has proved to be cost competitive with other delivery
systems in other schools built in the Atlanta region. It was determined that the building
shell could be closed in more quickly using precast concrete, which would facilitate the tight
construction schedule. Studies also showed that a total–precast concrete school would be
less expensive to heat and cool and less expensive to maintain than traditional construction
systems.
Cunningham, Forehand, Matthews and Moore Architects is heading up the design team
along with the structural engineer of record Pruitt, Eberly and Stone Structural Engineers,
both of Atlanta. McKnight Construction Co. Inc. of Augusta, Ga., is the general contractor.
The grade work had been completed at the time of award of the contract, and the school
is scheduled to be open for students in August 2013. The new high school is designed for
1800 students.
—Source: Metromont Corp.

Donald Pfeifer
Donald Pfeifer, PCI Fellow, died December 4, 2011. He was 75.

As a nationally recognized expert in concrete structures, concrete materials, and
the investigation and remediation of concrete-related problems, Pfeifer was a pioneer
in the development of solutions for corrosion-related distress in reinforced concrete
structures.
Pfeifer’s interest in engineering brought him to the University of Illinois at Urbana-
Champaign, where he earned a bachelor’s degree in civil engineering in 1959 and a
master’s degree in theoretical and applied mechanics in 1960.
Pfeifer’s professional career began at the Portland Cement Association (PCA) in
Skokie, Ill. From 1960 to 1970, he carried out a wide range of research investigations
there. His major work involved structural lightweight aggregate concretes, precast concrete panels, and architectural
concrete, as well as laboratory and field instrumentation studies of state-of-the-art tall reinforced concrete buildings.
After leaving PCA in 1970, Pfeifer served as the director of the Structural Precast division of PCI from 1970 to
1972, with responsibility for technical, marketing, and educational efforts. During this time, he visited more than
80 prestressed concrete plants to discuss production and engineering problems.
In 1972, Pfeifer joined the Westinghouse Prestressed Concrete division. As the manager of strategic engineering
for Westinghouse from 1972 to 1976, Pfeifer conducted in-plant studies of production problems associated with
precast, prestressed concrete. He performed structural studies that involved fire tests, lateral-load-distribution tests,
and prestress losses on American Association of State Highway and Transportation Officials bridge beams. He also
developed a pretensioned concrete crosstie that was tested by the Federal Railroad Administration.
Pfeifer joined Wiss, Janney, Elstner Associates Inc. (WJE) in 1976, where he founded and managed a materi-
als group. His consulting activities at WJE involved the investigation and repair of corrosion of reinforcement in
concrete structures, concrete mixture designs for nuclear construction and large bridge projects, underwater tremie
concreting, architectural concrete, long-term creep and shrinkage studies on concrete, corrosion studies of alterna-
tive deicing products for structural steel and reinforced concrete bridge members, facade and laboratory investiga-
tions involving glass-fiber-reinforced concrete panels, elastomeric bearing pads, and numerous projects dealing with
concrete materials.
Pfeifer served as principal investigator on four major corrosion-related studies sponsored by the Federal Highway
Administration (FHWA) or the National Cooperative Highway Research Program (NCHRP). These studies
resulted in the 1981 NCHRP 244 report, Concrete Sealers for Protection of Bridge Structures; the 1987 FHWA
report RD-86/193, Protective Systems for New Prestressed and Substructure Concrete; the 1989 NCHRP 313
report, Corrosion Protection of Prestressing Systems in Concrete Bridges; and the 1998 five-year FHWA study,
Corrosion Resistant Reinforcement for Concrete Components.
Pfeifer retired from WJE in 1998.
Pfeifer was a Fellow of the American Concrete Institute (ACI) and a member of the honorary fraternities Chi
Epsilon and Sigma Tau. He served on technical committees for PCI and ACI, and he published more than 50 tech-
nical papers.
Holcim plant becomes EPA Green Power Partner

The Midlothian, Tex., plant of Holcim (US) Inc. has qualified for a U.S. Environmental Protection Agency
(EPA) Green Power Partnership and is now an EPA Green Power Partner.
The EPA Green Power Partnership is available to renewable energy leaders whose green power use meets or
exceeds the EPA’s green power use requirements. Partner organizations must buy green power in amounts pro-
portional to their annual electricity use.
—Source: Holcim (US) Inc.
James H. Gilbert
James H. Gilbert, PCI Fellow and president from 1965 to 1966, died September 4,
2011. He was 91.
After obtaining a BA from Dartmouth College, Gilbert joined C. W. Blakeslee
and Sons Inc., a subsidiary of the Westinghouse heavy construction division, in 1946.
Subsequently, he rose to become vice president, president, and chief operating officer
of the firm.
In 1958, Gilbert started the Prestressed division at Blakeslee. That same year, he
became a member of PCI. Gilbert was named chairman of the Blakeslee board in 1974.
He retired from Blakeslee in 1976 and then served as business advisor and special con-
sultant to the Heavy Construction division of Westinghouse Electric Corp. Gilbert
also played an important role in developing Precast Systems Inc. (PSI) and served as the first PSI president.
In addition to serving on the PCI Board of Directors, he was president of the Connecticut Society of Civil
Engineers. He was active in developing the PCI Plant Certification Program, and his firm served as a charter
certified plant. Gilbert was also chair of the 1974 FIP/PCI Congress Organizing Committee charged with over-
all administration of the New York, N.Y., congress under the direction of the PCI Board of Directors. In 1994,
Gilbert was named one of PCI’s first 25 fellows in recognition of his contributions to the precast/prestressed con-
crete industry and PCI.
Fabcon celebrates 40th anniversary

Fabcon celebrated 40 years in the precast concrete industry in 2011. Fabcon’s first plant began operations
in 1971 as a plank manufacturer. Later, the majority of Fabcon’s business was multistory systems buildings for
apartments and condominiums. Today, Fabcon has a national accounts program focused on big-box retailers,
including Home Depot, Lowe’s, Menards, and Target.
The patented manufacturing process used in Fabcon’s first plant helped lay the foundation for the com-
pany’s current success. Leveraging rolling-bed technology, the forms move to where the concrete is mixed.
This allows Fabcon to house all of its equipment in a small, quality-controlled area, as opposed to a large space
required for fixed-bed production. Fabcon still uses this manufacturing process at its facilities in Minnesota,
Indiana, Pennsylvania, and Ohio.
With headquarters in Savage, Minn., Fabcon gained two plants—one in Indianapolis, Ind., and the other in
Columbus, Ohio—after acquiring the American Precast Co. in 1995. Fabcon expanded to Allentown, Pa., in
2000.
Each day, Fabcon’s plants can produce 1.5 mi (2.4 km) of precast concrete panels, or enough to cover one
city block. Almost 10% of its employees in Minnesota have worked for Fabcon for more than 20 years. A few of
the innovations that Fabcon is known for include its rolling-bed technology, three-dimensional design software
that interfaces with a computer-aided manufacturing system to project a laser image of designs directly onto the
casting beds, and a segmented screed that enables automated equipment to bypass obstructions in the concrete,
such as structural connections or blockouts, that would otherwise need to be hand finished.
—Source: Fabcon

This rendering shows the new Cruise Terminal 6 808-stall parking structure for the Canaveral
Port Authority in Florida. The parking structure is one of five new Finfrock design-build projects.
Courtesy of Finfrock.
Finfrock awarded five parking structure projects

Finfrock Construction Inc. was recently awarded five new parking structure projects.
Design-build services will be provided for three new projects in central Florida: a 617-stall
parking structure for Adventist Health System/Sunbelt Healthcare Corp. in Altamonte
Springs; a 330-stall parking structure for Osceola Regional Medical Center in Kissimmee;
and an 808-stall parking structure serving the new Cruise Terminal 6 for the Canaveral Port
Authority.
In addition, complete design-build services will be provided to Alta Congress LLC
for a 587-stall parking structure project in Del Ray Beach, Fla., and to Metropolitan Life
Insurance Co. for a 1452-stall expansion to an existing Finfrock parking structure in Tampa,
Fla. Construction costs for the five parking structures are expected to total nearly $30 mil-
lion. Design work is under way on each project, with all projects scheduled to begin con-
struction in 2012. Finfrock has begun hiring to fill positions that will be created by these five
new projects.
—Source: Finfrock
Nitterhouse adds Paczewski to sales team

Nitterhouse Concrete Products Inc. recently added Allan P. Paczewski as regional sales
representative in Virginia and Washington, D.C.
In his new position, Paczewski is responsible for providing services to new and existing
clients within Virginia and Washington. His services include personal consultation with
developers, designers, and other construction professionals in the design of precast/pre-
stressed concrete structures, as well as direction and recommendations in the application of
precast/prestressed concrete products.
—Source: Nitterhouse Concrete Products Inc.
Allan Paczewski
First CarbonCast application used in Canada
International Precast Solutions LLC of River Rouge, Mich., a member of The Prestressed
Group and AltusGroup, produced and erected the first CarbonCast Enclosure System in
Canada at the new Cineplex Galaxy. CarbonCast High Performance Insulated Wall Panels
were used in the project.
International Precast Solutions is licensed to produce CarbonCast High Performance
Insulated Wall Panels, CarbonCast Insulated Architectural Cladding, and CarbonCast
Architectural Cladding after becoming the 14th producing member of AltusGroup in
August 2011. It is one of 30 manufacturing facilities nationwide that share more than
400 specification-oriented sales, marketing, and engineering professionals operating PCI-
certified plants.
—Sources: AltusGroup and International Precast Solutions
Stresscon begins precast concrete installation

at National Renewable Energy Laboratory
Stresscon recently began installation of precast concrete on the Energy Systems
Integration Facility (ESIF) at the National Renewable Energy Laboratory (NREL) in
Golden, Colo., with JE Dunn Construction. The ESIF is a 182,500 ft2 (16,950 m2) facil-
ity that includes a 55,055 ft2 (5100 m2) footprint high‐bay structure located on the U.S.
Department of Energy’s NREL campus.
The building is situated on the hillside of a topographically challenging site. This project
features 901 precast/prestressed concrete components, including 290 double-tee floor and
roof elements and 204 insulated wall panels. The insulated wall panels are a structural gray
or acid‐etched finish and are 9 in., 14 in., or 16 in. (230 mm, 360 mm, or 400 mm) thick.
The wall panels are a Thermomass system using polyisocyanurate as the primary insulation
material. The exterior of the structure features a complex system of horizontal reveals in the
precast concrete walls.
The projected erection duration is about 52 days with two (and sometimes three) cranes
working inside and around the perimeter of the structure with about 600 loads delivering
the precast/prestressed concrete components to the jobsite.
—Source: EnCon Design LLC

Gary Gunter Blaine Withers
Heldenfels names new division VPs

Heldenfels Enterprises Inc. has named Gary Gunter vice president and general manager of its Highway and
Bridge division. Gunter joined Heldenfels in September 2010.
Blaine Withers, formerly vice president of Operations has been named vice president and general manager of
the Marine and Industrial division overseeing the Heldenfels plant in Corpus Christi, Tex.
Gil Heldenfels continues serving as vice president and general manager of the Building Systems division,
which designs, manufactures, and erects commercial structures and sports facilities from both plants.
—Source: Heldenfels Enterprises Inc.
Cretex’s Haeder receives ASTM committee award

Corey L. Haeder, chief engineer at Cretex Concrete Products West Inc. in Rapid City, S.D., has received the
Merlin G. “IB” Spangler Award from ASTM International Committee C13 on Concrete Pipe for his outstand-
ing technical contributions to the committee.
A member of ASTM International since 2003, Haeder also serves on Committees C09 on Concrete and
Concrete Aggregates and C27 on Precast Concrete Products.
Haeder joined Cretex Concrete Products West Inc. in 1997 as an office engineer and assumed his current
role in 2003. Haeder specializes in buried precast concrete drainage structures and bridges and was involved in
a project that received the Montana Contractors’ Association Concrete Excellence Award for Technical Merit
in 2010.
—Source: ASTM International
Bentley invests in developer of software

to manage codes and standards for projects
Bentley Systems Inc. has inaugurated a strategic relationship with The Engineering Essentials Co. (TEEC)
by investing in a minority equity position and by placing a Bentley representative on its board.
Based in Philadelphia, Pa., TEEC is developing SpecWave, software to help architectural, engineering, and
construction organizations intelligently manage their engineering specifications and related codes and stan-
dards. The principals of TEEC are Leon Gorbaty, CEO, and Adam Klatzkin, CTO.
—Source: Bentley Systems Inc. J
the search for excellence
the 2012 design awards
200 West Adams Street 200 West Adams Street 200 West Adams Street
Suite 2100 Chicago, IL 60606 Suite 2100 Chicago, IL 60606 Suite 2100 Chicago, IL 60606
Phone: 312-786-0300 Phone: 312-786-0300 Phone: 312-786-0300
Fax: 312-621-1114 Fax: 312-621-1114 Fax: 312-621-1114
www.pci.org www.pci.org www.pci.org
The 50th annual PCi design awards program will be

open for submissions on January 16, 2012. All entries must be submitted
electronically by May 21, 2012.
Visit www.pci.org
200 West Adams and
Street I Suite 2100 I Chicago, click on the “Design Awards” icon for more information.
IL 60606-5230
Phone: 312-786-0300 I Fax: 312-621-1114 I www.pci.org
Contact: Jennifer Peters, jpeters@pci.org or

200 West Adams Street I Suite 2100 I Chicago, IL 60606-5230
Phone: 312-786-0300 I Fax: 312-621-1114 I www.pci.org Brian Miller, P.E., LEED AP, bmiller@pci.org
Project Spotlight
Beach Prisms reduce ing waves before they reach the shoreline. Waves pass
through the slots in the triangular 4 ft × 4 ft × 10 ft (1.2
shore erosion near m × 1.2 m × 3 m) structures.
Chesapeake Bay home Several years of working through the permitting
process led to a final approval meeting in front of the
On September 1, 2006, Hurricane Ernesto roared up Virginia Marine Resources Commission (VMRC). John
Chesapeake Bay and nearly cut Bob and Chris White’s Klein, a consulting coastal engineer working with SMC,
bay-front property in Virginia in half. presented statistical data to the VMRC representing 20
“I can’t believe how powerful the water was,” Bob years of Beach Prism installation and testing. The board
White says. “Our home office was knocked off its foun- voted 11 to 0 in favor of the permit.
dation, the pool was full of mud and debris, and the Tractor trailers traveling from the SMC plant trans-
water had risen over our tennis court and was threaten- ferred the Beach Prisms in five loads to a barge. The full
ing the house.” Subsequent storms have continued to do installation consisted of 22 pieces of Beach Prism with 6
additional damage. in. (150 mm) spacing for a total of about 230 ft (70 m)
The Whites tried several methods to solve their of protection. Each piece was placed parallel to the beach
problems: bulkheads, stone revetments, and five wooden in about 1 ft (0.3 m) of water at low tide. Installation was
groins. More than four years ago, the Whites discovered completed the next day. The Whites hope that this will
Beach Prisms, made by Smith-Midland Corp. (SMC). provide some protection from storm damage and slow or
Beach Prisms are an open precast concrete product stop the erosion.
that works by reducing the amount of energy in incom-
A Virginia couple has Beach Prisms installed in their home on Chesapeake Bay to reduce beach erosion. Courtesy of Smith-Midland Corp.

These Beach Prisms in Chesapeake Bay are designed to mitigate beach erosion while allowing marine life to migrate freely. Courtesy of Smith-Midland Corp.
The slotted design of the Beach Prism allows water to

flow through while dissipating the energy of the waves,
thus reducing their impact on the shoreline. The open
design also allows for marine life such as fish and turtles
to migrate freely. Beach Prisms have been proven in some
cases to allow sand to accumulate in front of and behind
them.
Beach Prisms are for river- and bay-front property
owners who want an alternative to traditional armor
stone, or groins and jetties. They are jointly permitted by
state and federal agencies, including the Army Corps of
Engineers.
—Source: Smith-Midland Corp.

Precast concrete T-wall
performs well in Iraq
Twelve-foot-high (3.7 m) precast concrete T-walls
have been used for blast protection in Iraq and
Afghanistan. The name derives from the T-shaped
cross section. The 12 ft size is also referred to as the
“Texas barrier” to distinguish it from the 6 m (20 ft)
“Alaska barrier.”
The T-wall in the photo shows minimal damage
after being subjected to rocket fire. The crater about
3 ft (1 m) in front of the wall marks the impact loca-
tion of a 107 mm (4.21 in.) insurgent rocket. By con-
trast, the Hesco barrier, a cardboard and wire mesh
caisson filled with sand, was destroyed when subjected
to the same type of rocket landing the same distance
away. Both barriers provide effective blast protection,
but the T-wall can withstand more than one attack.
—Rachel J. Detwiler and James E. Doten J
This precast concrete T-wall in southern Iraq is shown following an impact from a
107 mm (4.21 in.) insurgent rocket about 3 ft (1 m) away. The crater is indicated with an
arrow. Courtesy of MSG James Doten, Minnesota Army National Guard.
This Hesco barrier in southern Iraq is shown following an impact from a 107 mm (4.21 in.)
insurgent rocket about 3 ft (1 m) away. Courtesy of MSG James Doten, Minnesota
Army National Guard.

Save the Date!
What: PCI Committee Days and Membership Conference

When: March 29 – April 1, 2012
Where: Wyndham Hotel, Chicago, Illinois
Join PCI for this outstanding opportunity to network, share ideas and insight with fellow
PCI members, and provide input to the industry’s Body of Knowledge.
You don’t need to be Suite

on
200 West Adams Street200 a committee
West Adams Street
2100 Chicago, IL 60606
to attend – Adams
200 West all members
Street are welcome!
Suite 2100 Chicago, IL 60606 Suite 2100 Chicago, IL 60606
More information will be
Phone: 312-786-0300 available
Phone: 312-786-0300 soon on the PCI website.
Phone: 312-786-0300
Fax: 312-621-1114 Fax: 312-621-1114 Fax: 312-621-1114
Registration will open in early January.

This year’s PCI Committee Days and Membership Conference will take
place concurrently with the Structural Engineering Institute’s Structures
Congress. PCI is a proud sponsor of the 2012 Structures Congress and is
able to offer members these great opportunities to network and build
relationships with structural engineers:
• PCI Committee Days and Membership Conference registration includes a ticket to a

joint PCI/SEI reception on Thursday, March 29.
• PCIAdams
200 West members can
Street I Suite 2100receive
I Chicago, discounted
IL 60606-5230 registration to the Structures Congress.
• PCI has organized a full day of sessions at the Structures Congress on Thursday, March 29.

PCI Continuing Education
Approved Provider
PCI is a registered continuing education provider with the American Institute of Architects, the National Council of
Examiners of Engineers and Surveyors, and the U.S. Green Building Council, and offers continuing education credit
accepted in all 50 states! PCI provides several options for you to engage in continuing education programs including
webinars, seminars, lunch-and-learns, and the new PCI eLearning Center. To learn more, visit www.pci.org/education.
PCI eLearning Center

The new PCI eLearning Center is the first education management system dedicated to the precast concrete and
precast structures industries. Architects and engineers can log on to the site, complete free hour-long courses, and
receive continuing education credit. Check back frequently as new courses are added often.
Webinars
200 West Adams Street
Suite 2100 Chicago, IL 60606
Phone: 312-786-0300
Fax: 312-621-1114
www.pci.org
Phone: 312-786-0300
Fax: 312-621-1114
www.pci.org
Phone: 312-786-0300
Fax: 312-621-1114
PCI webinars are presented live each month by industry experts on a variety of topics, from design and construc-
www.pci.org
tion to sustainability and more. All webinars are FREE, one-hour long, and scheduled to provide a noon start time in
each time zone in the contiguous United States. Webinars provide an inexpensive way to keep up to date on new
and improved materials, products, concepts, and more while earning continuing education credits.
Seminars
PCI and its regional affiliates offer seminars all over the United States on a variety of topics. These seminars include
Quality Control Schools, Structural Design Seminars, Architectural Precast Seminars, lunch-and-learns, and more.
Visit www.pci.org/education for up-to-date seminar listings, additional information, and to register.
U C AT IO
ED
N
NUING

TI
CON
Seismic
Preview of PCI’s New Zealand Earthquake 42
Reconnaissance Team Report
Robert B. Fleischman, José I. Restrepo, Joseph R. Maffei,
and Kim Seeber
Preview of PCI’s Japan Earthquake 47

Richard Sause, Robert Frosch, S. K. Ghosh, Jason Lien,
Clay Naito, Larbi Sennour, and Toshi Yamanishi
Observations from the February 27, 2010, 52

Earthquake in Chile
Effects of Loading Conditions 76

on the Behavior of Unbonded
Post-tensioning Strand-Anchorage Systems
Seismic Performance and Retrofit 98

of Precast Concrete
Grouted Sleeve Connection
Structural Behavior of an Innovative 110

All–Precast Concrete Dual System
for Residential Buildings
Architectural Precast Concrete Panel 124

Systems Used for Lateral-Force Resistance
J. Paul Hobelmann, Macarena Schachter,
and Matthew C. Cooper
P h o t o : J o s e p h R . Ma f f e i
PCI Reconnaissance Team Report
Preview of PCI’s
New Zealand earthquake
reconnaissance team report
Robert B. Fleischman, José I. Restrepo, Joseph R. Maffei,

and Kim Seeber
Precast concrete structural Emulative design, which requires critical sections to re-
design in New Zealand spond under load as they would for cast-in-place concrete,
together with the capacity design philosophy, was crucial
recast concrete has been used in New Zealand for gaining the acceptance of structural engineers in New
since the early 1960s and became widely accepted Zealand. In capacity design, the primary structural system
in the 1980s as an effective construction method for resisting seismic forces is first designed to have a
for seismic resistance. Since then, precast concrete suitable mechanism of accommodating nonlinear lateral
has been extensively used there. Park described the evolution deformation, with the engineer explicitly defining the loca-
of precast concrete design in that country.1 tions of plastic hinges. Plastic hinge regions are designed
and detailed for strength and ductility during a severe
During the building boom of the mid- to late 1980s, the speed earthquake. Then the remainder of the structural system is
of construction and high quality of precast concrete gave it provided with sufficient strength to avoid other possible
significant advantages over cast-in-place concrete. The New failure modes and maintain ductility after flexural over-
Zealand concrete design standard2 of the time included com- strength develops at the plastic hinges.
prehensive provisions for the seismic design of cast-in-place
concrete structures but did not cover all aspects of precast Resisting lateral forces
concrete structures. A study group of the New Zealand from earthquakes
Concrete Society was formed in 1988 to summarize data on
precast concrete design and construction, recommend best Moment-resisting frames have been the most common
practices, and identify areas for further research. A 1995 revi- seismic-force-resisting system for tall buildings in New
sion of the New Zealand concrete design standard3 included Zealand. A beam sidesway mechanism, in which plas-
new provisions for precast concrete structures based on the tic hinges form in the beams due to strong columns and
research of the early 1990s.1 weak beams, is the preferred nonlinear mechanism. To
ensure that flexural hinging does not occur at locations not

designed for ductility, or that shear failure cannot occur
anywhere in the structure, the maximum actions likely to
be imposed are calculated from the flexural overstrengths
at the plastic hinges.1 Tall buildings in New Zealand were
typically constructed using perimeter moment frames with
relatively closely spaced stiff columns and stiff beams.
Interior beams and columns carried gravity load; their
contribution to seismic resistance was neglected.
Structural precast concrete walls (shear walls) have also

been used to provide seismic resistance in medium-rise
buildings in New Zealand. In some cases precast concrete
walls and frames are combined to form dual systems.
These walls are designed to develop a plastic hinge at the
base. The most important consideration in detailing plastic
hinge regions for ductility is to provide sufficient trans-
verse reinforcement, whether rectangular stirrups, hoops,
or spirals. The transverse reinforcement acts as shear rein-
forcement, confines the concrete, and prevents premature
buckling of the longitudinal reinforcement.1
Earthquake
of February 22, 2011
A magnitude 6.3 earthquake struck Christchurch, New

Zealand, at 12:51 p.m. on Tuesday, February 22, 2011. The
epicenter was located 10 km (6 mi) southeast of the city
center.4 The earthquake caused 182 deaths and damaged
thousands of buildings.5
New Zealand lies on the Pacific Rim and has a history Figure 1. PricewaterhouseCoopers building under construction in Christchurch,
New Zealand.
of strong earthquakes. However, Christchurch is about
240 km (150 mi) from the nearest known major fault, and
thus was not expected to experience a large earthquake. It Case study:
is in a moderate seismic hazard zone comparable to that of PricewaterhouseCoopers
Portland, Ore. The February earthquake, along with a less- office building
damaging earthquake the previous September, indicates the
existence of fault lines much closer to Christchurch. The 21-story PricewaterhouseCoopers office building was
constructed in the 1980s (Fig. 1). It was designed with
In the February 22 earthquake, an interval of extremely an exterior moment-resisting frame and constructed us-
high-intensity shaking lasted approximately 8 sec. The ing a detail in which the precast concrete joint (Fig. 2) is
average response spectrum of the ground motions recorded placed over the column below. Vertical ducts in the precast
in the city center was approximately equivalent to the concrete joint allow the vertical reinforcement from the
Christchurch maximum considered earthquake and was column to be threaded through and into the column above.
stronger for substantial ranges of typical building natural The ducts are grouted once the joint is in place.1
frequencies. For individual records, the ground motion
contained frequency ranges where it was substantially Figure 3 shows the perimeter moment-resisting frame
stronger, similar to the maximum considered earthquake in (background) and the interior “gravity” frame (foreground)
Berkeley, Calif.6 The February event was extremely strong following the earthquake of February 22. As can be seen,
near the epicenter, with vertical accelerations of up to 2g. the columns are more closely spaced in the perimeter
Soil liquefaction was significant and widespread due to the frame.
presence of a high water table in sandy soil. Deep founda-
tions are fairly rare in Christchurch, and several buildings A visual survey of this building by the authors showed a
and bridges showed evidence of support movement or dif- pattern of damage consistent with capacity design (Fig. 4).
ferential settlement. The lowest two floors, situated above a parking garage and
undergoing less deformation, showed no plastic hinging.
The lower- to midlevel floors showed the greatest beam de-

formation and damage, and the upper floors less damage.
Figure 4 shows the damage to the perimeter frame on the
interior of the 10th floor. The north-south beam exhibits
flexural-torsional cracks from earthquake displacement
perpendicular to the frame, while the east-west beam
exhibits flexure cracks. The authors judge that the flexural
cracking observed does not reduce the strength of the
beams and does not substantially diminish their continued
deformation capacity or the ability of the structure to resist
future earthquakes.
The exterior moment-resisting frame of the Pricewater-

houseCoopers building performed as intended, forming the
desired flexural plastic hinges at the ends of the beams. No
structural damage was observed in the columns. Where the
damage was greatest, cracking and flexural torsion were
associated with these hinges. The extent of the damage
is not surprising given the extreme intensity of the earth-
quake motions. The authors judge the office building to
have met or exceeded its design requirements to avoid col-
lapse in the maximum considered earthquake.
Final report
A full report covering the performance of a number of

Figure 2. Typical precast concrete corner unit being installed in perimeter other types of precast concrete structural systems during
moment frame of the PricewaterhouseCoopers building.
the February 22, 2011, New Zealand earthquake, some
Figure 3. Corner joint in first-level parking podium area after February 22, 2011, earthquake in New Zealand. Photo courtesy of Franz Kahn.

Figure 4. Corner joint in upper floor indicated damage suffered in February 22, 2011, earthquake in New Zealand. The north-south beam (upper left) exhibits flexural-
torsional cracks, while the east-west beam (upper right) exhibits flexure cracks.
of which suffered serious damage, will be published in a 5. The Canterbury Earthquakes Royal Commission.
future issue of the PCI Journal. canterbury.royalcommission.govt.nz.
References 6. Earthquake Engineering Research Institute (EERI).

2011. Learning from Earthquakes: The M 6.3 Christ-
1. Park, Robert. 2002. Seismic Design and Construction church, New Zealand, Earthquake of February 22,
of Precast Concrete Buildings in New Zealand. PCI 2011. EERI Special Earthquake Report. Oakland, CA:
Journal, V. 47, No. 5 (September–October): pp. 60–75. EERI.
2. Standards Association of New Zealand. 1982. Code of

Practice for the Design of Concrete Structures. NZS
3101:1982. Wellington, New Zealand: Standards As-
sociation of New Zealand.
3. Standards Association of New Zealand. 1995. The

Design of Concrete Structures. NZS 3101:1995. Wel-
lington, New Zealand: Standards Association of New
Zealand.
4. GNS Science. 2011. The Canterbury Earthquake

Sequence and Implications for Seismic Design Levels.
Consultancy report 2011/183 to Canterbury Earth-
quakes Royal Commission. Compiler: T. H. Webb.

About the authors Reader comments
Robert B. Fleischman, PhD, is an Please address any reader comments to journal@pci

associate professor in the Depart- .org or Precast/Prestressed Concrete Institute, c/o PCI
ment of Civil Engineering and Journal, 200 W. Adams St., Suite 2100, Chicago, IL
Engineering Mechanics at the 60606. J
University of Arizona.
José I. Restrepo, PhD, is a

professor in the Department of
Structural Engineering at the
University of California at San
Diego.
Joseph R. Maffei, S.E., PhD,

LEED AP is a principal at
Rutherford & Chekene Engineers
in San Francisco, Calif.
Kim Seeber, P.E., FPCI, is a

senior vice president at Seaboard
Services of Virginia Inc. in
Taylors, S.C.

Preview of PCI’s
Japan Earthquake
Richard Sause, Robert Frosch, S. K. Ghosh, Jason Lien,

Clay Naito, Larbi Sennour, and Toshi Yamanishi
he authors formed the PCI reconnaissance team The tsunami inundated many locations along the north-
that visited selected locations (Fig. 1) in Japan east coast of Honshu. Depending on offshore topography
from June 27–30, 2011, following the Tohoku and distance from the epicenter, the depth of water above
earthquake of March 11, 2011. grade on shore varied from 1 m (3.3 ft) to 8 m (26 ft) in
the coastal plain region of Natori to more than 18 m (59
Tohoku earthquake ft) in the Ria coastal town of Onagawa.3 A tsunami wave
typically crests before reaching the shore, creating a bore,
At 2:46 p.m. on March 11, 2011, a magnitude 9.0 earthquake which can move rapidly and produce impulsive forces on
occurred near the east coast of Honshu, the main island of Ja- structures in its path. The subsequent inundation produces
pan.1 This earthquake, known as the 2011 Tohoku earthquake, high-velocity water flow around and through structures.
caused strong ground shaking throughout much of eastern The effects of these forces on structures were widespread
Honshu and triggered tsunami waves that caused widespread in the tsunami-affected coastal regions.
damage at many locations along the northeast coast of Japan.
Earthquake-resistant
The epicenter of the earthquake was located at 38.297° building design in Japan
N, 142.372° E,1 about 70 km (44 mi) off the coast. Strong
ground shaking occurred in Sendai, the major city (130 km The lack of widespread significant structural damage to
[80 mi]) nearest the epicenter. Peak ground accelerations buildings from ground motions associated with the 2011
exceeding 2g were recorded at locations near the east coast Tohoku earthquake has been noted by other observers.2
of Honshu.1 Some ground shaking–induced damage was This result has been attributed to important revisions over
observed even in the Tokyo metropolitan region, more than the past 40 years to the Japanese national building code
350 km (210 mi) southwest of the epicenter; 1 however, wide- and related national standards. The Japanese national
spread structural damage was not observed across eastern code, the Building Standard Law of Japan (BSLJ), which
Honshu.2 includes the Building Standard Law, the Enforcement

Figure 1. Locations in Japan visited by PCI’s reconnaissance team.
Order of the Building Standard Law, and the Enforcement

Regulations of the Building Standard Law,4–6 specifies de-
sign loads, allowable stresses, and other requirements. The Figure 2. Tension-tie precast concrete frame retrofit of Shiogama Municipal No.
details of structural design are specified in standards issued 2 Junior High School.
by the Architectural Institute of Japan (AIJ). These AIJ
standards, prepared separately for each structural material, Performance of
are supplements to the BSLJ. precast concrete structures
The 1968 Tokachi-oki earthquake caused significant dam- The authors noted a lack of widespread earthquake damage
age to buildings, and a revision to the BSLJ5 reduced the to buildings in the affected areas, including Sendai, where
spacing of steel ties in reinforced concrete columns to 100 the recorded peak ground acceleration on rock exceeded
mm (4 in.). In 1971, a major revision of the AIJ standard 0.34g. The authors observed some damage to conventional
for reinforced concrete7 incorporated ultimate-strength reinforced concrete buildings at the Aobayama Campus of
design for shear of beams and columns, including more Tohoku University, which is located in a hill zone above
stringent shear reinforcement requirements. These changes the city of Sendai.
are comparable to important code changes in the United
States following the 1971 San Fernando earthquake in The authors observed the exteriors of several precast
California. Post-1971 reinforced concrete structures per- concrete structures, all of which performed well during
formed much better in the 1995 Kobe earthquake than their the earthquake with no indication of significant damage.
pre-1971 counterparts, primarily because of the improved Several of these structures are discussed in the following
shear design of columns.8 paragraphs.
The 1978 Miyagiken-oki earthquake caused significant The Shiogama Municipal No. 2 Junior High School
damage to buildings and led to a 1981 revision of the (Fig. 2) was retrofitted in 2010 using an external precast
BSLJ,6 which introduced two-phase earthquake-resistant concrete frame braced with tension ties. The existing
design. The first-phase design (essentially the allowable cast-in-place concrete building was deficient in torsional
stress design from the previous BSLJ) is intended to pro- resistance to earthquake loading. The north facade had stiff
tect a building against loss of function in ground motions shear walls, while the south facade had moment frames,
expected to occur several times during its lifetime, with creating a torsional stiffness irregularity. The tension-tie-
peak ground accelerations in the range of 0.08g to 0.10g. braced precast concrete frame increased the lateral stiff-
The second-phase design is intended to ensure safety under ness along the south facade. The pieces included precast
a ground motion expected to occur once in the lifetime of concrete columns anchored to the foundation using type II
a building, with peak ground accelerations in the range of mechanical splices, architectural medallions on top of the
0.3g to 0.4g. Post-1981 structures designed by the two- columns, and drop beams between the medallions post-
phase procedure performed well in the 1995 Kobe earth- tensioned together on-site. To enhance the rigidity of the
quake. precast concrete frame, X-bracing with steel rods covered
with plastic sleeves was used in the openings. Cast-in-
Current earthquake-resistant design of reinforced concrete place concrete diaphragms (Fig. 3) between the existing
buildings in Japan is consistent with the 1981 BSLJ6 and structure and precast concrete frame were used to transfer
the 1971 AIJ standard for reinforced concrete.7 lateral shear forces to the external precast concrete frame.
No damage to this structure was observed.

Figure 3. Cast-in-place concrete diaphragm of retrofit of Shiogama Municipal
No. 2 Junior High School.
Figure 5. Minor spalling at the base of the columns of Namagachi Lions Tower.
to have performed well.
The four-story total–precast concrete residential apartment

building shown in Fig. 7 was constructed in Natori City in
2006. The building has shear walls to provide earthquake
resistance. Viewed from the exterior, the building appeared
to have no structural damage or distress. Similar apartment
buildings in the Kobe and Osaka area performed well in
the 1995 Kobe earthquake.8
The final report

Figure 4. Total–precast concrete Namagachi Lions Tower in Sendai.
A full report covering all of the observations made by

The Namagachi Lions Tower (Fig. 4) in Sendai is a PCI’s Japan earthquake reconnaissance team will be pub-
29-story total–precast concrete tower on a two-story, cast- lished in a future issue of the PCI Journal.
in-place concrete podium. The structure was completed in
1999, and the lateral-force-resisting system consists of pre- References
cast concrete moment-resisting frames. There was minor
spalling near the base of three columns on the front of the 1. U.S. Geological Survey. Magnitude 9.0—Near the
building (Fig. 5). The only other visible damage was spall- East Coast of Honshu, Japan. earthquake
ing of the underside of the slab (Fig. 6) at two opposite .usgs.gov/earthquakes/eqinthenews/2011/
corners of the building, where two orthogonal beams that usc0001xgp/#summary.
appeared to be unconnected came together. This damage
did not appear to be significant, and the building appeared 2. Building Research Institute. Summary of the Field

Figure 7. Three-story total–precast concrete apartment building at Natori City.
Figure 6. Spalling of slabs at corners of Namagachi Lions Tower.
Survey and Research on “The 2011 off the Pacific

Coast of Tohoku Earthquake” (the Great East Japan
Earthquake). www.kenken.go.jp/english/contents/
topics/20110311/0311summaryreport.html.
3. D. Cox, Oregon State University, Department of Civil

and Construction Engineering, Corvallis, Ore., email
message including unpublished data, 2011.
4. Building Guidance Division, Housing Bureau, Min-

istry of Land, Infrastructure, Transport and Tourism
(MLIT). 2007. Building Standard Law of Japan.
Tokyo, Japan. MLIT. CD-ROM.
5. MLIT. 1971. Building Standard Law of Japan. Tokyo,

Japan: MLIT.
6. MLIT. 1981. Building Standard Law of Japan. Tokyo,

Japan: MLIT.
7. Architectural Institute of Japan (AIJ). 1971. Standard

for Structural Calculation of Reinforced Concrete
Structures. Tokyo, Japan: AIJ.
8. Ghosh, S. K. 1995. Observations from the Kobe

Earthquake of January 17, 1995. Engineered Concrete
Structures, V. 8, No. 1 (April): pp. 1–5.

About the authors Larbi Sennour, PhD, P.E., S.E.,
FPCI, FACI, is president of The
Richard Sause, PhD, P.E., is the Consulting Engineers Group Inc.
Joseph T. Stuart Professor of in San Antonio, Tex. He is a
Structural Engineering and the member of the PCI Board of
director of the Advanced Technol- Directors, chair of PCI’s Techni-
ogy for Large Structural Systems cal Activities Council, and
(ATLSS) Center at Lehigh member of the Industry Handbook, Connections, and
University in Bethlehem, Pa. Materials Technology committees.
Sause is past chair of the PCI Seismic Committee.
Toshi Yamanishi is president of
Robert J. Frosch, PhD, P.E., FACI, Splice Sleeve North America Inc.
is a professor of civil engineering He worked with a Japanese
at Purdue University in West general contractor before joining
Lafayette, Ind. He is a member of Splice Sleeve and was based in
the PCI Research and Develop- Japan, the United States, Hong
ment Committee and the Techni- Kong, Vietnam, and Thailand.
cal Activities Council.
Abstract
S. K. Ghosh, PhD, is president of
S. K. Ghosh Associates Inc., in This preview highlights observations made by PCI’s
Palatine, Ill., and Aliso Viejo, Japan earthquake reconnaissance team, which visited
Calif. He has influenced seismic selected locations in Japan June 27–30, 2011, follow-
design provisions in the United ing the Tohoku earthquake of March 11, 2011. A full
States for many years by author- report covering all of the team’s observations will be
ing many publications and by published in a future issue of the PCI Journal.
serving on or chairing numerous committees and
advisory panels. He is PCI’s code consultant. Keywords
Jason Lien is vice president of Code, earthquake, Japan, preview, seismic, tsunami.
design for the corporate office of
EnCon. He is on the Technical Reader comments
Activities Coouncil and the
Continuing Education Committee. Please address any reader comments to journal@pci
He is chair of the Building .org or Precast/Prestressed Concrete Institute, c/o PCI Jour-
Information Modeling Committee. nal, 200 W. Adams St., Suite 2100, Chicago, IL 60606. J
He is a registered Professional Civil Engineer in
Colorado, California, New Mexico, Wyoming, and the
Territory of Guam.
Clay Naito is an associate

professor of structural engineering
and associate chair of the Lehigh
University Department of Civil
and Environmental Engineering.
He is a licensed professional
engineer in Pennsylvania and
California. Naito is chair of the PCI Blast Resistance
and Structural Integrity Committee.

Observations
from the February 27, 2010,
earthquake in Chile
he earthquake that shook Chile at 3:34 a.m. on

Saturday, February 27, 2010, was one of the
most devastating in the history of the country,
which has a 2650 mi (4270 km) coastline along
the Pacific Ring of Fire. The moment magnitude issued by
the U.S. Geological Survey1 was 8.8. The earthquake was
followed by hundreds of aftershocks, the strongest
■ The authors were part of a PCI reconnaissance team of inves- measuring from 6.0 to 6.9 on the moment magnitude scale.
tigators who went to various locations affected by the February Table 1 gives the details of the earthquake.1
2010 earthquake in Chile.
Maximum ground acceleration of up to 0.65g was recorded
■ The 1996 Chilean seismic code was similar to the then-current at Concepción, and more than 6.6 ft (2 m) of uplift was
UBC and ACI 318 codes, except that boundary elements and observed near Arauco on the coast.1
special transverse reinforcement were not required in structural
walls. This exception was revoked in the 2009 version of the The earthquake was generated at the gently sloping fault
Chilean code due to a trend towards thinner walls. along which the Nazca plate moves eastward and downward
beneath the South American plate (Fig. 1). The two plates
■ The number of deaths and the amount of property loss were are converging at 2¾ in. (70 mm) per year. The fault rupture,
not disproportionate to the severity of the earthquake. Much of largely offshore, exceeded 60 mi (100 km) in width and
this is attributable to Chile’s history of adoption and implemen- extended nearly 300 mi (500 km) parallel to the coast.
tation of adequate building codes.
A comprehensive written record beginning in the mid-16th
■ The 2010 emergency changes to the Chilean Building Code century describes large damaging earthquakes throughout
have far-reaching implications for the special structural wall the region that was affected by the February 27, 2010,
design provisions in ACI 318. earthquake. An 1835 M8.2 (M = moment magnitude)

Table 1. Earthquake details
Concepción earthquake is notable because famed naturalist
Moment magnitude 8.8
Charles Darwin and naval officer Robert FitzRoy provided
observations and comments.1 Since the beginning of the Saturday, February 27, 2010,
Date-time
20th century, there have been M8.2 earthquakes in 1906, at 03:34:14 a.m. at epicenter
1943, and 1960, and an M8.0 earthquake in 1985.1 The
1960 M8.2 earthquake was a foreshock that occurred a day Location 35.909°S, 72.733°W
before the great M9.5 Chilean earthquake of 1960.1 Depth 21.7 mi
The 2010 earthquake that is the subject of this paper struck Region Offshore Maule, Chile
in an area previously identified as a seismic gap extending 60 mi NW of Chillán, Chile
from Constitución in the north to Concepción in the south
with a projected worst-case potential to produce an earth- 65 mi NNE of Concepción, Chile
quake with M between 8.0 and 8.5.2 The rupture extended Distances
beyond the northern and southern boundaries of the gap, 70 mi WSW of Talca, Chile
overlapping extensive zones already ruptured in 1985 and
1960.3 210 mi SW of Santiago, Chile
Strong-motion records Source: U.S. Geological Survey. Note: 1 mi = 1.61 km.
The University of Chile’s strong-motion instrumenta-

tion array recorded motions at several sites in the heav- ment response spectra for the same ground motions. The
ily stricken region. Some of the digital data have been acceleration spectra show unusual second peaks at periods
processed and reported by the University of Chile4 and by of 1.5 sec and longer. The acceleration and displacement
Boroschek et al.5 Figure 2 shows three sets of recorded ac- spectra are compared with the design spectra for soil types
celerograms and the corresponding response spectra from II, III, and IV as defined in NCh433-2009.
the Santiago area. Ground accelerations exceeding 0.05g
lasted more than 60 sec according to most of the records. Figure 4 shows horizontal ground motion accelerograms
Elastic response spectra of several records are higher than from Colegio San Pedro, across the Bio Bio River south-
the elastic design spectrum of the Chilean seismic code, west of downtown Concepción, along with their accelera-
NCh433-2009.6 tion and displacement response spectra. The acceleration
spectra show second peaks at periods of about 0.8 sec.
Figure 3 shows horizontal ground motion accelerograms Again, the design spectra for soil types II, III, and IV are
from downtown Concepción. A special characteristic of also shown.
the records is the long duration of strong shaking (90 sec
or more). Also shown are the acceleration and displace-
Figure 1. The source of the Chile earthquake is at the convergence of the Nazca and the South American plates. Source: Roberto Leon presentation at www.eqclear
inghouse.org/20100227-chile/wp-content/uploads/2010/04/Leon-Chile-Earthquake.pdf. Note: 1 mi = 1.61 km.

Figure 2. Accelerograms and corresponding acceleration response spectra (β = 5%) from the Santiago, Chile, area. Source: Boroschek et al. 2010. Note: g = accel-
eration due to earth’s gravity; EW = east-west; NS = north-south; Sa = spectral acceleration; T = time; UD = up-down; β = damping coefficient.
Figure 3. Accelerograms and corresponding acceleration and displacement response spectra (β = 5%) from downtown Concepción, Chile. Source: Boroschek et al.
2010. Notes: g = acceleration due to earth’s gravity; Sa = spectral acceleration; Sd = spectral displacement; β = damping coefficient. 1 cm = 0.4 in.

Figure 4. Accelerograms and corresponding acceleration and displacement response spectra (β = 5%) from Colegio San Pedro, Chile. Source: Boroschek et al. 2010.
Notes: g = acceleration due to earth’s gravity; Sa = spectral acceleration; Sd = spectral displacement; β = damping coefficient. 1 cm = 0.4 in.
PCI investigation The Structural Engineering Institute of the American

Society of Civil Engineers (ASCE) sent an earthquake
The entire team visited Chile April 26–29, 2010, to inves- assessment team to assess the effectiveness of Chile’s
tigate damage in Santiago, Concepción, Talca, Chillán, building methods and codes, which closely parallel those
Coronel, and Chillán Viejo. Three team members spent an used in the United States. The primary purpose of the
additional day visiting Valparaiso/Viña del Mar. team was to determine whether changes are warranted to
the U.S. codes, standards, or practice in general and to
Other investigations ASCE 7-10 Minimum Design Loads for Buildings and
Other Structures8 and ASCE 41-06 Seismic Rehabilita-
A team organized by the Earthquake Engineering Re- tion of Existing Buildings9 in particular. Part of the team
search Institute (EERI) investigated the effects of the Chile traveled to different locations to study structures built after
earthquake. The team was assisted by local university the 1985 earthquake, when more-detailed building codes
faculty and students. Geotechnical Extreme Events Recon- were implemented. The remainder of the team focused on
naissance (GEER) contributed geosciences, geology, and structures such as steel mills and power plants.
geotechnical engineering findings. The Technical Council
on Lifeline Earthquake Engineering (TCLEE) contributed The report of the assessment team has not been published
a report based on its reconnaissance. Based on its own yet, so no definite conclusions are available. The team ob-
investigation and the GEER and TCLEE input, EERI pub- served several differences between the Chilean standards
lished the EERI Special Earthquake Report—June 2010 as and those in the United States. For example, the walls of
an insert in EERI’s monthly newsletter.3 EERI also set up a buildings are much thinner than is required in the United
Chile Earthquake Clearinghouse.7 States and do not contain as much reinforcement. Despite
its observations of significant nonstructural damage and
their review of plans, the team did not identify anything

ment in walls of bearing-wall buildings may be necessary
at the extremities of walls having T-shaped, L-shaped, or
similar cross sections, but that confinement is typically not
required for symmetrically reinforced, rectangular wall
cross sections. Wallace and Moehle15 went on to state,
“The good performance of the majority of these [bear-
ing wall] buildings during the March 3, 1985, earthquake
suggests that bearing walls with limited detailing may be
an effective construction form for earthquake resistance.
Although buildings in Chile are designed for roughly the
same lateral forces as those in regions of high seismic risk
in the United States, the typical structural wall in a Chilean
building does not require boundary elements or special
transverse reinforcement.”
Figure 5. Damaged coupling beam in Viña del Mar, Chile.
Based on this, NCh433-199611 contained clause B.2.2,
which states, “When designing reinforced concrete walls,
that would necessitate substantive changes to U.S. stan- it is not necessary to meet the provisions of paragraphs
dards such as ASCE 7-108 or ASCE 41-06.9 21.6.6.1 through 21.6.6.4 of the ACI 318-9513 code.”
These ACI 318-95 sections are for specially confined
The Los Angeles Tall Buildings Structural Design Council boundary elements at the edges of shear walls. NCh433-
also sent a team. Its report is not yet available; however, a 20096 rescinded this exception before the February 2010
presentation is posted on its website.10 earthquake because of a trend to use thinner walls more in
recent years than in the past.
With so much information already available or coming
soon, this report concentrates largely on the performance of Tall concrete buildings are typically found in the metro-
precast concrete structures, though some other aspects are politan areas around Santiago, Valparaiso/Viña del Mar,
also included. and Concepción. In Viña del Mar, a number of buildings
that were damaged in the 1985 earthquake and repaired
Building performance suffered significant damage once again. However, damage
was largely concentrated in newer buildings. The failure
Mid- to high-rise buildings in Chile are predominantly of of one tall building in Viña del Mar was due to the wide
reinforced concrete construction. Most of these buildings use spacing of transverse reinforcement in shear walls, which
structural walls to resist both gravity loads and earthquake caused the vertical bars to buckle, in this particular case,
forces. Dual systems of walls and frames are occasionally without fracture. In many other cases the vertical bars did
used in newer construction. Typical wall cross-sectional fracture.
area–to–floor area ratios are high compared with values
commonly used in U.S. concrete building construction. Coupling beams over doorways typically have inadequate
reinforcement. Many of these beams suffered damage
In 1996, the Chilean seismic code (NCh433-1996)11 adopt- (Fig. 5). Some buildings lacked coupling beams. In many
ed analysis procedures similar to those in the 1997 Uniform of those cases, damage resulted from the slab acting as
Building Code (UBC).12 However, there are no prohibitions the coupling element. There were several instances of
or penalties related to vertical or horizontal system irregu- doors that jammed because of displacements in the walls
larities. NCh433-1996 also enforces provisions of Building on either side. Spalled cover on top of lap splices of wall
Code Requirements for Structural Concrete (ACI 318-95) boundary reinforcement was a common occurrence.
and Commentary (ACI 318R-95),13 with one significant
exception, as noted in the following paragraph. Four mid- to high-rise concrete buildings collapsed com-
pletely or partially. Two of these were nearly identical,
Having observed and investigated the performance of side-by-side, five-story buildings in Maipú, Santiago (not
reinforced concrete buildings during the 1985 Chile earth- visited by the PCI team). According to the EERI Special
quake, Wood14 and Wallace and Moehle15 reached nearly Earthquake Report,3 these buildings had four stories of
identical conclusions. The primary variables that determine condominium units atop a first-story parking level with an
the need for confined boundary elements in shear walls irregular wall layout. Wall failures apparently contributed
were found to be the ratio of wall cross-sectional area to to the collapses.
floor-plan area, the wall aspect ratio and configuration, the
axial load on the wall, and the reinforcement ratio of the A third collapsed building was the 15-story Alto Río con-
wall. Wallace and Moehle concluded that concrete confine- dominium in Concepción (Fig. 6). The team was unable

to examine closely the side of the building toward which it
collapsed. According to the EERI Special Earthquake Report,3
the structural drawings indicated that concrete walls on the
facade were discontinuous and that the wall lengths were
decreased in the first story on the side toward which the build-
ing collapsed. There was ample indication that the building
had rotated about its corridor walls as it collapsed, leading to
tension failures of the transverse walls on the side from which
the photo was taken. Some of the wall vertical reinforcement
fractured, and some lap splices failed on the tension side.
The fairly new 23-story O’Higgins office building in Con-

cepción suffered partial story collapses at levels 10, 14, and
18, each coincident with a framing setback (Fig. 7). The
perforated shear walls on the east face (shown) and south
Figure 6. The 15-story Alto Río Condominium in Concepción, Chile.
face showed damage to both wall piers and spandrels. The
exterior north and west faces appeared undamaged.
crushing and buckling or fracture of reinforcement
The following observations of building performance across entire wall (unzipping) should be investigated.”
emerged:
Precast concrete buildings
• Axial stress in shear walls. As mentioned previously,
Chilean buildings typically contain many shear walls. The precast concrete construction market in Chile does not
This contributed to their relatively good performance include parking structures but does include bridges, office
during the 1985 earthquake. Newer buildings appear to buildings, stadiums, warehouses, and industrial buildings.
have the same shear wall area in terms of the percent- Some systems did not fare well during the February 27,
age of floor area, but many are significantly taller than 2010, earthquake. Many buildings of more recent construc-
before. This suggests that the axial stress in the walls tion did well, and some advanced precast concrete con-
of newer buildings is significantly higher than in older cepts proved their merit.
buildings. This may, at least in part, be responsible for
the widely observed localized wall damage character- Gable frames
ized by buckling of vertical reinforcement.
One precast concrete system that did not perform well was
• Confinement of wall boundary elements. The a precast concrete gable frame system at Parque Industrial
exception made in NCh433-19966 to the specially Escuadrón. The roof of the single-story San José Fishery
confined boundary zone requirements of ACI 318-9513 was formed by a series of these frames. The structure was
was explained previously. Considerable damage was reported to be 23 years old but appeared older. Figure 8
observed in many wall boundary elements, including shows the portion of the building still standing after the
crushing of concrete and buckling and fracture of lon- earthquake.
gitudinal reinforcement. The exception was rescinded
in NCh433-2009. However, the ACI 318-0816 require-
ments have now come into question. The trigger for
requiring specially confined boundary zones should be
reexamined. The reduction of the boundary element
confinement requirements when specially confined
boundary zones are not triggered should also be
reviewed.
• Vertical wall reinforcement. This item has been

described by Wallace:17 “Many damaged walls were
lightly reinforced and had unconfined lap splices.
These walls were observed to have problems at lap
splices or to suffer tension failures (or fractures during
buckling following tensile elongation). Due to the long
duration of the earthquake, the walls likely underwent
a large number of cycles of loading. The possibility
of a failure mode consisting of progressive concrete
Figure 7. The 23-story O’Higgins office building in Concepción, Chile.

roof infill did not form a continuous diaphragm.
Column base failures revealed base anchor bars lapped with

column bars lacking standard hooks and a lack of confine-
ment reinforcement. The remaining debris also showed that
some sections were hollow, formed with expanded poly-
styrene cores. These gable frames lacked the strength and
ductility of special moment frames suitable for high seismic
Figure 8. Precast concrete gable framed San José Fishery after the earthquake. application.
It did not appear that this framing system was in common

The gable frames were assembled using three standard use. The PCI team did not find any examples of this fram-
parts, which include end columns, interior columns, and ing of more recent vintage than at the San José Fishery.
drop-in gables (Fig. 9). The columns have monolithic knee
joints and include parts of the sloping gable members. Precast and cast-in-place concrete
shear walls
The roof was constructed over spaced precast concrete
purlins that spanned between rows of these frames. The Based on the examples of precast concrete construction
gable frame construction was similar to precast concrete that the PCI team was able to find, the industrial buildings
frames that performed poorly in the 1999 earthquake near that were constructed with precast concrete often included
Izmit, Turkey. There are a few notable differences. In Tur- precast concrete walls as cladding. Unlike the practice in
key, the frame across the top of the column was a separate the United States, however, walls were not often used as
precast concrete element spliced to the column and not cast the primary lateral-force-resisting system (LFRS). Exam-
monolithically. The connections at the drop-in gables in the ples where walls provided the lateral bracing were found
Turkish frames were pinned. The connections observed at in a pair of warehouse buildings that were constructed with
the Chilean fishery were welded and apparently intended long-span gable beams on columns and clad with double-
to provide strong connections for continuity. Some welded tee walls.
connections failed by fracture of the reinforcement welded
to the embedded parts for lap and development with the The bays were 39 ft (12 m) wide and 66 ft (20 m) long.
precast concrete frame reinforcement. It is likely that the Gable-shaped girders spanned the 66 ft, and spaced precast
bars were not weldable. Another difference was that the concrete purlins spanned the 39 ft. The ends of the girders
Turkish precast concrete frames lacked any lateral bracing were fixed to the tops of the columns, but the purlins were
perpendicular to the plane of the gable frame. At the Chil- pinned to the girders. The purlins were not continuous. At
ean fishery, there were precast concrete diagonal braces in the outside edges, two-stemmed channels spanned between
the end bays that remained standing. It is unclear whether the girder-column frames, providing a stiff lateral support to
there was additional diagonal bracing in the collapsed bays, the exterior cladding made with precast concrete double-tee
but it was reported that the collapse started in one corner walls. Figure 10 shows a building portion with this framing.
and progressed across the building to the braced bays still At the front of the building, the tee stems were turned out,
standing. The system of spaced purlins and light corrugated and at the rear they were turned in. Concrete planks spanned
Figure 9. Gable frame system.

Roof plank
Purlins
Edge channel Gable-shaped girder
Column
Figure 10. Framing system that uses long-span gable beams on columns and is clad with double-tee walls in an industrial building with offices.
over the purlins, but there did not appear to be connections the other end, just inside the two-level bay with offices at
between planks to form a diaphragm. There was no diagonal the front of the building. There were interior walls across
bracing in the plane of the roof deck. The wall cladding was the building at the first bay that included two floors of of-
connected to the precast concrete roof using long threaded fice space, but this did not provide effective bracing to the
rods that connected to the channels on the sides and the framing at the distant end. The column-to-girder connec-
girders at the ends. These connections appeared to provide tions apparently provided sufficient continuity for frame
out-of-plane resistance but not a load path to transfer lateral behavior in the direction of the frame. Figure 11 shows a
forces into the plane of the wall cladding. sketch of the building plan.
The shear walls for these buildings were isolated cast-in- The team learned from the owner’s representative that
place concrete walls in two bays on each side. These bays one of the buildings had soil saturation problems during
were the last short bay at one end and the second bay from construction that required soil improvement to a depth of
Shear wall
Shear wall
Interior wall Purlins
Purlins
Girder
Shear wall Shear wall

Office area (two levels)
Figure 11. Roof framing plan used in an industrial building with offices. Note: All measurements are in meters. 1 m = 3.28 ft.

Figure 12. Front elevation of an office and warehouse building showing failed Figure 14. Reinforced concrete column with bar buckling and lack of
beam-column framing. confinement.
Figure 13. Rear elevation of the same office and warehouse building showing Figure 15. Bent and broken connection angle that failed to hold walls to edge
fallen exterior walls and exposed transverse walls and end columns. beam at roof.
4.6 ft (1.4 m). The other did not. The structure without soil the front, with column spacing at about 20 ft (6 m) and
improvement dropped the precast concrete purlins and roof stairways every 80 ft (24 m). Transverse walls separating
in the two bays with cast-in-place concrete shear walls. the units were spaced at 40 ft (12 m), and the length from
The other building did not suffer damage. It appears that front to back was from 66 ft to 82 ft (20 m and 25 m). The
the lateral bowing in the roof girders caused a failure at the rear was enclosed by vertical precast concrete walls with
pins and loss of bearing for the purlins. loading docks and doors. Figures 12 and 13 show photos
of the collapse.
Although many precast concrete buildings constructed
using shear walls have performed well in past earthquakes, Although there appeared to be an ample number of walls
the LFRS requires a complete load path that ties all compo- to provide lateral bracing for the structure, failures likely
nents together. In this case, it appears that the roof framing occurred because of inadequate connections and a lack of
with spaced purlins and without connections between roof seismic detailing for strength and ductility. Most of the
planks lacked a diaphragm. Failures occurred at the pinned walls that clad the exterior of the stairs on the front eleva-
purlin bearings where movement between the roof and the tion fell away from the structure because of out-of-plane
supporting girders was not sufficiently restrained. forces that caused the connections to fail. Failed gravity
columns showed bar bucking and a lack of confinement
Another combined office and warehouse consisting of a (Fig. 14).
precast concrete building with transverse and longitudinal
walls suffered major local failures and partial collapse. On the rear elevation, the tops of the walls were connected
The structure had two-level beam and column framing on to the spandrel girders at the roof through a thin angle that

Figure 16. Misaligned wall-girder connection that failed.
spanned between embedments in the walls and girders. bays. The framing included precast concrete girders in both
Figure 15 shows a remnant of bent and broken angle still the longitudinal and transverse directions, bearing on top of
welded to embedments in the back of the fallen spandrel precast concrete columns with wet-cast joints. Figure 17
panel. The welds to the plates embedded in the spandrel shows one interior joint. The roof sloped from the sides
girder were torn loose. Figure 16 shows that the continu- to the center girder line. The roof deck was supported be-
ous angle allowed the welds to be made even if the plates tween the girders with simply supported precast concrete
in the walls did not align with the plates in the spandrel joists. The roof plane, in lieu of a diaphragm, had diagonal
girder. The out-of-plane forces, however, caused bend- bracing below the roof joists that was connected between
ing in the angle and prying on the welds. There were also girders to plates and gussets. The diagonal bracing was
locations that appeared to have field-installed anchor rods also composed of precast concrete joists, similar to those
in the walls that projected into the cast-in-place concrete in Fig. 18. The only significant damage to this structure
topping over the roof. Walls had shallow breakout cones from the earthquake was that all of these diagonal braces
that appeared to correspond to short lengths of bent dowels disconnected at the plates at the girders and fell.
projecting from the edge of the topping. In addition, the
walls were relatively thin, about 6 in. (150 mm), and the The girders were formed as I-shaped sections with rect-
wall reinforcement comprised at least three sizes of mild angular end blocks, similar to bridge girders used in the
steel reinforcement no larger than no. 3 (10M) bars.
The structure may have had sufficient area of walls to sus-

tain the lateral forces from the earthquake, but the connec-
tions were insufficient in strength and ductility to support
them against either in-plane or out-of-plane forces. The
panel design deficiencies in thickness and in reinforcement
may also have contributed to out-of-plane failures.
Moment-resisting frames
There were examples of precast concrete column and beam

framing that had cast-in-place concrete closure joints that
created continuity and formed moment-resisting frames.
These systems generally performed well.
One industrial building that used long-span girders with

wet-cast joints was formed with 52 ft × 79 ft (16 m × 24 m) Figure 17. Wet-cast joint between girders and column in an industrial building
that performed well during the earthquake.

United States. This building was clad using vertically span-
ning precast concrete double-tee wall panels. The panels
were connected to the structure with long threaded rods
that projected across the open void formed by the I shape
and bolted through the webs. These connections provided
out-of-plane support, but they did not engage the cladding
as shear walls to provide any assistance to the LFRS.
The field-cast joints for the building were described as

having reinforcement projecting from the ends of the gird-
ers into the space over the tops of the columns with bar
laps and hooks that were engaged by the closure pour.
Other examples with similar framing and bracing were

found at the Weir Vulco plant. Figure 18 shows the precast
Figure 18. Diagonal precast concrete bracing for precast concrete girder roof
concrete diagonal framing in place. These buildings did
framing in an industrial building that did not experience structural damage dur-
ing the earthquake. not experience damage to the structural systems. Some
bracing for these buildings was provided by shear walls
that framed large door openings with drop-in walls over
the doors. Precast concrete column and beam framing
made continuous at the columns was also used for the con-
struction of 40 total–precast concrete schools and buildings
at five universities. No failures were reported in any of
these buildings.
The diaphragms and continuity of joints in these

systems were developed with cast-in-place concrete
topping for floors and wet joints. The floor framing
was constructed with precast concrete double-tees and
tapered end stems and flanges that formed the pocket for
the wet joint. Figure 19 shows a view of the underside
of this framing.
Cantilevered column systems
Figure 19. Total–precast concrete school framing system used in more than 40 There were several examples of industrial buildings with
schools that experienced no damage from the earthquake. precast concrete framing supported by cantilevered col-
umns as the LFRS.
One example in Santiago was a large exhibition hall. The

structure was framed with three consecutive bays, each
130 ft (40 m) long. It had eight 40-ft-wide (12 m) bays
spanned by spread precast concrete beams with a trap-
ezoidal section. Figure 20 shows an interior view of the
framing. The 130-ft girders are tapered I-beams that form
gable roofs. These beams bear on 40-ft columns and are
held with vertical rods that pin the ends to the bearings.
The columns are 35 in. (900 mm) square. Without moment
continuity between the ends of the beams or between the
beams and the columns, the lateral support for the structure
is provided by only the cantilever action of the columns at
the footings. The footings are not tied together with grade
beams, but the columns were designed for a combined
lateral force equal to 25% of the weight.
Figure 20. Interior of Espacio Riesco Exhibition Hall with cantilevered column
framing. The spread precast concrete beams do not form moment-
resisting frames, but they have wet-cast connections at

Figure 21. Interior view of the roof framing at a can factory. Figure 22. Base isolation bearings at Weir Vulco.
their bearings on the roof girders, so the secondary framing were fixed to the precast concrete columns with exterior
is made continuous. This detail adds some redundancy to plates bolted around the columns. The earthquake caused
the roof system and contributes to the overall structural the canopy to sag and the columns to crack near the beam
integrity. The spread-beam system, however, does not form connections. The cantilevered behavior of the columns
a continuous diaphragm capable of redistribution of the resulted in flexural cracks near the bottoms of the columns.
forces. The building suffered no structural damage to the The team also found that there were some local spalls at
primary LFRS. the bearing of a roof beam where the width of the beam
spanned across the joint between the ends of the roof gird-
This building was clad to about two-thirds of the exterior ers. The observed damage in these areas was not severe
wall height with horizontal precast concrete walls. These and was being repaired.
walls were not intended to act as shear walls, and some of
the upper panels fell from the structure as the cladding con- The design of cantilevered column systems was shown to
nections failed. This aspect is discussed in the section on be effective.
precast concrete cladding.
Advanced seismic-force-resisting
Another example of precast concrete framing with cantile- systems
vered columns was found in Coronel, south of Concepción,
at the Parque Industrial Escuadrón, adjacent to the failed There were several examples of framing systems using
gable frames described earlier. This recent construction advanced concepts that proved effective during the earth-
also used long-span gable-shaped roof girders and spread- quake.
beam framing to form a warehouse for the fish meal opera-
tion. The gable-shaped girders are pierced with round holes Base-isolated offices As a demonstration of lami-
in the webs to reduce their weight. Again, the trapezoidal nated base-isolation rubber bearings, the manufacturer con-
spread roof beams were made continuous across these gird- structed a total–precast concrete office building supported
ers with cast-in-place concrete joints. One section of this on slide bearings at the corners and on base isolators at the
building was about 52 ft (16 m) tall, with beams framing interior bays. Figure 22 shows a view of the isolation bear-
with pinned end connections at two levels above and below ings on one side.
the girders. This building survived the earthquake without
damage. The building is two bays by five bays, with a square
module of 26 ft (8 m). The structure is only two stories
A second example in Coronel was found at a manufactur- tall, but the company has supplied isolation bearings to
ing facility. The framing was similar to that of the fish meal fifteen other buildings with similar design. The structure
facility. Figure 21 shows an interior view of this building experienced no damage during the earthquake, though the
showing the gable-shaped girders and spread trapezoidal slide bearings showed movement of about 5 in. (130 mm)
roof beams. Field-cast joints created continuity in the roof in both longitudinal and transverse directions. The owner
beams across the girders. reported that books standing on end in the structure did not
fall over.
This structure was damaged in three areas. In one loca-
tion there was a long exterior cantilevered canopy over a It was reported that these bearings were used in some
loading area that was framed with steel beams. The beams residential buildings, some bridges, and at buildings at the

Reinforced concrete braced frames with
cable-stayed roof The main building in the conven-
tion center complex is a large conference hall constructed
with precast concrete braced frames. The frames are tilted
in from the side walls so that the clear span at the floor level
is 200 ft (61 m) but the girder span for the roof is reduced
to 160 ft (49 m). Figure 23 shows an interior view of this
framing.
The clear height under the roof girders is 40 ft (12 m), and
they are 5 ft (1.5 m) deep. Roof beams span between the
girders to support the roof decking. Because of the long
spans of the girders, there is additional support provided
by cable stays. To hold these stays above the roof, columns
were added above the top intersection of the tilted braced
Figure 23. Interior view of a convention hall with tilted braced frames.
frame columns; these added columns lean outward. The
joints tying the lower and upper columns together were
made with field-cast concrete. Figure 24 shows an exterior
view of the cable stays, braced frames, and columns.
At one end of the hall, bracing columns extend to the

edge girder. At the other end, a large room for staging and
service support is framed with seven sides and clad to
about half the wall height with horizontal stacked precast
concrete wall panels. These cladding panels were not used
as lateral bracing for the structure and were connected to
columns with erection angles between slotted inserts. The
slots are oriented horizontally in the walls and vertically
in the columns, apparently to allow compensation during
erection for casting tolerance between walls and interior
framing. The roof for this side room is supported with
tapered precast concrete girders that span the width of the
room and are supported by, and cantilever over, another
Figure 24. Exterior view of a convention hall with braced frame and cable-stay interior long-span girder. This girder spans 160 ft (49 m)
supports.
and is 8 ft 2 in. (2.5 m) deep.
Catholic University and the University of Chile. They were The only damage to this building from the earthquake
used in a dock at the port of Coronel, which was reported was from failure of the cladding panel connections. It was
to be the only dock not damaged by the earthquake. reported that the top panels pulled out of plane and fell off
the structure. Some of the panels shown in the photograph
Unbonded prestressed concrete frames and were replaced temporarily while waiting for new panels
walls A precast concrete manufacturer constructed a five- with additional connections to be fabricated.
story structure at its convention/exhibition site that uses
unbonded post-tensioned walls and frames following the Precast concrete stadiums
research of the PCI PRESSS (Precast Seismic Structural
Systems) program. The structure is braced in the short The PCI team investigated stadiums framed with precast
direction by post-tensioned shear walls placed at the ends concrete columns, beams, rakers, and risers in Chillán and
of the building. The post-tensioning strands are located in nearby Chillán Viejo.
near the center of the walls. In the other direction, there are
three bays framed with unbonded post-tensioned moment- The stadium in Chillán was framed with precast concrete
resisting frames. Although the erection of the structure was for the seating areas on four sides enclosing the playing
complete, the building was unfinished. The first floor was in field. The seating was shaded with a fabric roof within
operation as a kitchen for the convention center. The upper steel frames supported by cantilevered steel columns
floors were not yet completed. The structure experienced no attached to the tops of precast concrete columns on the
damage from the earthquake. perimeter; these were braced by the raker beams and
transverse beam framing. The project was planned on a
tight schedule, and when the precast concrete manufacturer

could not supply sufficient components, the framing of the
press box and supporting building was converted to cast-in-
place concrete, with a vertical line of separation at the back
of the seating area.
The precast concrete framing included rakers, columns,

beams, and single-step risers. Figure 25 shows a view of the
framing during erection. The precast concrete framing was
tied transverse to the rakers by beams with welded connec-
tions. With the exception of the failed roof structure falling
on the seating, the precast concrete framing withstood the
earthquake with only minor damage to bearing surfaces,
which showed some cracking. The primary failure occurred
at the connections at the tops of columns where the steel
framing for the fabric roof tore from column-base connec-
Figure 25. Stadium with precast concrete rakers on precast concrete columns.
tions and at some of the bracing cable anchor connections.
The failures occurred only at cast-in-place concrete columns
on the press box side of the stadium. buildings had lightweight non-diaphragm roofs and the
lateral forces were resisted by the column-beam framing
The stadium in Chillán Viejo was a smaller structure with system, sometimes with roof-level diagonal bracing, and a
precast concrete columns, rakers, and risers on opposite few cast-in-place concrete shear walls. The double-tees are
sides of the playing field. The framing was a simple single- commonly about 8 ft (2.4 m) wide and 30 ft (9.1 m) tall,
span raker on exterior columns. On one side, the seating with conventional (nonprestressed) reinforcing.
was backed up with a press box structure.
The team investigated two buildings with double-tee clad-
Damage to this stadium appeared to be relatively minor. ding. One used flat precast concrete panel cladding at the
There was a spall at the bearing of a riser stem at the top corners and double-tee walls along the sides. The other
of a raker where the bearing area was not confined, but used double-tee wall panel cladding that included walls
the concrete appeared to be intact over most of the bear- supported above a wide loading dock opening. In these two
ing length. There were also cracks at the bearing of a raker examples, the full-height double-tee wall panels were sup-
beam at a column corbel, but most of that bearing appeared ported on the foundation and with projecting reinforcing
intact. cast into the floor slab. Both the floor slab and an exterior
slab were cast against the base of the wall panel. Near the
It appeared that the precast concrete framing performed top of each double-tee leg, there were long bolt tiebacks
well during the earthquake. that projected through the interior perimeter beams
(Fig. 26). The bolts were more than 1 ft (0.3 m) long to
Performance provide out-of-plane restraint while allowing movement
of precast concrete cladding parallel to the wall of the framing system and the roof
and cladding connections system without transmitting force to the wall system.
In Chile, precast concrete cladding panels are used on At the convention center complex described previously, the
industrial buildings and on some low-rise office buildings. exhibition and convention halls were clad with long hori-
Precast concrete cladding was not observed on high-rise zontal precast concrete panels. The panels were stacked
structures. Most of the panels observed performed well. two to four units high, with the primary weight transferred
There were some cases where connections between the through the panel below and then to the foundation. One of
cladding and the supporting structure failed. the buildings had sloped precast concrete columns, creat-
ing a braced frame, so that part of the panel weight was
Most of the precast concrete cladding panels observed were carried by its connections to the columns (Fig. 23 and 24).
nonstructural. These panels were subject only to inertial The end annex to that building (Fig. 27) and the adjacent
seismic forces and wind loads. For effective support, the exhibit hall had vertically stacked panels. In neither case
connections of these panels should accommodate move- did the precast concrete walls reach the level of the roof.
ment of the supporting structure. Without this flexibility, Metal cladding was used to complete the enclosure.
cladding panels can attract unintended forces. Examples of
successful and unsuccessful performance were observed. The wall panels were attached to the concrete columns
with long slotted embedments (Fig. 28). The embedments
There were several warehouse-type buildings using vertical are commercial inserts commonly used for precast con-
double-tee wall panel cladding that performed well. These crete connections. The slotted inserts are oriented vertically

Figure 26. Double-tee panel connection to perimeter beams used in industrial buildings.
in the columns and horizontally in the wall panels to provide

ample alignment tolerance. The connection appeared to be
for out-of-plane forces only. Although the photos show the
precast concrete wall panels in place, some of those panels
fell away from the buildings during the earthquake and were
replaced. As designed and installed, these connections did
not have sufficient strength to withstand the earthquake
forces. The slotted insert embedment had deformed at the
lips and allowed the bolts to pull out. One large all–precast
concrete multioccupant structure that was virtually de-
stroyed was described earlier in this report. The building’s
nonstructural precast concrete cladding was damaged or
simply collapsed. Cladding on the building was damaged as
the supporting structure failed. Exterior panels fell away due
to connection failures at the roof level.
Figure 27. Annex of Preansa Convention Hall uses horizontal precast concrete
panels. Precast concrete cladding was used on the steel-framed
structure of a building supply warehouse store in Con-
cepción. Panel collapse at this structure appeared to be
caused by the failure of the supporting structure. It was not
possible to determine whether inertial forces from the clad-
ding contributed to the failure.
An office structure in Concepción clad with precast con-

crete panels appeared intact, though much of the infill glass
was broken. Precast concrete cladding was also used on the
base-isolated Weir Vulco building. The base isolation of that
building protected the cladding as well as the structural pre-
cast concrete framing. Other buildings with precast concrete
wall cladding were observed from a distance to have been
damaged, but with access limited, the configuration and the
extent of damage could not be determined.
In general, the precast concrete cladding panels in Chile

performed well when the effects of and requirements for
Figure 28. Slotted inserts to connect wall panels to concrete columns and the seismic resistance were considered in design and detailing.
deformation of the slotted inserts due to the earthquake load.

Figure 29. Bridge section showing recent construction practice that eliminates the transverse diaphragm and uses vertical steel rods to prevent overturning.
Precast concrete bridges
Many highway bridges in Chile are constructed with pre-

cast concrete I-girders or bulb-tee sections. Many of these
crossings performed well, but there were also many notable
failures that shared common characteristics.
Precast concrete
bridge construction
Moderate-span precast concrete bridges in Chile are con-

structed using I-girder or bulb-tee girder shapes, similar to
AASHTO sections used in the United States. The design
includes cast-in-place concrete diaphragms between gird-
ers. The girders bear in pockets with direct lateral restraints
against displacement at each girder. The diaphragms tie Figure 30. Failed bearing in the Llacolen Bridge.
the girders together laterally and prevent overturning from
lateral loads above the bearings. configuration are more susceptible to loss of bearing, the
geometry of the rigid concrete deck must contact the piers
Local engineers reported that design practice for these or abutments and cause some lateral displacement of the
bridges changed in the late 1990s following Spanish prac- support for the rotation to continue. This strongly suggests
tice so that diaphragms providing lateral support between that a lack of longitudinal restraint of these bridge girders
girders were reduced or eliminated. Lateral support at the at the abutments or piers at one end of the span at least
bearings was reduced to end stoppers at the ends of the contributed to these failures.
piers or abutments, constructed with small reinforced con-
crete sections projecting above the beam bearing surface. Bridge inspection
Vertical steel rods from the bearing to the underside of the
upper flange of the girders compensate for the loss of over- The PCI team inspected two bridges in Concepción. The
turning resistance. Figure 29 shows a section representing first bridge was the Llacolen Bridge, which includes sev-
these features of precast concrete bridge design. eral moderate-length spans across the Bio Bio River. This
bridge was constructed without diaphragms between the
The bridges with designs based on the more recent practice girders at the piers. The construction included the vertical
experienced more damage than bridges constructed with restraint bars enclosed in galvanized steel tubes. Figure 30
traditional details. The more recent bridges suffered lateral shows the bearing surface with concrete debris and twisted
displacements at the bearings, failure in end stoppers, and reinforcing from the failed girders and deck. The photo
some overturning of the beams. also shows the bent and twisted galvanized tubes that held
the failed restraint bars.
Bridges with skewed bearings failed because a lack of lat-
eral restraint permitted the global rotation of the bridge and As seen in the photo, the bridge girders pulled away from
allowed the beams to fall from the bearings. The absence the bearing and dropped the end of the span. This was
of lateral restraints at the bearings is not sufficient to pro- not a skewed span, but the span lacked both lateral and
vide a mechanism for collapse. Although bridges with this longitudinal restraint at this bearing. The length of the

Diaphragm
Figure 31. Bridge section at the Bio Bio River crossing with partial diaphragms, vertical restraint bars, and end stoppers.
bearing was not sufficient for the longitudinal displacement The abutment bearing was cast as a level ledge, and then
between the piers. This was one of several failed multispan the bearing blocks with varying thicknesses were placed
bridges that dropped the ends of spans. to form the final bearing elevations. This convenient
method used to construct the correct bearing elevations
Figure 30 also shows an elevated bearing surface that is lacks the lateral restraint of bearing details using pockets.
a reminder that actual construction geometry is often not The earthquake forces caused lateral displacement and the
as simple as the illustration of the geometry shown in the failure of the stopper at one end of the pier. The failure did
earlier figures. not indicate a lack of reinforcing, though the horizontal
confinement did not appear to contain all of the vertical
Figure 31 shows another bridge section that includes reinforcement. The design, however, imposed all the re-
the center crown and cross slope for drainage that must quirement of lateral restraint on the end stopper. Figure 33
often be included. It is common for the beam bearings shows the failed stopper at the abutment. This view also
to be stepped to provide the crowned shape or to provide shows a wide crack in the bottom of the edge girder from
for cross slope or superelevation in the roadway above. the impact with the stopper.
These features of practical geometry can result in the loss
of lateral restraint to the girders; this was observed in the A bridge in Santiago designed with vertical restraint bars
Llacolen Bridge. The second bridge was a multispan bridge attached to precast concrete girders and bridge pier showed
crossing an inlet of the Bio Bio River. no evidence of displacement. Damage at this bridge was
seen in subsidence of the fill that formed the bridge ap-
This bridge had some of the characteristics of more recent proach and embankment at the grade separation.
bridge construction details but did include partial-depth
diaphragms between the girders at the bearings. Figure 31 Building code
also includes the diaphragms, the vertical restraint bars,
and the end stoppers. Figure 32 shows a view of the bear- Buildings and other structures in Chile must be designed
ing at two interior girders. and constructed in compliance with the Chilean Building
Code NCh433-2009.6 This code is applied in addition to
the specific design code for each of the materials and aims
to achieve structures that meet the following objectives:
• resist moderate intensity of seismic actions without

damage
• limit damage to nonstructural elements during earth-

quakes of moderate intensity
• prevent collapse during earthquakes of severe inten-

sity, even though they show some damage
Compliance with the provisions of this code does not guar-

antee that the objectives will be achieved.
In particular, the provisions for reinforced-concrete-wall

Figure 32. Lateral displacement due to earthquake forces of the Bio Bio River buildings are based on their satisfactory behavior dur-
bridge with partial diaphragms and vertical restraint bars.

Table 2. Soil parameters for base shear calculation in NCh433-2009
Soil type S T', sec n
I 0.9 0.20 1.00
II 1.0 0.35 1.33
III 1.2 0.85 1.80
IV 1.3 1.35 1.80
Note: n = soil parameter for base shear calculation indicated in
Where is the referencef for these tables?

NCh433-2009; S = site coefficient; T' = soil parameter for base shear
calculation indicated in NCh433-2009.
Table 3. Maximum values of seismic coefficient C in NCh433-2009
R Cmax
2 0.90SA0 /g
3 0.60SA0 /g
4 0.55SA0 /g
5.5 0.40SA0 /g
6 0.35SA0 /g
7 0.35SA0 /g
Note: A0 = maximum effective acceleration value; Cmax = maximum

seismic coefficient; g = acceleration due to Earth’s gravity; R = reduc-
Figure 33. Failed stopper at the abutment of the Bio Bio River bridge, with a
crack in the bottom of the girder. tion factor ranging from 2 for structural systems of limited ductility to 7
for ductile structural systems; S = site coefficient.
ing the earthquake of March 1985. Those buildings were

designed in accordance with NCh433-1972.18
type I is rock; soil type IV is soft soil.
The Chilean seismic code NCh433-199611 was in effect
until it was replaced by the 2009 update (NCh433-2009)6 The base shear Qo is determined from Eq. (1).
shortly before the earthquake. A 2010 update has been
developed in direct response to the earthquake. Qo = CIP (1)
NCh433-20096 has four building categories: A, B, C, The seismic coefficient C is obtained from Eq. (2).
and D. These are comparable to ASCE 7/05 occupancy
2.75A0 a T lk
categories IV (essential facilities, hazardous facilities), III n
(high-occupancy buildings where many people congregate C= (2)

in one space at one time), II (standard-occupancy buildings gR a T * kn
such as office buildings and apartments), and I (miscella-
neous-occupancy buildings where no life safety is at stake),
respectively. The importance factor I for building types A, In no case shall the value of C be less than A0/6g.
B, C, and D is 1.2, 1.2, 1.0, and 0.6, respectively.
The value of C need not be greater than that indicated in
The country is divided into three seismic zones: zone 1 is Table 3.
along the foothills of the Andes, zone 3 is along the Pacific
coast, and zone 2 is between zones 1 and 3. The maximum There are no specific provisions for or prohibition of build-
effective acceleration values A0 corresponding to zones 1, ings with structural irregularities.
2, and 3 are 0.20g, 0.30g, and 0.40g, respectively.
ACI 318-95 seismic provisions were referenced by
NCh433-1996 considers four soil types: I, II, III, and IV NCh433-199611 in annex B, except the confinement re-
(Table 2). These are comparable to soil profile types S1, S2, quirements for wall boundaries were specifically exempted
S3, and S4, respectively, in UBC editions before 1997. Soil in clause B.2.2 as previously discussed. This clause has

Figure 34. Longitudinal and transverse reinforcing bar diameter limitations. Note: bw = thickness of wall; db = nominal diameter of longitudinal bar; dbt = nominal
diameter of transverse bar; fy = yield strength of longitudinal reinforcement; fyt = yield strength of transverse reinforcement.
been deleted from NCh433-2009.6 • Studies should be nonlinear-analytical and experi-

mental. Analytical results should be compared with
The following changes19 to NCh433-20096 were being observations and new findings that confirm or reject
considered and were in draft form at the time of the team’s the proposed explanations.
visit:
• A building responds to an earthquake with the struc-
• Limit axial forces on columns and walls that are sub- ture as constructed and with material properties that
ject to lateral displacements to 0.35 fclAg . exist at the time of the earthquake. This may be obvi-
ous but is often ignored.
• All hooks on hoops and cross ties of confinement rein-
forcement should be 135° rather than 90°. At a presentation before ACI 318 subcommittee H in Oc-
tober 2010, Bonelli discussed the following changes made
• Confine at least 0.15ℓw from each edge and laterally to ACI 318-0816 requirements in the emergency changes to
support every vertical bar, not just every other one. NCh433-20096,21 and NCh430-2008:22,23
• Apply capacity design concepts to determine axial 1. The whole flange width of a flanged section (T, L, C,
force and shear, considering the effect of connection or other cross-sectional shapes) must be considered in
with the slabs. calculating combined flexural and axial load strength.
• Add slenderness restrictions to avoid transverse bend- 2. The contribution of the total amount of longitudinal
ing of boundary elements and the panel. For this, study reinforcement must be considered in determining
the New Zealand19 and Canadian20 code recommenda- combined flexural and axial load strength.
tions.
3. Longitudinal reinforcing bar diameter must be less
• Revise the displacement spectrum. than or equal to one-ninth of the least dimension of
the boundary element (wall thickness) (Fig. 34).
• Avoid adding to the resistance side; add to the demand
side. Study the effect of resistance (strength) on dis- 4. Transverse reinforcing bar diameter must be greater
placement demand. than or equal to one-third of the diameter of the re-
strained longitudinal bar (Fig. 34).
• Optional displacement-based design has been intro-
duced in annex B. 5. Transverse reinforcement must be anchored to ex-
treme longitudinal bars in a wall.
Patricio Bonelli, a professor at Universidad Técnica Fed-
erico Santa María, also shared the following recommenda- 6. Standard hooks must be used with transverse rein-
tions and reflections: forcement as defined in section 7.1 of ACI 318-08:
135° or 180° bend plus 6db extension, but not less than
• Carefully study the demands of displacement and 3 in. (75 mm) at the free end of the bar. In ACI 318-
rotation. In Concepción, 15 cycles of 0.2g acceleration 08,16 a standard hook is defined as a 180° bend plus
were observed at a period of 1.5 sec. 4db extension or a 90° bend plus 12db extension. Thus,
these two requirements are contradictory. The second
• Microzonation of cities is desirable. requirement (135° or 180° bend plus 6db extension),

Figure 35. Transverse reinforcement through the length of lap splice. Note: Ab = area of an individual longitudinal bar; A b‘ = area of an individual transverse bar;
fy = yield strength of longitudinal reinforcement; fyt = yield strength of transverse reinforcement.
being more restrictive than the first, is assumed to 11. Transverse reinforcement through the length of lap
govern. splices of longitudinal bars in walls must satisfy the
condition in Fig. 35.
7. In special structural walls, the net tensile strain in the
extreme tension steel εt must be equal to or greater 12. In section 21.9.6.2 of ACI 318-0816 on boundary ele-
than 0.004 when the concrete in compression reaches ments of special structural walls, the lower-bound limi-
its assumed strain limit of 0.003. (Section 10.3.5 of tation of 0.007 on δu /hw in Eq. (21-8) shall not apply.
ACI 318-08 requires εt at nominal strength to be not
less than 0.004 for nonprestressed flexural members 13. Section 21.9.6.4(a) of ACI 318-0816 may be replaced
and nonprestressed members with factored axial com- by the following:
pressive load less than 0.10 fclAg .)
The boundary element shall extend horizontally from
8. The transverse dimension (thickness) of special struc- the extreme compression fiber a distance not less than
tural walls must be greater than or equal to one-six- cc determined from Eq. (3).
teenth of the lateral unsupported member length under
cc c 1
compression, ℓu /16. = - (3)
,w ,w dul
600
9. In special structural walls, when shear has not been hlw
calculated using capacity design rules, the maximum
shear obtained from design load combinations that When this option is applied, the term δu/hw in ACI 318-
include load effects of earthquake E shall be calculated 0816 section 21.9.6.2 must be replaced by d'u /h 'w . The
with E assumed to be 1.4 times that prescribed by the lateral design displacement δu comes from NCh433-
legally adopted general building code for earthquake- 199611 (modified in 20096) section 5.9.5.
resistant design. ACI 318-08 section 21.3.3(b) lists the
comparable factor as 2, which is applicable to beams 14. Replace ACI 318-0816 section 21.9.6.2(b) with “The
and columns of intermediate moment frames. boundary element reinforcement shall extend vertical-
ly from the critical section a distance not less than ℓu.”
10. Transverse reinforcement through the length of lap
splices of longitudinal bars in walls must satisfy items 4, 15. Transverse reinforcement in boundary elements in
6, and 11 of this list when either (a) or (b) is true. walls, when required, must satisfy ACI 318-08 section
21.9.6.4 and (a) and (b).
a) T
he longitudinal reinforcement ratio, defined as
ΣAb/hs, is greater than 2.8/fy.
b) T
he cover of a longitudinal bar with nominal diam-
eter db is less than 2db.

a) T
he spacing of cross ties or legs of rectilinear resistance without some of the constraints imposed on U.S.
hoops hx within a boundary element in a wall shall practice by restrictive building code provisions.
not exceed the smaller of 8 in. (200 mm) and the
least dimension of the boundary element. In general, precast concrete cladding panels performed
well in Chile as long as the effects of and requirements for
b) T
he spacing of transverse reinforcement in a bound- seismic resistance were considered in design and detailing.
ary element in a wall shall not exceed the smaller of
six times the diameter of the smallest longitudinal bar The bridge infrastructure in Chile experienced large-
and half the minimum boundary element dimension. magnitude shaking, often larger than what the bridges
were designed for, with varying degrees of damage. With
16. Replace ACI 318-0816 section 21.9.6.5(a) with the fol- the exception of the more recently constructed bridges in
lowing: which diaphragms were reduced or eliminated, the precast
concrete bridges observed by the PCI team generally per-
a) W
here the longitudinal reinforcement ratio at formed well. In some cases, the girders performed well but
the wall boundary is greater than 2.8/fy, bound- the absence or weakness of diaphragms or lateral restraint
ary transverse reinforcement shall satisfy sections resulted in failures at bearings or piers. Failures at bearings
21.6.4.2 and 21.9.6.4(a) or item 13 of this list. The and piers and the rotation of spans causing loss of bear-
maximum longitudinal spacing of transverse rein- ing were not confined to bridges constructed with precast
forcement in the boundary shall not exceed 8 in. concrete girders but were also seen in steel-girder bridges
(200 mm). constructed with similar end details.
b) W
here vertical reinforcement can yield, the maxi- The 2010 emergency changes to Chile's building code have
mum longitudinal spacing of transverse reinforce- far-reaching implications for the special structural wall de-
ment in the boundary shall not exceed the smaller sign provisions in ACI 318-0816 section 21.9. Changes will
of six times the longitudinal bar diameter and 8 in. be considered for possible inclusion in the next version of
(200 mm). ACI 318 and, if adopted, may have significant effects on
design practice in the United States.
Conclusion
Acknowledgments
The majority of structures performed acceptably or better,
considering the severity of the 2010 Chile earthquake. The A version of this report is under review for inclusion in a
number of deaths and the amount of property loss, while special edition of Earthquake Spectra to be published by
quite significant, were not disproportionate to the severity the Earthquake Engineering Research Institute (EERI) in
of the earthquake. Much of this is attributable to Chile’s June 2012. The guest editors are Jack P. Moehle and Da-
history of adoption and implementation of adequate build- vid Frost. This report appears here by permission of EERI.
ing codes.
The PCI team consisted of Ned Cleland of Blue Ridge
The PCI team concentrated on precast concrete structures. Design in Winchester, Va.; Susan Dowty of S. K. Ghosh
With the exception of the out-of-date gable frame system Associates Inc. in Aliso Viejo, Calif.; S. K. Ghosh of S. K.
observed in one location south of Concepción, the pre- Ghosh Associates Inc. in Palatine, Ill. (team leader); Ray
cast concrete building systems generally performed well. McCann a structural engineer in Napa, Calif.; and Dante
In some cases, the LFRS performed, but the absence or Sanguineti of Pomeroy Corp, in Perris, Calif. Augusto
weakness of diaphragm framing resulted in local failures. Holmberg of Instituto del Cemento y del Hormigón de
Where lateral forces were resisted by cantilevered columns Chile was the PCI team’s local contact and host. Patricio
and distribution of loads through diaphragm action was not Bonelli, a professor at Universidad Técnica Federico Santa
essential, the structural framing of the buildings did not María, provided much valuable information concerning
experience significant damage. Chile's building code and changes to that document. He
also led three of the team members on a tour of Viña del
Some of the buildings inspected showed the success of Mar. His help is gratefully acknowledged. The contribution
advanced precast concrete seismic systems, which reflects of the section on cladding by Ray McCann and review of
research conducted in the United States. Some used tech- the original version of this report by all team members are
nology associated with other materials on the U.S. market. much appreciated. The authors are grateful to Prabuddha
The example of the reinforced concrete braced frames Dasgupta of S. K. Ghosh Associates Inc., whose help in
showed the success of a system that is not included in the the preparation of the manuscript was invaluable.
defined systems in ASCE 7-10.8 The PCI team found a
mature and sophisticated precast concrete industry that has
successfully considered and solved problems of earthquake

References 14. Wood, S. L. 1991. Performance of Reinforced Con-
crete Buildings during the 1985 Chile Earthquake:
1. U.S. Geological Survey. Magnitude 8.8—Offshore Implications for the Design of Structural Walls. Earth-
Bio-Bio, Chile. earthquake.usgs.gov/earthquakes/eqin- quake Spectra, V. 7, No. 4: pp. 607–638.
thenews/2010/us2010tfan/#summary.
15. Wallace, J. W., and J. P. Moehle. 1992. Ductility and
2. Ruegg, J. C., A. Rudloff, C. Vigny, R. Madariaga, J. B. Detailing Requirements of Bearing Wall Buildings.
de Chabaliera, J. Camposc, E. Kauselc, S. Barrientosc, Journal of Structural Engineering, V. 118, No. 6: pp.
and D. Dimitrovd. 2009. Interseismic Strain Accumu- 1625–1644.
lation Measured by GPS in the Seismic Gap between
Constitución and Concepción in Chile. Physics of the 16. ACI Committee 318. 2008. Building Code Require-
Earth and Planetary Interiors, V. 175: pp. 78–85. ments for Structural Concrete (ACI 318-08) and Com-
mentary (ACI 318R-08). Farmington Hills, MI: ACI.
3. Earthquake Engineering Research Institute (EERI).
2010. EERI Special Earthquake Report—June 2010. 17. Wallace, J. W. 2010. The 27 February 2010 Chile
EERI Newsletter Suppl. Earthquake: Implications for U.S. Building Codes and
Standards. Meeting summary, The 27 February 2010
4. RENADIC. Red de Cobertura Nacional de Acceleró- Chile Earthquake: Implications for U.S. Building
grafos, Departamento de Ingeniería Civil, Universidad Codes and Standards, organized by American Society
de Chile, Santiago, Chile. www.renadic.cl of Civil Engineers (ASCE), National Institute of Stan-
dards and Technology (NIST), and Pacific Earthquake
5. Boroschek, R., P. Soto, and R. Leon. 2010. Maule Engineering Research (PEER) Center, San Francisco,
Region Earthquake—February 27, 2010 – Mw = 8.8. Cal., June.
RENADIC report 10/08 Rev. 1. Civil Engineering
Department, Faculty of Mathematics and Physical Sci- 18. INN. 1972. Diseño Sísmico de Edificios [Earthquake-
ences, University of Chile, Santiago, Chile. resistant design of buildings]. Official Chilean stan-
dard NCh 433. 1972. Santiago, Chile: INN.
6. Instituto Nacional de Normalización (INN). 2009. Dis-
eño Sísmico de Edificios [Earthquake-resistant design 19. Standards New Zealand. 1997. Concrete Construction.
of buildings]. Official Chilean standard NCh 433.Of NZS 3109.97. Including Amendments A1 and A2.
1996, Modified in 2009. 2nd ed. Santiago, Chile: INN. Wellington, New Zealand: Standards New Zealand.
7. EERI. EERI Chile Earthquake Clearinghouse. www 20. Canadian Standards Association (CSA). 2004 (reaf-
.eqclearinghouse.org/20100227-chile/. firmed 2010). Design of Concrete Structures. CAN/
CSA-A23.3-04 (R2010). Toronto, ON, Canada: CSA.
8. American Socity of Civil Engineers (ASCE) 7. 2010.
Minimum Design Loads for Buildings and Other 21. 2011. Decreto numero 117—Aproeba Reglamento que
Structures. ASCE 7-10. Reston, VA: ASCE. fija el Diseño Sísmico de Edificios [Decree number
117—approves regulation which establishes seismic
9. ASCE 41. 2006. Seismic Rehabilitation of Existing design of buildings]. In Diario Oficial de la República
Buildings. ASCE 41-06. Reston, VA: ASCE. de Chile [Official diary of the Republic of Chile].
Santiago, Chile.
10. Los Angeles Tall Buildings Structural Design Council.
www.tallbuildings.org. 22. INN. 2008. Hormigón Armado—Requisitos de Diseño
y Cálculo [Reinforced concrete—requirements for de-
11. INN. 1996. Diseño Sísmico de Edificios [Earthquake- sign and calculation]. Official Chilean Standard NCh
resistant design of buildings]. Official Chilean stan- 430. 2008. Santiago, Chile: INN.
dard NCh 433. 1996. Santiago, Chile: INN.
23. 2011. Decreto numero 118—Aproeba Reglamento
12. International Code Council. 1997. 1997 Uniform que fija los requisites de Diseño y Cálculo para el
Building Code. International Conference of Building Hormigón Armado [Decree number 118—approves
Officials: Whittier, CA. regulation which establishes the requirements for de-
sign and calculation of reinforced concrete]. In Diario
13. American Concrete Institute (ACI) Committee 318. Oficial de la República de Chile [Official diary of the
2005. Building Code Requirements for Structural Con- Republic of Chile]. Santiago, Chile.
crete (ACI 318-05) and Commentary (ACI 318R-05).
Farmington Hills, MI: ACI.

Notation P = total weight of the building above the base level
A0 = maximum effective acceleration Qo = base shear
Ab = area of an individual longitudinal bar R = reduction factor ranging between 2 for structural
systems of limited ductility and 7 for ductile struc-
A 'b = area of an individual transverse bar tural systems in NCh433
Ag = gross area of concrete section s = center-to-center spacing of longitudinal reinforcement
bw = thickness of wall S = site coefficient
c = largest neutral axis depth calculated for the factored Sa = spectral acceleration
axial force and nominal moment strength consistent
with δu Sd = spectral displacement
cc = extension of the confined area measured from the T = time

extreme compression fiber
T ' = soil parameter for base shear calculation indicated in
C = seismic coefficient NCh433-2009
Cmax = maximum value of seismic coefficient T* = period of mode with the highest translational
equivalent mass in the direction of analysis
db = nominal diameter of longitudinal bar
β = damping coefficient
dbt = nominal diameter of transverse bar
δu = lateral design displacement
E = effects of earthquake
d'u = design drift measured between the top and the con-
fcl = compressive strength of concrete sidered level
fy = yield strength of longitudinal reinforcement εt = net tensile strain in the extreme tension steel
fyt = yield strength of transverse reinforcement
g = acceleration due to Earth’s gravity
hw = height of entire wall from base to top or height of

the segment of wall considered
h 'w = height of wall between top and considered level
hx = spacing of cross ties or legs of rectilinear hoops
I = importance factor for building categories, as speci-

fied in NCh433
ℓu = lateral unsupported member length under compres-

sion
ℓw = length of wall
M = moment magnitude of earthquake
n = soil parameter for base shear calculation indicated in

NCh433-2009 (Table 3)

About the authors or better, considering the severity of the earthquake.
Much of this is attributable to Chile’s history of adop-
S. K. Ghosh, PhD, is president of tion and implementation of adequate building codes.
S. K. Ghosh Associates Inc., in The success of advanced precast concrete structural
Palatine, Ill., and Aliso Viejo, systems, based on research in the United States, was
Calif. He has influenced seismic demonstrated. The Chilean precast concrete indus-
design provisions in the United try is mature, sophisticated, and, in the absence of
States for many years by author- constraints imposed by restrictive building codes, quite
ing many publications and by innovative. Code implications of lessons from the Feb-
serving on or chairing numerous committees and ruary 2010 earthquake are also discussed.
advisory panels. He is PCI’s code consultant.
Ned M. Cleland, PhD, P.E., is Keywords

president of Blue Ridge Design
Inc. in Winchester, Va. He has Bridge, cladding panels, code, construction, earth-
over three decades of experience quake, lateral resistance, seismic, structural system,
in the analysis and design of structural wall.
precast/prestressed concrete
structures. He chaired the PCI Reader comments
Seismic Committee for 14 years and is the incoming
chair of the Technical Activities Council. Please address any reader comments to journal@pci
.org or Precast/Prestressed Concrete Institute, c/o PCI
Journal, 200 W. Adams St., Suite 2100, Chicago, IL
Abstract 60606. J
This paper reports on the findings and observations

of the team sent by PCI to investigate damage caused
by the February 2010 earthquake in Chile. The paper
concentrates on the performance of precast concrete
structures, although some other aspects are also includ-
ed. The majority of structures performed acceptably

Effects of loading conditions
on the behavior
of unbonded post-tensioning
strand-anchorage systems
N
ew types of seismic-resistant building and bridge
structural systems (for example, shear wall, mo-
ment-resisting frame, bridge pier) using unbonded
post-tensioning tendons have emerged since the 1990s.1–5
The post-tensioning tendon anchorages in these structures
are among the most important components because the
entire post-tensioning force is transferred at these loca-
tions. While unbonded post-tensioning has been available
in the United States since the 1950s (for example, concrete
slab systems) and standards exist for its application,6–8 the
■ This paper presents an experimental investigation of unbonded use of this construction method in seismic zones is more
post-tensioning strand-anchorage systems subjected to a recent. Unbonded tendons result in a uniform distribu-
variety of loading conditions. tion of the strand elongations over the unbonded length,
thus reducing the maximum strand strains as the structure
■ The research focuses on the strand stresses and strains at wire deflects laterally during an earthquake. However, depend-
fracture inside the anchors, considering the following loading ing on the geometry and type of the structure, the intensity
parameters: loading rate, eccentricity between the strand ends, of the earthquake, and the initial strain (that is, prestrain)
postyield cyclic loading, and initial stress. level, maximum strains significantly greater than those
under gravity loading can develop in the strands during
■ In addition to the strand-anchorage experiments, updated seismic loading.
stress-strain relationships are proposed for the PCI Design
Handbook: Precast and Prestressed Concrete for 0.5-in.- For example, consider the hybrid precast concrete shear
diameter (13 mm) and 0.6-in.-diameter (15 mm) strands based wall described in Smith et al.4 Despite the placement of the
on experiments that achieve free-length fracture of the test post-tensioning tendons near the wall centerline to mini-
specimens. mize the strand elongations, the expected maximum strand

strain in the structure at the validation-level9 roof drift
of 2.3% is about 0.01. During the reversed-cyclic lateral
load testing of another type of unbonded post-tensioned
structure (a precast concrete coupling beam subassembly),5
brittle fracture of individual strand wires inside the anchors
(Fig. 1) was observed at strains less than 0.01, limiting the
seismic performance of the structure.
A comprehensive evaluation of unbonded post-tensioning

strand-anchorage systems is needed before they can be
used in extreme seismic applications. Walsh and Kurama10
focused on the behavior of monostrand-anchorage sys-
tems under monotonic concentric tensile loading. These
experiments showed a number of strand wire fractures at
strains as small as 0.01 (or slightly less for a small number
Figure 1. Individual wires inside the anchors of unbonded post-tensioning
of samples), which is significantly less than the mini-
strands can fracture under extreme seismic loading.
mum fracture strain of 0.02 required by the International
Code Council Evaluation Service (ICC-ES) Acceptance
Criteria for Post-tensioning Anchorages and Couplers of the strand-anchorage system to perform differently at
Prestressed Concrete.11 This paper builds on Walsh and the live end than at the dead end.
Kurama’s previous study10 by providing experimental
information on the strand wire fracture stresses and strains Six different one-time-use monostrand anchor configura-
from a representative sampling of strand-anchorage con- tions and two multistrand anchor types are investigated in
figurations considering the following loading parameters to this paper. The primary objectives are as follows:
better simulate field conditions in seismic regions:
• investigate the effects of the previously listed loading
• Loading rate: During an earthquake, the strands are parameters on the ultimate fracture stress and strain
subjected to varying loading rates. The loading rate capacity of a post-tensioning strand inside an anchor
may affect the fracture strain of a strand inside an
anchor. • make recommendations for the acceptance testing of
post-tensioning strand-anchorage systems for use in
• Strand end eccentricity: Eccentricity between the ends seismic applications
of a strand can occur due to reasons such as accidental
misalignment during construction or relative displace- • make recommendations for the design of unbonded
ments of the strand ends as the structure undergoes post-tensioned structures in seismic regions
lateral motion. The presence of strand flares inside the
trumpet at the end of a multistrand anchorage, where • investigate the stress-strain relationship of seven-wire,
the individual strands in a multistrand tendon spread uncoated, low-relaxation 270 ksi (1860 MPa) post-
out from the post-tensioning duct into the wedge tensioning strand through free-length fracture
holes of the anchor, can also cause the strand ends to experiments
be loaded eccentrically. These conditions may result
in earlier fracture of the strand wires compared with More information on the research can be found in the full
strands loaded concentrically. project report.16 Ultimately, the performance data measured
in this research should be compared with the maximum
• Postyield cyclic loading: Design guidelines12–14 for un- strain demands expected in unbonded post-tensioning
bonded post-tensioned structures in seismic regions re- strands under extreme seismic loading conditions.
quire that the strands remain essentially linearly elastic
during a “design basis earthquake”; however, postyield Research program
cyclic tensile stresses and strains can develop in the
strands during a “maximum considered earthquake.”4,15 Testing requirements
This may reduce the strand ductility by causing earlier
wire fractures inside the anchorages. The control test setup reported in this paper conforms to
the static testing requirements of the ICC-ES.11 Accord-
• Initial strand stress: In practice, a hydraulic jack is ing to these requirements, tests should be conducted on a
used to prestress the strand and power seat the anchor minimum of three samples for each strand-anchor configu-
wedges at the stressing (“live”) end. This may affect ration. The minimum free length of the test strand between
how the wedges are seated around the strand causing anchors should be 3.5 ft (1.1 m) and the strand must be

TW
L TOD IW
TOD TID Notch
TID H for ring
H H
BW
Boss
BID BID
BOD BOD
Two piece Three piece
TID Anchor wedges

W
H
TOD dmw
L dp
BID
dow
Cast anchor Barrel anchor
Nine-hole multistrand anchor Seven-wire strand
Figure 2. Materials tested in this research. Note: Diagrams are not drawn to scale. BID = anchor bottom inside diameter; BOD = anchor bottom outside diameter
(equal to TOD in case of barrel anchors); BW = wedge bottom outside width; dmw = diameter of middle strand wire; dow = diameter of outer strand wire; dp = outside
crown-to-crown diameter of strand; H = anchor or wedge height; IW = wedge inside width; L = cast anchor length; TID = anchor top inside diameter; TOD = anchor
top outside diameter (equal to BOD in case of barrel anchors); TW = wedge top outside width; W = cast anchor width.
tested with a monotonic machine position displacement rate • anchor wedges (two-piece or three-piece wedges with
between 0.197 in./min (5.00 mm/min) and 0.887 in./min or without binding rings)
(22.5 mm/min), which corresponds to a strand strain rate of
0.0047/min to 0.021/min for the minimum strand free length • ASTM A41617 seven-wire, uncoated, low-relaxation
of 3.5 ft required by ICC-ES. An extensometer with a mini- post-tensioning strand (0.5 in. [13 mm] or 0.6 in.
mum gauge length of 36 in. (910 mm) must be used within [15 mm] nominal diameters) with a nominal ultimate
the free-length of the strand to measure the strains. strength fpu of 270 ksi (1860 MPa)
The ICC-ES requirements also serve as the basis for the An important factor in the selection of each anchor con-
loading parameter tests in this paper, though most of these figuration for testing was the use of wedges and anchors
other procedures exceed the requirements of ICC-ES in that were compatible with each other and were provided by
vigor. The ICC-ES cyclic loading requirements, for exam- the same manufacturer.
ple, are provided for fatigue load tests. As described later
in the paper, the ICC-ES 50-cycle fatigue load upper stress Table 1 lists the geometric properties of the anchor speci-
limit of 0.85fpu (where fpu is the nominal strand strength), mens used in the loading parameter tests, named accord-
which roughly corresponds to a strand strain of 0.008, is ing to the anchor type and the nominal diameter of the
not high enough to observe the postyield cyclic behavior strand used with the anchor. Four monostrand cast anchors,
that may occur in strands under extreme seismic loading. two monostrand barrel anchors, one machined nine-hole
multistrand anchor (multi-0.6A), and one cast seven-hole
Test specimen properties multistrand anchor (multi-0.6B) representing three differ-
ent manufacturers were included in the investigation. As
The materials tested in this research (Fig. 2) consist of the described in the section titled “Multistrand Anchorages,”
following: the multistrand anchors were tested using only one hole
per test, as if they were monostrand anchors.
• monostrand cast anchors
Table 2 lists the tested wedges, which are named ac-
• single-use monostrand barrel anchors cording to the number of wedge pieces (2P for two-piece
wedges or 3P for three-piece wedges) and the nominal
• multistrand anchors (shown here as a nine-hole cylin- strand diameter (0.5 in. [13 mm] or 0.6 in. [15 mm]).
drical anchor but also tested with a seven-hole anchor)

Table 1. Anchor properties
Anchor H, in. L, in. W, in. TOD, in. TID, in. BID, in. BOD, in.
Cast-0.5B 1.48 4.96 2.19 1.67 1.00 0.66 1.34
Cast-0.5C 1.31 5.01 2.26 1.69 0.96 0.64 1.38
Cast-0.5D 1.53 5.00 2.23 1.67 1.00 0.67 1.27
Cast-0.6A 1.65 5.86 3.03 2.05 1.12 0.72 1.32
Barrel-0.5C 1.44 n.a. n.a. 1.62 1.01 0.65 1.62
Barrel-0.6A 1.76 n.a. n.a. 1.87 1.14 0.74 1.87
Multi-0.6A 2.20 n.a. n.a. n.a. 1.09 0.70 n.a.
Multi-0.6B 2.14 n.a. n.a. n.a. 1.12 0.67 n.a.
Note: BID = bottom inside diameter; BOD = bottom outside diameter; H = height; L = length; n.a. = not applicable; TID = top inside diameter; TOD =
top outside diameter; W = width. 1 in. = 25.4 mm.
Table 2. Wedge properties

Wedge Ring H, in. TW, in. BW, in. IW, in.
2P-0.5C No 1.17 1.00 0.74 0.48
2P-0.5E No 1.21 0.99 0.73 0.47
2P-0.5F No 1.30 1.00 0.69 0.47
3P-0.5B Yes 1.28 0.82 0.58 0.37
2P-0.6A No 1.61 1.12 0.79 0.57
2P-0.6C Yes 1.78 1.12 0.73 0.57
3P-0.6B Yes 1.60 0.94 0.66 0.46
3P-0.6C Yes 1.66 0.94 0.60 0.46
Note: BW = bottom outside width; H = height; IW = inside width; TW = top outside width. 1 in. = 25.4 mm.
The properties provided in the tables are average measure- and outside crown-to-crown diameter of strand dp—are all
ments taken from a minimum of three samples. Variations within ASTM A416 tolerance limits. The cross-sectional
exist in the average dimensions of both the anchors and area ap of each strand was determined from measured
the wedges depending on the manufacturer and the model. dimensions and weights of the strand samples10,16 (that is,
Binding rings hold the wedge pieces together, particularly measured strand areas rather than nominal areas were used
during seating. The wedges without rings (always two- in the research).
piece wedges in this sample set) do not have the notches
shown in Fig. 2. Test setup
Table 3 shows the properties of the tested strands, named As described in Walsh and Kurama,10 each strand speci-
according to the nominal diameter and differentiated men was positioned through the crossheads of a 600 kip
for each different source (spool) from which the mate- (2700 kN) hydraulic universal testing machine with a
rial was obtained. Strands 0.5A and 0.6A, from the same post-tensioning anchor placed on the outer surface of each
manufacturer, were used exclusively in the testing of the crosshead. All of the monostrand barrel anchors and the
strand-anchorage assemblies, while all of the 10 different nine-hole multistrand anchor (multi-0.6A) bore directly
strands listed, representing five different manufacturers, on 1.5-in.-thick (38 mm) steel plates, each with a central
were used in the free-length fracture tests described at the hole oversized 1/16 in. (1.6 mm) with respect to the nominal
end of this paper. The strand properties in the table are strand diameter. The bearing plates were screwed to the
average measurements from a minimum of three samples. machine crossheads and the central holes were aligned
The measured wire and strand diameters—diameter of vertically using a laser alignment tool such that each strand
middle strand wire dmw, diameter of outer strand wire dow, sample was placed in the same position with minimal

Table 3. Strand properties
Strand spool dmw, in. dow, in. dp, in. ap, in.2 fpm,free-length, ksi εpf,free-length
0.5A 0.171 0.162 0.500 0.149 286.7 0.0736
0.5B 0.171 0.165 0.501 0.151 286.4 0.0739
0.5C 0.171 0.165 0.501 0.152 285.2 0.0673
0.5D 0.171 0.165 0.501 0.153 288.9 0.0638
0.5E 0.173 0.165 0.502 0.153 285.7 0.0672
0.6A 0.205 0.196 0.600 0.219 280.9 0.0752
0.6B 0.207 0.198 0.597 0.215 275.7 0.0676
0.6C 0.207 0.197 0.597 0.216 283.2 0.0726
0.6D 0.205 0.197 0.601 0.219 278.7 0.0688
0.6E 0.205 0.197 0.598 0.218 280.1 0.0718
Note: a p = strand cross-sectional area; dmw = middle-wire diameter; dow = outer-wire diameter; dp = strand diameter; fpm,free-length = maximum free-length
strength; εpf,free-length = free-length fracture strain. 1 in. = 25.4 mm; 1 ksi = 6.895 MPa.
accidental eccentricity between the ends of the strand. the extensometer gauge length, little strand deformation
Further validation on the concentric axial loading of the occurred at the fracture location due to the lack of neck-
test specimens (for example, by testing strand samples with ing (that is, reduction in wire diameter) in the fractured
extended free-lengths) was also conducted.10,16 wires (Fig. 1). Thus, the extensometer strains from the
post-tensioning strand-anchorage tests provide an accu-
At the beginning of each test, the crossheads of the testing rate measurement of the strand strains up to fracture. All
machine were positioned such that a strand free length of test equipment components, including the load cell and
about 72 in. (1800 mm) was achieved between the top and position rate of the testing machine, as well as the gauge
bottom post-tensioning anchors after seating the anchor lengths and strain measurements of the extensometers,
wedges. The anchor wedges were placed and tapped in light- were calibrated at least yearly on-site. Comparisons and
ly by hand using the back of a spare anchor barrel before a validation of the measurements from the two extensom-
preload of approximately 750 lb (3300 N) was applied to eters were also made.10,16
the strand. Then the wedges were checked for proper initial
seating, the extensometers were initialized, and the strand For the monostrand cast anchors with the wedge hole
was loaded to fracture by moving the top machine crosshead protruding from the bottom surface of the anchor (referred
while the bottom crosshead remained stationary. to as the anchor boss [Fig. 2]), each anchor was screwed to
a fitted steel support plate10 to use the full bearing area of
During each test, the strand stress and strain data, as well the anchor, simulating an anchor embedded inside con-
as the time elapsed, applied load, crosshead displacement, crete. The necessity for these support plates was proved by
and strand elongation (based on the strain data) were preliminary tests that compared strand-anchorage perfor-
recorded continuously at a frequency of 100 Hz. The strand mance with and without support plates. The support plates
stress was calculated by dividing the measured load by the were machined such that the bottom of the anchor boss
measured strand area ap (Table 3). Each test was concluded was flush with the underside of the plate.
following the point at which the first strand wire fracture
occurred, resulting in a sudden drop in stress. After the test, Multistrand anchorages
the number of simultaneous wire fractures and the loca-
tions of the fractures were recorded. The multistrand anchorages tested in this research had
multiple wedge holes on the perimeter and one central
Two extensometers (with 2.0 in. [50 mm] and 36 in. hole (Fig. 2 shows a nine-hole anchor). As specified by the
[910 mm] gauge lengths) were placed within the free Post-Tensioning Institute’s (PTI’s) Acceptance Standards
length of the strand (as required by ICC-ES) between the for Post-Tensioning Systems,18 these anchors were tested
anchors. While both extensometers were used to measure one strand at a time because of the uncertainties associ-
the strand strains up to fracture, the strain measurements ated with loading multiple strands simultaneously (for
presented in this paper are from the 36 in. extensometer example, difficulties in loading the strands evenly). Only
unless stated otherwise. While the strand wire fractures oc- the outside wedge holes were used because of the different
curred inside the post-tensioning anchors, and thus outside anchor geometry around the middle hole. After each test,

300,000
0.95fpu
0.85fpu 0.90fpu
250,000 0.80fpu
0.75fpu
200,000
Stress, psi
0.70fpu
150,000
100,000
50,000
0
0 1 2 3 4
Time, min
Cyclic loading with strain rate of

Eccentric strand 0.056/min Prestressing jack
Figure 3. The effects of various loading parameters on the strand fracture strains were investigated through different testing procedures. Note: fpu = nominal ultimate
strand strength. 1 in. = 25.4 mm; 1 psi = 6.895 kPa.
the anchor wedges and a short gripped length of the test procedure was applied to three to twelve samples of each
strand were left inside the wedge hole. For the nine-hole of the anchor-wedge configurations from Tables 1 and 2.
anchors (multi-0.6A), which were tested resting directly on
the 1.5-in.-thick (38 mm) bearing plates connected to the • ICC: This control procedure10,16 conformed to the
top and bottom machine crossheads, the strand was ground static testing requirements of ICC-ES. After a preload
flush with the anchor bottom to allow subsequent testing of of approximately 750 lb (3300 N) was applied, the
the remaining holes. strand was loaded to fracture by monotonically mov-
ing the top machine crosshead upward to result in a
The seven-hole anchors (multi-0.6B) were each composed constant nominal strain rate of about 0.02/min.
of two parts, with a wedge plate placed on top of a 4-in.-
deep (100 mm) hollow cylindrical anchor pedestal. The • SLO: In the SLO procedure, a slower monotonic
wedge plate–pedestal assembly was placed directly on the strand strain loading rate of about 0.01/min was used.
bearing plate at each end of the strand, and the relative This slow loading rate was studied because the control
position of the machine heads was decreased to maintain ICC rate of 0.02/min was at the high end of the ac-
a 72 in. (1830 mm) strand free-length between the wedge ceptable ICC-ES range, allowing for the investigation
plates. Also, because the wedge plates were tested resting of any differences in the strand behavior due to the
on the hollow pedestals, after each test, the short gripped different strain rate.
lengths of the fractured strand only needed to be cut short
(that is, not ground flush with the bottom of the wedge • ECC: In the ECC procedure, a strand end eccentricity
plate) to fit inside the pedestals. of 3.3% was applied at the control monotonic strain
rate of 0.02/min. The importance of testing specimens
During the investigation of the multistrand anchors, both eccentrically stems from the differential transverse
the wedge hole locations used in the different tests and the displacements the strand ends can undergo in an
test results were closely monitored to determine whether unbonded post-tensioned structure under lateral load-
differences in strand fracture strains might have occurred ing. Furthermore, load eccentricity may occur due to
due to the presence of previously loaded wedge holes the presence of strand flares inside the trumpet at the
(containing short lengths of fractured strand) next to the end of a multistrand tendon. A strand end eccentricity
hole being tested. No significant change in strand-anchor as large as 3.3% can develop under such conditions,
performance was found because of the previously loaded which was simulated by offsetting the hole in the top
holes in the wedge plates. bearing plate by 2.4 in. (61 mm) (Fig. 3).
Testing procedures • CYC: The CYC procedure applied postyield cyclic

loading to simulate a seismic event. The testing ma-
The effects of various loading parameters on the strand chine crosshead was position controlled to produce
fracture strains were investigated through the seven dif- the control strain rate of 0.02/min. The movements of
ferent testing procedures described in this section. Each the machine head were bounded by percentages of the

nominal ultimate design strength fpu of the strand, as mitted to be jacked to a stress of about 0.80fpu, which
shown in Fig. 3 for the CYC-FST procedure described is the same as the maximum jacking stress applied in
subsequently. The major cycles were in 0.05fpu incre- this research; however, after losses they typically have
ments, and the intermediate cycles were targeted at the a greater effective prestress of about 0.65fpu. The short
summation of the baseline stress (0.70fpu) and 30% of length of the test strands prevented these prestress
the difference between the peak stress of the preced- levels from being achieved in this project.
ing major cycles and the baseline stress. This cyclic
load pattern is not in accordance with the ICC-ES • EQK: In the EQK procedure, the loading variables from
fatigue test requirements because the ICC-ES upper the ECC, CYC-FST, and PRE tests were combined to
stress limit of 0.85fpu is not high enough to observe the create the best simulation of a seismic event possible
possible postyield cyclic behavior of a strand under with the available equipment by prestressing the strand
extreme seismic loading. incrementally per the description of the PRE procedure
and applying a fast, cyclic, and eccentric load.
• CYC-FST: Because the control strain rate used in the
CYC procedure is unrealistically slow for an earth- Results from the strand-
quake, the CYC-FST procedure was run at the testing anchorage experiments
machine’s fastest possible strain loading rate of 0.056/
min. This loading procedure was also not in accor- All of the strand wire fractures described in this section
dance with the ICC-ES fatigue loading requirements. occurred inside the post-tensioning anchor-wedge assembly
The target loading history (Fig. 3) was only roughly within the first few full-depth wedge teeth where the strand
met under the increased strain rate of the CYC-FST entered the anchor. Most of the tests resulted in a single
procedure because of limitations in controlling the fractured wire; however, as many as four near-simultaneous
movements of the testing machine head at this load wire fractures were observed. Due to the brittle nature of the
rate. wire fractures in these experiments (that is, no observable
necking in the strand wires), the fracture strain point from
• PRE: In the PRE procedure, a prestressing force was each test virtually coincided with the point of peak stress.
applied to the strand before continuing to load it
monotonically at the control strain rate of 0.02/min. Initially, each test procedure was applied to three to five
The prestressing force was applied using a hydraulic samples each of six anchor-wedge configurations from
jack equipped with a power seater (Fig. 3). There are Tables 1 and 2. Then, the CYC procedure was applied
two differences between prestressing with a jack and to twelve samples each of two additional anchor-wedge
simply allowing the movement of the testing machine configurations. No consistent performance differences
head to stress the strand: were observed among the anchor-wedge specimens from
different manufacturers.
— The jack creates a stressing-end effect at one
of the anchors (in this research, the top anchor) Loading parameter test results
by reducing the diameter of the strand (due to
the tightening of the strand wires) before power Tables 4–10 list the initial six anchor-wedge configurations
seating the wedges, possibly resulting in deeper tested; the sample size used for each configuration per pro-
wedge seating at the stressing end. cedure; the average, standard deviation, and coefficient of
variation (CV) of the fracture strains from the ICC control
— Prestressing with the hydraulic jack occurs at a tests; and comparative data from the other test procedures
faster strain rate than is achieved by the movement with respect to the ICC test results. Similarly, Fig. 4 and
of the test machine head. 5 contain a series of strand fracture stress versus fracture
strain plots comparing a base parameter (represented by the
The 0.5-in.-diameter (13 mm) strands were incre- red/dark markers, typically the ICC test data) with a varied
mentally jacked to 750 lb, 25,000 lb, and 33,000 lb parameter (represented by the yellow/light markers). As
(3300 N, 111,200 N, and 146,800 N), which corre- shown by the diamond and circle markers in these figures,
sponds to a maximum jacking stress of about 0.80fpu. the 0.5-in.-diameter (13 mm) strand specimens had a higher
After wedge seating losses, this procedure resulted yield stress than the 0.6-in.-diameter (15 mm) strand speci-
in a strand prestress of 0.46fpu to 0.50fpu. Similarly, mens, a trend that is discussed later in the paper.
the 0.6-in.-diameter (15 mm) strands were jacked to
750 lb, 36,000 lb, and 46,750 lb (3300 N, 160,100 N, Looking at the entire six-anchor ensemble, the loading
and 208,000 N), resulting in a prestress of 0.54fpu parameters associated with the CYC, CYC-FST, and EQK
to 0.57fpu. These jacking procedures were found to procedures result in significantly reduced average fracture
provide the maximum prestress levels possible without strains. Based on the entire data set, the differences be-
yielding the tested strands. In practice, strands are per- tween the fracture strains from these three test procedures

Table 4. Numerical parametric test results for ICC
Anchor Wedge Number of samples Fracture strain Standard deviation CV
Cast-0.5C 2P-0.5C 4 0.0271 0.0081 0.30
Cast-0.5D 2P-0.5E 4 0.0273 0.0095 0.35
Cast-0.6A 2P-0.6A 5 0.0269 0.0021 0.08
Barrel-0.5C 2P-0.5F 5 0.0186 0.0044 0.24
Barrel-0.6A 3P-0.6B 3 0.0229 0.0030 0.13
Multi-0.6A 2P-0.6C 4 0.0261 0.0041 0.16
Total 25 0.0247 n.a. n.a.
Note: CV = coefficient of variation; ICC = standard monotonic test procedure based on International Code Council Evaluation Service11 requirements;
n.a. = not applicable.
Table 5. Numerical parametric test results for SLO
Number Difference in Difference in

Anchor Wedge Difference in CV, %
of samples fracture strain, % standard deviation, %
Cast-0.5C 2P-0.5C 4 18.8 -24.5 -36.5
Cast-0.5D 2P-0.5E 4 -9.6 -54.3 -49.4
Cast-0.6A 2P-0.6A 5 19.2 56.3 31.1
Barrel-0.5C 2P-0.5F 5 -7.8 40.9 52.8
Barrel-0.6A 3P-0.6B 3 -14.8 48.1 73.7
Multi-0.6A 2P-0.6C 4 9.3 65.0 51.0
Total 25 4.5 n.a. n.a.
Note: CV = coefficient of variation; ICC = standard monotonic test procedure based on International Code Council Evaluation Service11 requirements;
n.a. = not applicable; SLO = monotonic test procedure with slower loading rate than ICC procedure.
Table 6. Numerical parametric test results for ECC

Cast-0.5C 2P-0.5C 4 -15.6 4.3 23.6
Cast-0.5D 2P-0.5E 4 9.5 -40.7 -45.8
Cast-0.6A 2P-0.6A 5 17.6 164.0 125.0
Barrel-0.5C 2P-0.5F 5 1.0 22.4 21.2
Barrel-0.6A 3P-0.6B 3 -10.7 56.0 74.7
Multi-0.6A 2P-0.6C 4 -5.8 103.0 116.0
Total 25 0.7 n.a. n.a.
Note: CV = coefficient of variation; ECC = monotonic test procedure with eccentric loading; n.a. = not applicable.

Table 7. Numerical parametric test results for CYC

Cast-0.5C 2P-0.5C 4 -13.1 -8.6 5.2
Cast-0.5D 2P-0.5E 4 -28.3 -26.0 3.3
Cast-0.6A 2P-0.6A 5 2.9 176.0 168.0
Barrel-0.5C 2P-0.5F 5 -28.6 -64.9 -50.9
Barrel-0.6A 3P-0.6B 3 3.5 -43.6 -45.5
Multi-0.6A 2P-0.6C 4 -17.7 47.6 79.3
Total 25 -13.6 n.a. n.a.
Note: CV = coefficient of variation; CYC = cyclic test procedure with same loading rate as ICC procedure; ICC = standard monotonic test procedure
based on International Code Council Evaluation Service11 requirements; n.a. = not applicable.
Table 8. Numerical parametric test results for CYC-FST

Cast-0.5C 2P-0.5C 4 -13.5 -30.6 -19.8
Cast-0.5D 2P-0.5E 4 -5.8 -72.1 -70.3
Cast-0.6A 2P-0.6A 5 -11.9 222.0 266.0
Barrel-0.5C 2P-0.5F 5 -28.5 -43.0 -20.3
Barrel-0.6A 3P-0.6B 3 -20.0 -32.9 -16.1
Multi-0.6A 2P-0.6C 4 -5.5 -23.9 -19.5
Total 25 -13.4 n.a. n.a.
Note: CV = coefficient of variation; CYC-FST = cyclic test procedure with faster loading rate than CYC procedure; n.a. = not applicable.
Table 9. Numerical parametric test results for PRE

Cast-0.5C 2P-0.5C 4 -1.1 -32.0 -31.3
Cast-0.5D 2P-0.5E 4 -5.9 -51.9 -49.0
Cast-0.6A 2P-0.6A 5 -1.4 71.4 73.8
Barrel-0.5C 2P-0.5F 5 -28.3 36.1 89.8
Barrel-0.6A 3P-0.6B 3 0.8 29.0 28.0
Multi-0.6A 2P-0.6C 4 9.1 -7.2 -15.0
Total 25 -4.1 n.a. n.a.
Note: CV = coefficient of variation; n.a. = not applicable; PRE = monotonic test procedure including a prestress.

Table 10. Numerical parametric test results for EQK

Cast-0.5C 2P-0.5C 4 -25.0 -14.7 13.7
Cast-0.5D 2P-0.5E 4 -5.8 20.7 28.0
Cast-0.6A 2P-0.6A 5 4.7 75.1 67.3
Barrel-0.5C 2P-0.5F 5 -17.6 -23.8 -7.5
Barrel-0.6A 3P-0.6B 3 -2.4 50.7 54.4
Multi-0.6A 2P-0.6C 4 -25.2 124.0 199.0
Total 25 -11.6 n.a. n.a.
Note: CV = coefficient of variation; EQK = cyclic test procedure with a prestress, eccentricity, and fast loading rate; n.a. = not applicable.
280,000 280,000
Fracture stress, psi
270,000 270,000
260,000 260,000
250,000 250,000
240,000 240,000
230,000 230,000
0.00 0.01 0.02 0.03 0.04 0.05 0.00 0.01 0.02 0.03 0.04 0.05
Fracture strain Fracture strain
ICC (0.5 in.) SLO (0.5 in.) ICC (0.5 in.) ECC (0.5 in.)
ICC (0.6 in.) SLO (0.6 in.) ICC (0.6 in.) ECC (0.6 in.)
ICC versus SLO ICC versus ECC
280,000 280,000
270,000 270,000
260,000 260,000
250,000 250,000
240,000 240,000
230,000 230,000
0.00 0.01 0.02 0.03 0.04 0.05 0.00 0.01 0.02 0.03 0.04 0.05
ICC (0.5 in.) CYC (0.5 in.) ICC (0.5 in.) CYC-FST (0.5 in.)
ICC (0.6 in.) CYC (0.6 in.) ICC (0.6 in.) CYC-FST (0.6 in.)
ICC versus CYC ICC versus CYC-FST
Figure 4. Effects of loading parameters on strand fracture stress versus fracture strain. Note: CYC = cyclic test procedure with same loading rate as ICC procedure;
CYC-FST = cyclic test procedure with faster loading rate than CYC procedure; ECC = monotonic test procedure with eccentric loading; ICC = standard monotonic test
procedure based on International Code Council Evaluation Service11 requirements; SLO = monotonic test procedure with slower loading rate than ICC procedure. 1 in.
= 25.4 mm; 1 psi = 6.895 kPa.

280,000 280,000

270,000 270,000
260,000 260,000
250,000 250,000
240,000 240,000
230,000 230,000
0.00 0.01 0.02 0.03 0.04 0.05 0.00 0.01 0.02 0.03 0.04 0.05
CYC (0.5 in.) CYC-FST (0.5 in.) ICC (0.5 in.) PRE (0.5 in.)
CYC (0.6 in.) CYC-FST (0.6 in.) ICC (0.6 in.) PRE (0.6 in.)
CYC versus CYC-FST ICC versus PRE
280,000 280,000

270,000 270,000
260,000 260,000
250,000 250,000
240,000 240,000
230,000 230,000
0.00 0.01 0.02 0.03 0.04 0.05 0.00 0.01 0.02 0.03 0.04 0.05
ICC (0.5 in.) EQK (0.5 in.) CYC-FST (0.5 in.) EQK (0.5 in.)
ICC (0.6 in.) EQK (0.6 in.) CYC-FST (0.6 in.) EQK (0.6 in.)
ICC versus EQK CYC-FST versus EQK
Figure 5. Further comparisons among the test results. Note: CYC = cyclic test procedure with same loading rate as ICC procedure; CYC-FST = cyclic test pro-
cedure with faster loading rate than CYC procedure; EQK = cyclic test procedure with a prestress, eccentricity, and fast loading rate; ICC = standard monotonic
test procedure based on International Code Council Evaluation Service 11 requirements; PRE = monotonic test procedure including a prestress. 1 in. = 25.4 mm;
1 psi = 6.895 kPa.
are small on average, indicating that the dominant param- dard deviation and coefficient of variation due to the varied
eter is the postyield cyclic loading. The loading parameter loading parameters are generally greater than the decreases
from the PRE procedure has a slightly negative effect, as in these measures, and the overall data ranges from the
controlled by the greatly reduced fracture strains for only varied loading parameter procedures in Fig. 4 and 5 tend to
one of the six anchor-wedge configurations, whereas the be larger than those from the base procedure.
SLO procedure has a slightly positive effect and the ECC
procedure has almost no effect on average. However, these Stressing-end effect
are average trends from the entire ensemble and the effects
of the loading parameters greatly depend on the anchor- Figure 6 combines the data from the EQK and PRE
wedge configuration. procedures, exhibiting the strand fracture end from each
test. Twenty-five tests from this combined data pool had
The smallest fracture strains from the varied loading wire fractures occur inside the bottom anchor, 24 tests had
parameter test data are always smaller than the smallest fractures occur inside the top anchor (the prestressing end),
fracture strains from the base test data (that is, the left-most and one test had wire fractures occur simultaneously inside
yellow markers are always to the left of the left-most red both anchors. Based on these data, it is concluded that the
markers) (Fig. 4 and 5). This means that the varied loading wire fractures occurred fairly randomly inside the top or
parameter always results in the single worst strand-anchor bottom anchor, and the use of a prestressing jack in the
performance in each comparison chart. The effects of the EQK and PRE procedures did not have an observable ef-
loading parameters on the standard deviation and coef- fect on the fracture end. Tests with fractures inside the top
ficient of variation of the fracture strains vary significantly. anchor at the prestressing end, however, did appear to have
Nevertheless, Tables 4–10 show that the increases in stan- a slightly larger range of fracture strains.

300,000
0.05
0.85fpu
250,000
0.04
200,000
Stress, psi
Fracture strain
0.03
150,000
0.02 100,000
0.40fpu
0.01 50,000
0.00 0
-2 -1 (25 points)0 Top anchor1(24 points) 2 0 2 4 6 8 10 12
Bottom anchor
Time, min.
PRE EQK
Figure 7. Fifty-cycle loading pattern with strain rate of 0.056/min com-
Figure 6. The test data from the EQK and PRE procedures are combined, exhib- pleted without strand fracture. Note: fpu = nominal ultimate strand strength.
iting the fracture end from each test. Note: EQK = cyclic test procedure with a 1 psi = 6.895 kPa.
prestress, eccentricity, and fast loading rate; PRE = monotonic test procedure
including a prestress.
Figure 7 shows the stress-time history from one of the

Fifty-cycle load tests tests, exhibiting that all 50 cycles were completed without
strand fracture, which was, indeed, the case with all three
The 50-cycle ICC-ES “fatigue” load test requires that a strand- specimens.
anchorage system not fracture when subjected to 50 cycles of
loading ranging from 0.40fpu to 0.85fpu at a frequency of 1 Hz Further investigation
to 3 Hz. The upper stress limit required by the ICC-ES fatigue of postyield cyclic loading
load test is too low to observe the postyield cyclic behavior
that may occur in strands under extreme seismic loading. Postyield cyclic loading appears to be the common loading
Nevertheless, three specimens, each consisting of a cast-0.5B parameter among the procedures investigated in this paper
anchor with 3P-0.5B wedges, were subjected to 50-cycle load that tends to cause significant reductions in the strand frac-
tests within the ICC-ES fatigue stress range. Note that the ture strains. To further validate the influence of postyield
testing machine used in this research is not capable of meeting cyclic loading, twelve samples each of two additional
the ICC-ES fatigue cycle frequency requirement of 1 Hz to anchor-wedge configurations (Table 11) were subjected to
3 Hz, as the tests were conducted at the same strain rate as that the ICC and CYC procedures and the test data were ana-
in the CYC-FST and EQK procedures (that is, 0.056/min), lyzed in terms of the p-value of a two-sample t-test with a
corresponding to a loading frequency of about 0.08 Hz. one-sided alternative hypothesis.19
Table 11. Effect of postyield cyclic loading on additional anchors

Cast-0.5B Multi-0.6B
Wedge 3P-0.5B 3P-0.6C
Number of samples 12 12
Average fracture strain 0.0267 0.0204
ICC Standard deviation 0.0040 0.0056
CV 0.15 0.27
Average fracture strain 0.0203 0.0195
Difference in fracture strain, % -23.8 -4.7

Standard deviation 0.0059 0.0052
CYC
Difference in standard deviation, % 49.4 -7.6
CV 0.29 0.27
Difference in CV, % 96.2 -3.0
p-value × 100% 0.27 33.1
Note: CV = coefficient of variation; CYC = cyclic test procedure with same loading rate as ICC procedure; ICC = standard monotonic test procedure
based on International Code Council Evaluation Service11 requirements.

280,000 280,000

270,000 270,000
260,000 260,000
250,000 250,000
240,000 240,000
230,000 230,000
0.00 0.01 0.02 0.03 0.04 0.05 0.00 0.01 0.02 0.03 0.04 0.05
ICC CYC ICC CYC
Cast-0.5B/3P-0.5B anchor/wedge configuration Multi-0.6B/3P-0.6C anchor/wedge configuration
Figure 8. Among the procedures investigated in this paper, postyield cyclic loading appears to be the common loading parameter that tends to cause significant
reductions in the strand fracture strains. Note: CYC = cyclic test procedure with same loading rate as ICC procedure; ICC = standard monotonic test procedure based
on International Code Council Evaluation Service11 requirements. 1 psi = 6.895 kPa.
Strand
Differential seating
Anchor
Wedge Wedge
Poor wedge differential seating
0.05 0.05
0.04 0.04
Fracture strain
Fracture strain
0.03 0.03
0.02 0.02
0.01 0.01
0.00 0.00
0.00 0.05 0.10 0.15 0.20 0.00 0.05 0.10 0.15 0.20
Wedge differential seating, in. Wedge differential seating, in.
Cast-0.5B anchors with 2P-0.5C wedges Cast-0.5C anchors with 2P-0.5C wedges
Figure 9. Poor wedge seating may be associated with low fracture strains, but low fracture strains are not entirely dependent on poor wedge seating.
Note: 1 in. = 25.4 mm.

Given that even the relatively large sample size of twelve 0.05
specimens (compared with the three specimens required
0.04
by ICC-ES) is limited, the p-value expressed as a percent-
Fracture strain
age in Table 11 provides the probability of observing the 0.03
measured difference in the test averages between the two
procedures if the actual averages from the two procedures 0.02
(which could only be determined by running an unfeasi-
bly large number of tests) were, in fact, equal. A p-value 0.01
of 0.27% for the first anchor-wedge configuration, for
0.00
example, indicates that it is nearly impossible for the actual 0.00 0.05 0.10 0.15 0.20
averages from the two procedures to be equal. Conclu-
Wedge differential seating, in.
sively, the CYC procedure causes a significant reduction in
the average fracture strain of this particular anchor-wedge
Figure 10. Wedge differential seating measured for all of the tests conducted in
configuration. The p-value of the second anchor-wedge
this research project. Note: 1 in. = 25.4 mm.
configuration is not nearly as conclusively small, but it still
indicates a favorable degree of likelihood that the CYC
procedure causes a reduction in the average fracture strain the consistency and ease with which the 0.10 in. limit is
of that particular configuration. applied rendered it useful in invalidating certain test data in
this research. Of course, the cutoff differential seating could
Figure 8 plots the fracture stress and strain for each of the be adjusted anywhere from about 0.08 in. (2.0 mm) to about
anchor-wedge configurations in Table 11, comparing the 0.12 in. (3.0 mm) in accordance with these data.
data from the two loading procedures. Not only are the
averages of the ICC data higher than the averages of the Results from the strand free-
CYC data for both configurations but consistent with the length fracture experiments
trend observed in earlier plots, an ICC data point represents
the largest fracture strain in each configuration and a CYC As a corollary to the post-tensioning anchorage experiments,
data point represents the smallest fracture strain in each samples from all of the 10 strand spools (Table 3) were
configuration as well. tested to result in free-length seven-wire fracture of each
specimen in accordance with ASTM A370.20 These tests
Wedge differential seating were conducted using the same setup and equipment as the
strand-anchorage experiments described previously, but with
An analysis of the relationship between wedge differential special “sand-grip” anchors10,16 to achieve free-length failure.
seating and fracture strain was conducted. The data for Table 3 lists the average maximum free-length strength
this investigation was accumulated from the normal testing fpm,free-length and fracture strain εpf,free-length for each strand.
pool, and no test was run specifically to control the amount
of wedge differential seating because the investigators Figure 11 shows the average stress-strain behaviors from
wanted to observe the tendency of wedge seating under all tests of the five different spools in each strand diameter,
normal conditions. All of the tests presented thus far in the demonstrating considerable differences between the 0.5 in.
paper had small differential seating between the anchor (13 mm) and 0.6 in. (15 mm) strand sizes. As described
wedges. Any test with more than 0.10 in. (2.5 mm) of dif- in the following section, this information becomes impor-
ferential wedge seating at wire fracture was discarded from tant when determining a relationship that can be used to
the main test data pool but was recorded for the purposes correlate strand strains with stresses for design purposes.
of the particular investigation described in this section. The free-length fracture strain is reached at an apparent
stress slightly smaller than the maximum strength because
Figure 9 shows a photograph of well-seated wedges and of the necking (that is, reduction of wire diameter prior
also illustrates poor wedge differential seating. The test to fracture) that occurs in the strand wires in the case of
data in this figure, taken from two different anchor-wedge free-length fracture (Fig. 11). This behavior is different
configurations, also exhibit that poor wedge seating may from the brittle wire fractures observed inside the post-
be associated with low fracture strains but that low fracture tensioning anchors as described in the previous section,
strains are not entirely dependent on poor wedge seating. which occurred with no visible necking (Fig. 1), resulting
in the maximum strength and fracture strain being reached
Figure 10 shows the wedge differential seating measured for at virtually the same point on the stress-strain curve.
all of the tests conducted in this research project (including the
tests described in Walsh and Kurama)10 compared with their Strand linear-elastic modulus
fracture strains. This data pool includes wedges of several
different heights. Therefore, 0.10 in. (2.5 mm) represents a The PCI Design Handbook: Precast and Prestressed
different percentage of each wedge’s height. Nonetheless, Concrete21 assumes a linear-elastic modulus Ep equal to

300,000
250,000
Stress, psi
200,000
150,000
100,000
50,000
0
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Strain
0.5 in. average 0.6 in. average
Average 0.5-in.-diameter and 0.6-in.-diameter Free-length seven-wire strand fracture

strand stress-strain behaviors
Figure 11. Strand free-length fracture experiments were conducted to determine the actual strength and strain capacity of the strands.
Note: 1 in. = 25.4 mm; 1 psi = 6.895 kPa.
28,500 ksi (196,500 MPa) for prestressing strand. Table 12 ference between the two corresponding strains within the
lists the average elastic modulus measured for each of the initial linear-elastic range of the measured strand stress-
strand spools (from three samples each) as well as the aver- strain relationship. For most tests, the two stresses chosen
age elastic modulus for each diameter of strand tested. were 50 ksi (350 MPa) and 150 ksi (1030 MPa), but these
values were altered if irregularities were observed in the
The linear-elastic modulus from each test was calculated as stress-strain curve near these points.
the ratio of the difference between two stresses and the dif-
Table 12. Strand elastic moduli from free-length fracture tests
Average elastic modulus Ep, ksi Average elastic modulus Ep

Strand spool
(from three samples each) for each diameter, ksi
0.5A 28,028
0.5B 29,474
0.5C 28,707 28,682
0.5D 28,276
0.5E 28,926
0.6A 27,872
0.6B 27,827
0.6C 29,605 28,188
0.6D 27,308
0.6E 28,328
Total average 28,435
Note: 1 ksi = 6.895 MPa.

Although the elastic modulus was found to be fairly consis-
tent among samples from the same strand spool, it varied to 280,000
some degree between the different strand spools (Table 12).
In addition to the higher yield stress and maximum strength 270,000
associated with the 0.5-in.-diameter (13 mm) strand, it
Stress, psi
260,000
appears that 0.5-in.-diameter strand tends to have a slightly
250,000
higher elastic modulus as well. Nonetheless, the average of
the entire sample set is 28,435 ksi (196,050 MPa), which 240,000
is close to the value of 28,500 ksi (196,500 MPa) from the 230,000
PCI Design Handbook.21
220,000
0.00 0.01 0.02 0.03
Strand stress–strain relationship
Strain
The information derived from the free-length fracture PCI Combined test average Proposed combined
experiments is important not only for determining anchor-
independent strand capacities but also for determining PCI Design Handbook, combined test average,
and proposed combined relationship
idealized, approximate strand stress–strain relationships
that can be used to correlate strand strains with stresses for
design purposes. The PCI Design Handbook includes the
280,000
following expression to relate stress fps and strain εps for
strand with nominal 270 ksi (1860 MPa) strength:ksi 270,000
Stress, psi 260,000

For εps ≤ 0.0086, fps = 28,500εps ksi
250,000
For εps > 0.0086, fps = 270 - [0.040/(εps - 0.007)] ksi 240,000
230,000
This relationship gives a strand limit-of-proportionality
stress of fpl equal to 245.1 ksi (1690 MPa) at a strand 220,000
limit-of-proportionality strain of εpl equal to 0.0086. The 0.00 0.01 0.02 0.03
PCI Design Handbook proportionality limit stress is high Strain
compared with the results of this investigation and does not PCI 0.5 in. test average Proposed 0.5 in.
distinguish between 0.5 in. (13 mm) and 0.6 in. (15 mm)
strand diameters. Figure 12 plots the PCI Design Hand- PCI Design Handbook, 0.5-in.diameter strand test
average, and proposed 0.5-in.-diameter relationship
book stress-strain relationship against the combined test
average (that is, average of all 0.5-in.-diameter and 0.6-in.-
diameter strand test results) and a proposed relationship
derived from this combined average as follows: 280,000
270,000
For εps ≤ 0.0082, fps = 28,500εps ksi
Stress, psi
260,000
For εps > 0.0082, fps = 270 - [0.080/(εps - 0.006)] ksi 250,000
240,000
This relationship uses a strand limit-of-proportionality
stress of 233.7 ksi (1611 MPa) at a strain of 0.0082. The 230,000
proposed relationship provides a better approximation 220,000
than the PCI Design Handbook relationship does to the 0.00 0.01 0.02 0.03
measured average strand stress–strain curve from the Strain
proportionality limit up to a strain of about 0.02, which is
PCI 0.6 in. test average Proposed 0.6 in.
the most important strain range for most practical design
purposes. Although the differences in the proposed and PCI Design Handbook, 0.6-in.diameter strand test
PCI Design Handbook strand stress–strain relationships in average, and proposed 0.6-in.-diameter relationship
Fig. 12 may look insignificant, small variations in stress
result in rather large differences in strain in this nonlinear
range of behavior. The reason for the deviation of the PCI
Design Handbook strand stress–strain relationship from the
measured data in this paper is unknown. Figure 12. The free-length fracture experiments were used to determine ideal-
ized, approximate strand stress–strain relationships for design purposes.
Note: 1 in. = 25.4 mm; 1 psi = 6.895 kPa.

Table 13. Proposed strand stress–strain relationships
PCI Design Handbook Proposed combined Proposed 0.5 in. diameter Proposed 0.6 in. diameter
εpl 0.0086 0.0082 0.0083 0.0081
fpl, ksi 245.1 233.7 236.6 226.8
Ep, ksi 28,500 28,500 28,500 28,000
α 1.00 1.00 1.00 0.97
a 0.0398 0.0799 0.0435 0.0597
b 0.0070 0.0060 0.0070 0.0064
Note: For εps ≤ εpl, fps = Epεps ksi; for εps > εpl, fps = 270α - [a/(εps - b)] ksi; a = (270α - Epεpl)(εpl - b); b = equation variable; Ep = strand linear-elastic
modulus; fpl = strand limit-of-proportionality stress; fps = strand stress; α = ultimate strength factor; εpl = strand limit-of-proportionality strain; εps =
strand strain. 1 in. = 25.4 mm; 1 ksi = 6.895 MPa.
Similar to the PCI Design Handbook relationship, the tion format. These adjustments include lowering the elastic
proposed strand stress–strain relationship does not account modulus to emulate the average elastic modulus from the
for the strain hardening observed in the test data. Using 0.6-in.-diameter strand tests (Table 12) and adjusting the
the same form as the PCI Design Handbook relation- nominal strength by a factor of 0.97 as follows:
ship, the proposed model simply lowers the propor-
tionality limit and then asymptotically approaches the For εps ≤ 0.0081, fps = 28,000εps ksi
nominal 270 ksi (1862 MPa) strength as the PCI Design
Handbook curve does. The proposed proportionality For εps > 0.0081, fps = 0.97 × 270 - [0.060/(εps - 0.0064)] ksi
limit strain of 0.0082 is high enough for the jacking
stress limit of 0.80fpu permitted in American Concrete This relationship provides a proportionality limit stress
Institute’s (ACI’s) Building Code Requirements for of 226.8 ksi (1575 MPa) at a strain of 0.0081. Figure 12
Structural Concrete (ACI 318-08) and Commentary shows the relevant stress-strain curves. Of course, the
(ACI 318R-08)22 for 270 ksi (1860 MPa) strand. The adjustment factor 0.97 on the nominal strength limits the
proportionality limit stress of 233.7 ksi (1611 MPa) is ultimate strength of the strand, but the conservative nature
less than the minimum yield stress of 0.9fpu or 243 ksi of this adjustment and its necessity to accurately model the
(1675 MPa) specified for low-relaxation strand in ASTM test data between the proportionality limit and a strain of
A416; however, the ASTM A416 yield stress require- 0.02 warrant its use.
ment is defined at a strain of 0.01, which is satisfied by
the proposed stress-strain relationship. Table 13 gives a summary of the proposed stress-strain rela-
tionships. The difference in the average maximum free-length
Given the considerable difference in the yield stresses for strengths of the 0.5-in.-diameter (13 mm) strands (286.6 ksi
the 0.5-in.-diameter (13 mm) and 0.6-in.-diameter (15 mm) [1976 MPa]) and the 0.6-in.-diameter (15 mm) strands
strands, separate stress-strain relationships can also be (279.7 ksi [1928 MPa]) tested in this investigation is significant.
proposed for each diameter. The PCI Design Handbook Assuming any imprecision in load or area measurements is too
stress-strain relationship needs to be altered only slightly small to account for the notable strength differences between
to model the average test data from the 0.5-in.-diameter the two strand diameters, these differences may have been
strand samples as follows: caused by material and/or geometrical variations. To determine
steel material strengths, three middle-wire samples were taken
For εps ≤ 0.0083, fps = 28,500εps ksi from each of the two most tested strand spools (0.5A and
0.6A), which were from the same manufacturer. The middle
For εps > 0.0083, fps = 270 - [0.043/(εps - 0.007)] ksi wires are straight and are slightly larger in diameter than the
outer wires, permitting a more accurate measurement of the
This relationship results in a proportionality limit stress wire cross-sectional area and the application of concentric ten-
of 236.6 ksi (1631 MPa) at a strain of 0.0083. Figure 12 sile loading. The middle-wire samples were tested in a smaller
shows the relevant stress-strain curves. The PCI Design 60 kip (270 kN) hydraulic universal testing machine. Standard
Handbook curve is barely visible beneath the proposed wedge grips available with the testing machine were used to
0.5-in.-diameter (13 mm) strand curve. hold the wires at both ends, and all tests resulted in free-length
fracture.
Modeling the 0.6-in.-diameter (15 mm) strand average
test data requires more significant adjustments to the PCI The wire specimens were tested at a consistent strain rate,
Design Handbook relationship to maintain the same equa- resulting in an average maximum strength of 289.6 ksi

Table 14. Difference in average maximum strength (ksi) between 0.5-in.-diameter and 0.6-in.-diameter strands
0.5-in.-diameter strand 0.6-in.-diameter strand Difference, %
All strand spools 286.6 279.7 2.41
Strands 0.5A, 0.6A 286.7 280.9 2.02
Middle wires 0.5A, 0.6A 289.6 287.8 0.62
Note: 1 in. = 25.4 mm; 1 ksi = 6.895 MPa.
(1997 MPa) for the wires from the 0.5-in.-diameter the expected strand strains exceed this limit, anchors
(13 mm) strand samples and 287.8 ksi (1984 MPa) for the that have been proved to consistently achieve higher
wires from the 0.6-in.-diameter (15 mm) strand samples. strand wire fracture strains should be specified.
The difference between these middle-wire strengths, as-
sumed to depict material tensile strengths, does not account • Poor wedge seating may be associated with low strand
for the considerably greater difference between the average wire fracture strains, but low fracture strains are not
strengths of the entire 0.5-in.-diameter and 0.6-in.-diameter entirely dependent on poor wedge seating.
strand sample pools or between the average strengths of
strands 0.5A and 0.6A (Table 14). Thus the reason for the • The measured yield and maximum strengths of the
observed differences in the 0.5-in.- and 0.6-in.-diameter 0.5-in.-diameter (13 mm) strand samples tested in this
strand strengths is not known. research were higher than those of the 0.6-in.-diameter
(15 mm) strand samples.
Conclusion
• The prestressing strand stress–strain relationship
This research investigated the strand fracture strains of provided by the PCI Design Handbook needs to
unbonded post-tensioning strand-anchorage systems under be updated. Given the notable strength differences
different loading conditions. In addition, stress-strain between the 0.5-in.-diameter (13 mm) and 0.6-in.-
relationships were proposed for seven-wire, low-relaxation, diameter (15 mm) strands, separate stress-strain rela-
uncoated 270 ksi (1860 MPa) prestressing strand. tionships are proposed for each diameter, in addition
to a relationship for the combined 0.5-in.-diameter and
The following conclusions and recommendations are made 0.6-in.-diameter strand data pool.
based on the investigation:
Additional research should investigate the following topics
• Postyield cyclic loading is the parameter that most related to this paper:
consistently reduces the average wire fracture strains
of strand-anchorage systems. • The maximum strand strain demands expected in un-
bonded post-tensioned structures under extreme seis-
• Postyield cyclic loading should be used in the valida- mic loading should be researched and compared with
tion testing of unbonded post-tensioning strand-an- the performance data measured in this research. The
chorage systems for extreme seismic applications. investigation of the maximum strand strain demands
should consider the geometry and type of the structure
• The ICC-ES 50-cycle fatigue loading test uses an (for example, walls, frames, coupling beams), the
inadequate upper stress limit to observe the postyield intensity of the earthquake, and the initial strain level.
cyclic behavior that may occur in strands under ex-
treme seismic loading conditions. • Testing procedures that subject strand-anchorage
systems to cyclic end eccentricity should be devised.
• The large standard deviation and coefficient of variation By comparison, the strand end eccentricity used in
of the fracture strain data, with strains almost as large this research maintained the position of the transverse
as 0.04 and as small as 0.01, or slightly less for a small offset without mirroring it about the concentric strand
number of samples, suggests that a larger sample pool axis during the applied loading. A cyclic eccentricity
size should be used in the validation testing of post- condition (that is, full reversal of the transverse offset)
tensioning strand-anchorage systems than the pool size would more realistically model the effects of seismic
of three samples that is currently required by ICC-ES. loading.
• Unbonded post-tensioned structures under extreme • Testing equipment that is capable of subjecting the
seismic loading conditions should be designed such strand specimens to more realistic (that is, faster)
that the maximum strand strains are less than 0.01. If seismic strain rates should be used.

• Although about 400 strand-anchorage tests were crete Wall Specimen for Seismic Regions. Journal of
conducted as part of this project,16 additional anchor Structural Engineering, V. 137, No. 10 (October): pp.
and wedge configurations could be tested, particularly 1052–1062.
more multistrand anchors, and more tests could be
conducted with selected anchor-wedge configurations 5. Weldon, B., and Y. Kurama. 2010. Experimental Eval-
to increase the sample pool sizes. uation of Post-tensioned Precast Concrete Coupling
Beams. Journal of Structural Engineering, V. 136, No.
• Additional strand types and diameters could be tested 9 (September): pp. 1066–1077.
to view performance differences, such as the newer
(but not yet commonly used) 300 ksi (2070 MPa) 6. American Concrete Institute (ACI) Committee 423.
strength strand. 2007. Specification for Unbonded Single-Strand Ten-
dons and Commentary. ACI 423.7. Farmington Hills,
• Methods to increase the strand wire fracture strains MI: ACI.
inside the anchors, such as the use of longer wedges,
could be investigated. 7. ACI Committee 423. 2005. Recommendations for
Concrete Members Prestressed with Unbonded Ten-
Acknowledgments dons. ACI 423.3R. Farmington Hills, MI: ACI.
This research was funded by PCI under a Daniel P. Jenny 8. Post-Tensioning Institute (PTI). 2000. Specification for
Fellowship. The support of the members of the PCI Unbonded Single Strand Tendons. Phoenix, AZ: PTI.
Research and Development Committee and the project
advisory members, Tom D’Arcy (chair), Ned Cleland, and 9. Innovation Task Group 5. 2007. Acceptance Criteria
Richard Sause is gratefully acknowledged. In addition, the for Special Unbonded Post-Tensioned Precast Struc-
authors recognize the support provided by: Neil Hawkins tural Walls Based on Validation Testing (ACI 5.1-07)
of the University of Illinois at Urbana-Champaign; Susie and Commentary (ACI 5.1R-07). ITG-5.1. Farmington
Nakaki of the Nakaki Bashaw Group Inc.; John Pearson Hills, MI: ACI.
and Roger Pelletier of Wiss, Janney, Elstner Associ-
ates Inc.; John Reihl and Ken Everett of StresCore Inc.; 10. Walsh, K., and Y. Kurama. 2009. Behavior of Unbond-
Alan Stone of Aston Metallurgical Services; Brent Bach, ed Post-tensioning Monostrand Anchorage Systems
Michael Gekhtman, Kapil Khandelwal, Tracy Kijewski- under Monotonic Tensile Loading. PCI Journal, V. 55,
Correa, Albert Miller, Patrick Murren, and Brad Weldon of No. 1 (Winter): pp. 97–118.
the University of Notre Dame; and a number of material
suppliers and donors. The opinions, findings, and conclu- 11. International Code Council Evaluation Service Inc.
sions expressed in the paper are those of the authors and do (ICC-ES). 2007. Acceptance Criteria for Post-tension-
not necessarily reflect the views of PCI or the individuals ing Anchorages and Couplers of Prestressed Concrete.
and organizations acknowledged. AC 303. Whittier, CA: ICC-ES.
References 12. Stanton, J., and S. Nakaki. 2002. Design Guidelines

for Precast Concrete Seismic Structural Systems.
1. Priestley, M., S. Sritharan, J. Conley, and S. Pampanin. Report no. PRESSS-01/03-09. Department of Civil
1999. Preliminary Results and Conclusions from the Engineering, University of Washington, Seattle, WA.
PRESSS Five-Story Precast Concrete Test Building. PCI
Journal, V. 44, No. 6 (November–December): pp. 42–67. 13. Innovation Task Group 1 and Collaborators. 2003.
Special Hybrid Moment Frames Composed of Dis-
2. Ou, Y., M. Chiewanichakorn, A. Aref, and G. Lee. cretely Jointed Precast and Post-tensioned Concrete
2007. Seismic Performance of Segmental Precast Members (ACI T1.2-03) and Commentary (ACI
Unbonded Posttensioned Concrete Bridge Columns. T1.2R-03). ITG T1.2. Farmington Hills, MI: ACI.
Journal of Structural Engineering, V. 133, No. 11
(November): pp. 1636–1647. 14. Innovation Task Group 5. 2008. Design of a Special
Unbonded Post-Tensioned Precast Shear Wall Sat-
3. Kurama, Y., B. Weldon, and Q. Shen. 2006. Experi- isfying ACI ITG 5.1 Requirements. ITG-5.2, draft 5.
mental Evaluation of Post-tensioned Hybrid Coupled Farmington Hills, MI: ACI.
Wall Subassemblages. Journal of Structural Engineer-
ing, V. 132, No. 7 (July): pp. 1017–1029. 15. International Code Council (ICC). 2006. International
Building Code 2006. Falls Church, VA: ICC.
4. Smith, B., Y. Kurama, and M. McGinnis. 2011. Design
and Measured Behavior of a Hybrid Precast Con- 16. Walsh, K., and Y. Kurama. 2009. Behavior and Design

of Unbonded Post-tensioning Strand/Anchorage free-length fracture tests
Systems for Seismic Applications. Structural engineer-
ing research report NDSE-09-02. University of Notre fps = strand stress
Dame, Notre Dame, IN.
fpu = nominal ultimate strand strength (270 ksi in
17. ASTM International. 2006. Standard Specification for this research)
Steel Strand, Uncoated Seven-Wire for Prestressed
Concrete. ASTM A416. West Conshohocken, PA: H = anchor or wedge height
ASTM International.
IW = wedge inside width
18. PTI. 1998. Acceptance Standards for Post-Tensioning
Systems. Phoenix, AZ: PTI. L = cast anchor length
19. Ross, S. M. 2004. Probability and Statistics for Engi- TID = anchor top inside diameter
neers and Scientists. 3rd ed. Burlington, MA: Elsevier
Academic Press. TOD = anchor top outside diameter (equal to BOD in
case of barrel anchors)
20. ASTM International. 2008. Standard Test Methods and
Definitions for Mechanical Testing of Steel Products. TW = wedge top outside width (diameter in case of
Annex A7 Method of Testing Multi-Wire Strand for two-piece wedges)
Prestressed Concrete. ASTM A370. West Conshohock-
en, PA: ASTM International. W = cast anchor width
21. PCI Industry Handbook Committee. 2004. PCI Design α = ultimate strength factor in strand stress–strain
Handbook: Precast and Prestressed Concrete. MNL- equations
120. 6th ed. Chicago, IL: PCI.
εpf,free-length = strand fracture strain from free-length fracture
22. ACI Committee 318. 2008. Building Code Require- tests
ments for Structural Concrete (ACI 318-08) and Com-
mentary (ACI 318R-08). Farmington Hills, MI: ACI. εpl = strand limit-of-proportionality strain
Notation εps = strand strain
a = variable in strand stress–strain equations
ap = strand cross-sectional area
b = variable in strand stress–strain equations
BID = anchor bottom inside diameter 
BOD = anchor bottom outside diameter (equal to TOD

in case of barrel anchors)
BW = wedge bottom outside width
dmw = diameter of middle strand wire
dow = diameter of outer strand wire
dp = outside crown-to-crown diameter of strand
Ep = strand linear-elastic modulus
fpl = strand limit-of-proportionality stress
fpm,free-length = strand strength (maximum strand stress) from

About the authors nominal strand diameters were included in the investi-
gation.
Kevin Q. Walsh, MS, E.I., is a
project engineer with Frost While the effects of the loading parameters were found
Engineering and Consulting in to greatly depend on the anchor-wedge configuration,
Mishawaka, Ind. and an adjunct the test results show that extreme seismic loading con-
professor at the University of ditions can cause a significant reduction in the strand
Notre Dame. strain at fracture as well as an increase in the standard
deviation of the fracture strains. Current industry
requirements for the acceptance testing of post-ten-
Yahya C. Kurama, PhD, P.E., sioning strand-anchorage systems were assessed based
MPCI, is a professor at the on the investigation. In addition, as a corollary to the
University of Notre Dame. strand-anchorage experiments, stress-strain relation-
ships are proposed for prestressing strand based on
experiments that achieve free-length fracture of the test
specimens.
Abstract Keywords
This paper presents an experimental investigation Anchor, cyclic, eccentricity, post-tensioning, seismic,
on the strand wire fracture stresses and strains in strand, stress-strain relationship, wedge, wire fracture.
unbonded post-tensioning strand-anchorage systems
subjected to a variety of loading conditions that can Review policy
affect the performance of the strand inside the anchor.
While unbonded post-tensioned gravity load sys- This paper was reviewed in accordance with the
tems (for example, floor and roof slabs) are common Precast/Prestressed Concrete Institute’s peer-review
throughout the United States, the use of this construc- process.
tion technique for seismic-resistant structures can put
the strands under significantly greater strain demands. Reader comments
Strand wire fractures can occur inside the anchorages,
limiting the seismic performance of the structure. Please address any reader comments to journal@pci
The research focuses on the following loading param- Journal, 200 W. Adams St., Suite 2100, Chicago, IL
eters: loading rate, eccentricity between the strand 60606. J
ends, postyield cyclic loading, and initial strand stress
(that is, prestress). Eight representative anchor-wedge
configurations with 0.5 in. (13 mm) or 0.6 in. (15 mm)

Seismic performance and
retrofit of precast concrete
grouted sleeve connections
P
recast concrete structures are used worldwide because
of their shorter construction time, cost effectiveness,
and better plant control of structural elements and
materials compared with cast-in-place concrete structures.
In Italy, the most common applications of precast concrete

buildings are industrial warehouses and commercial malls.
The typical structural layout of these buildings consists of
cantilever columns connected by simply supported precast,
prestressed concrete beams that support prestressed con-
crete roof elements. The columns are inserted and grouted
in place in isolated precast concrete socket foundations.
■ Experimental tests on the cyclic behavior of column-to-founda- This structural layout reduces construction time and is cost
tion subassemblies were conducted to compare the response effective. However, this effectiveness could be compro-
of grouted sleeve connections with those of cast-in-place and mised for construction in seismic regions, especially if
pocket-foundation connections. the design follows the capacity design rule. The design of
column footings is based on the assumption that a plastic
■ Confinement by the grouted sleeves inhibited buckling of the hinge may develop at the column base in an earthquake
longitudinal reinforcement and increased the compressive event, while capacity design is used to prevent the forma-
strength of the grout. tion of an inelastic mechanism in the footing. The latter
goal is achieved by considering the bending moment asso-
■ The damage was localized to the column base, allowing ciated with the column base flexural capacity acting on the
easier postseismic column repair compared with traditional foundation amplified by an overstrength factor. Following
connections. this design approach, the foundation base becomes large,

even for medium-sized columns. This could limit the cost Experimental program
effectiveness of industrial and commercial precast concrete
structures. In fact, because of the foundation’s dimension, Test setup
the footings must be cast-in-place concrete, making the use
of isolated socket foundations less attractive. This experimental program included six tests: one on each
of five specimens with different column-to-foundation con-
Hence, mat foundations are more convenient than isolated nections and approximately the same maximum bending
footings. When using mat foundations, the following moment capacity and a test on one of the original speci-
column-to-foundation connections are typically used:1–3 mens after retrofitting. The columns had cross sections
of 400 mm × 400 mm (16 in. × 16 in.) and heights of
• bolted base plates embedded in the foundation 3200 mm (126 in.). Two hollow-core plunger jacks applied
a constant axial force of 600 kN (140 kip) to all columns.
• foundation pockets in which the columns are placed The axial force represents a gravity load compatible with
and grouted the structural framing system. An electromechanical screw
jack with a 500 mm (20 in.) maximum stroke and 1000 kN
• grouted sleeves (225 kip) capacity applied a cyclic horizontal displacement
at the top of the column, imposing an increasing column
• mechanical splices drift from 0.25% to 5%. Figure 1 shows the test setup.
This paper investigates the behavior and performance of The horizontal loads and displacements were recorded by
grouted corrugated steel sleeve connections under cyclic a load cell and a string pot transducer, respectively, while
loading. While the behavior of precast concrete socket base three pairs of linear displacement transducers evaluated
connections is well documented,4–6 no published experi- the average column curvature recorded at three consecu-
mental results concerning the cyclic response of grouted tive levels of the column base (Fig. 2). The lower pair of
sleeve column-to-foundation connections have been found. displacement transducers was connected directly to the
foundation. Therefore, the average curvature evaluated
Although temporary bracing is needed for erection comprises the strain penetration of the reinforcing bars into
stability, grouted steel sleeves are an inexpensive type of the foundation block.
connection, especially if the retrofit after a seismic event
is considered. The confinement provided by grouted steel Figure 3 shows the geometry of the tested specimens.
sleeves inhibits buckling of the longitudinal reinforcement Specimen CP represented a typical cast-in-place con-
and increases the strength of the grout. As a result, the crete column-to-foundation connection, and specimen
damage associated with this kind of connection is localized PF represented a typical grouted pocket foundation with
to the column base, allowing easier postseismic column no extra roughening of the surfaces at the column and
repair compared with traditional connections. foundation interface. Specimens GS4 and GS4B both had
four grouted sleeves, but each specimen had a different
Another advantage of grouted steel sleeve connections anchorage length of the reinforcing bars in the foundation.
compared with modern mechanical splice devices is the The specimen GS4 had 90° hooks at the reinforcing bar
ability to include an unbonded portion in the reinforc- ends, whereas GS4B had straight reinforcing bars, which
ing bar inside the sleeves. This distributes the reinforcing represented a connection sometimes adopted to increase
bar deformation over the unbonded length, increasing the construction speed. Specimen GS4U was equal to speci-
ultimate rotation capacity of the subassembly and reducing men GS4B with an additional 300 mm (12 in.) unbonded
the reinforcement strain demand for a given column drift length on the spliced reinforcing bars. Adhesive tape pro-
(defined as the interstory displacement divided by the inter- vided the unbonding for these bars. GS4UR was specimen
story height). This advantage is not typical of mechanical GS4U after retrofitting. The retrofitting sequence consisted
splice devices. The unbonded length of the reinforcement of clearing the grout crushed during the previous tests at
can be taken into account in the design process by consid- the column base, placing collar formwork, and filling it
ering an appropriate plastic hinge length. with high-strength shrinkage-compensating grout.
The responses of five columns subjected to a cyclic top Material properties

horizontal displacement history were investigated consider-
ing different connection details. All reinforcing steel had a specified yield stress greater
than 450 MPa (65 ksi), but an average yield stress of
The cyclic behavior of a grouted corrugated steel sleeve 530 MPa (77 ksi) was recorded. The concrete type was
connection with partly unbonded bars in the sleeves was C35/45 (that is, with a specified 28-day cylindrical and cu-
investigated and compared with the cyclic behavior of the bic compressive strength greater than 35 MPa [5.1 ksi] and
same specimen after retrofitting. 45 MPa [6.6 ksi], respectively). All of the grout that was

Hollow-core plunger jacks
for column axial load
Electromechanical
screw jack
(1000 kN capacity)
Column-to-
foundation
Self-equilibrated
subassembly
reaction frame
Figure 1. Test setup. Note: 1 kN = 0.225 kip.
used inside the sleeves, retrofitted connection for GS4UR, for the 2nd order load-displacement effects P-Δ, indi-
and pocket foundation PF was a shrinkage-compensating cated by the dashed lines. Any gain or loss of strength is
cement-based grout with a maximum aggregate size of obtained when the experimental curve in absolute terms
2.5 mm (0.1 in.) and a specified 28-day strength greater lies above or below the P-Δ lines, respectively. Figure 5
than 75 MPa (11 ksi). The ducts used for the grouted sleeve shows the horizontal load versus drift comparison between
connections were typical post-tensioning ducts made with the unbonded grouted sleeve connection before and after
corrugated galvanized strip steel with an 80 mm (3 in.) retrofitting.
internal diameter, an 84 mm (3.3 in.) external diameter, and
a 0.6 mm (24-gauge) thickness. Figure 6 shows pictures of the damaged region at the
base column corresponding to 5% drift for all tests except
Test results GS4UR, which failed at approximately 3.5% drift.
Figure 4 shows the experimental horizontal load versus The three pairs of linear displacement transducers at the
drift curves for the tested specimens. The lateral load was column base side computed the column average curvature
reported as measured by the load cell with no correction at three consecutive levels (low, intermediate, and high) at

Typical geometry and reinforcing layout Typical section AA - grouted sleeves
400 mm
220 mm
Stirrups 8 mm diameter
160 mm
50 mm
70 mm
50 mm
70 mm
at 50 mm center to center
18 mm diameter
Precast
concrete
1780 mm
column C
L
3200 mm
C
L 26 mm diameter
240 mm
80 mm
80 mm
Corrugated
steel sleeves
Displacement transducers at column base
900 mm
A A 215 mm High
250 mm
Grout 215 mm Intermediate

at 50 mm center to center 215 mm Low
600 mm
125 mm
125 mm
150 mm
2000 mm
Figure 2. Reinforcement layout and instrumentation details. Note: 1 mm = 0.0394 in.
215 mm (8.46 in.) spacing. For each level, the average cur- load due to the P-Δ effect.
vature was computed by dividing the difference between
the measurements of the two displacement transducers by Analysis of results
the level spacing (215 mm) and by the distance between
the transducers (150 mm [5.9 in.]). The curvature com- As expected, all of the columns had almost the same
puted at the lowest level comprises the strain penetration maximum lateral force capacity, though some differences
of the reinforcing bars into the foundation block because arose due to misalignment of the reinforcing bars during
the displacement transducers were connected directly to construction.
the foundation. The moment-curvature curves (Fig. 7) are
the envelope curves for each test, computed as the aver- The cast-in-place concrete column specimen CP failed at
age of both directions of excitation. The abscissa is made 5% drift due to the tensile failure of one of the reinforcing
dimensionless by multiplying the curvature by the column bars. This is likely due to low-cycle fatigue of the bar and
cross-section height B (400 mm [16 in.]). localized strain at the column base where a major crack
opened (Fig. 6). The column cyclic behavior was stable
Figure 8 shows the dimensionless energy for each test in up to 2.5% drift, and no damage other than base concrete
the drift range from 1% to 5%. The dimensionless energy spalling was observed. Significant pinching was present in
is defined as the ratio between the hysteretic energy (area the cycles after 2.5% drift.
inside each complete force-displacement cycle) and the
corresponding elastic energy in a semicycle (defined as one The grouted pocket foundation specimen PF showed a sta-
half of the maximum displacement times the corresponding ble response though a small amount of strength degrada-
lateral force), which accounts for the equivalent horizontal tion was registered in one loading direction after 3% drift.

Specimen CP Specimen PF
Section A-A Section A-A
A A A A
400 mm
400 mm
80 mm 80 mm 80 mm 80 mm
240 mm 240 mm
Grout
Specimen GS4 Specimen GS4B

A A A A
400 mm
400 mm
80 mm 80 mm 80 mm 80 mm
240 mm 240 mm
Grout Grout
Specimen GS4U Specimen GS4UR

A A A A
300 mm Repair collar
400 mm
400 mm
unbonded
80 mm 80 mm 80 mm 80 mm
240 mm 240 mm
Grout Grout
Figure 3. Details of column-to-foundation connections. Note: 1 mm = 0.0394 in.
In the same loading direction fatter hysteresis half loops dimension.

were observed compared with the other loading direction.
This asymmetry was probably due to imperfect centering Specimens GS4 and GS4B showed a stable hysteretic
of the column in the pocket foundation. Compared with response, though specimen GS4 showed some strength
others, specimen PF showed a more pronounced extension degradation starting at 2% drift. This was likely due to
of flexural cracks at the column base (Fig. 6), spreading longitudinal reinforcement slippage in the foundation
along a length approximately equal to the column base block. Specimen GS4 resembled a connection sometimes

100 22.5 100 22.5
Specimen CP Specimen PF
Horizontal load, kip

50 50
Horizontal load, kN

Horizontal load, kN
0 0 0 0
-50 -50
P-Δ P-Δ
Bar failure
-100 -22.5 -100 -22.5

-6 -3 0 3 6 -6 -3 0 3 6
Drift, % Drift, %
100 22.5 100 22.5

Specimen GS4 Specimen GS4B
50 50
Horizontal load, kN
Horizontal load, kN

0 0 0 0
-50 -50
P-Δ P-Δ
-100 -22.5 -100 -22.5

-6 -3 0 3 6 -6 -3 0 3 6
Drift, % Drift, %
100 22.5 100 22.5

Specimen GS4UR
Specimen GS4U

50 50
Horizontal load, kN
Horizontal load, kN
0 0 0 0
-50 -50
P-Δ P-Δ
Bar failure
-100 -22.5 -100 -22.5

-6 -3 0 3 6 -6 -3 0 3 6
Drift, % Drift, %
Figure 4. Lateral load versus column drift for the different connection types. Note: P-∆ = 2nd order load-displacement effects.
adopted in practice to increase construction speed even a value over which considerable damage is expected in
if full anchorage strength of the splice reinforcement is nonstructural members, such as cladding panels.
not provided. As expected, the limited anchorage length
resulted in strength degradation in the cyclic response. This The use of grouted steel sleeves on the column induced a
degradation is negligible for drift values smaller than 2.5%, major crack opening at the grout layer beneath the sleeves.
The compressive strain concentration at the column base
led to progressive damage at the unconfined grout next to

the steel sleeves. After crushing, the damaged grout was 100 22.5
expelled from the column base, leaving the role of resist-
ing compressive forces associated with cyclic bending to
only the confined grout in the steel sleeves and the vertical 50
Horizontal load, kN

reinforcing bars. On the other end, grouted steel sleeve
specimens showed no other noticeable sign of damage.
0 0
Compared with traditional connections (specimens CP and

PF), a slightly more pronounced pinching was observed in -50
P-Δ
the cycles of GS4 and GS4B. This was due to the progres- GS4U
sive damage of the base grout layer, which led to a larger GS4UR
strain localization at the column base. For these specimens, -100 -22.5
-6 -3 0 3 6
the test was extended to 6.5% drift with no reinforcing bar
Drift, %
failure. The higher displacement capacity recorded was due
to the confinement induced by the steel sleeves on the grout.
Figure 5. Comparison of lateral load–column drift between unbonded and retro-
Although the localized strain at the base of specimens fitted grouted sleeve solution. Note: P-Δ = 2nd order load displacement.
GS4 and GS4B could lead to a failure of the columns,
the existence of confined grout regions within the sleeves
effectively prevented an early failure of the connections. • For all of the tests, the majority of the curvature de-
Furthermore, the grouted sleeves prevented buckling of the mand was concentrated at the column base as expect-
vertical bars anchored in the foundation. ed. The average curvature computed in the low level
comprises the strain penetration of the reinforcing bars
Because the damage was limited to the grout layer exist- into the foundation. In fact, the displacement transduc-
ing at the base and the remaining part of the column was ers were connected directly to the foundation block.
mostly undamaged, it is concluded that the grouted steel
sleeve connections are easier to repair after a seismic event • The curvature of specimen CP showed a greater local-
than the traditional connections (specimens CP and PF). ization with respect to that of specimen PF; this led to
the early bar failure observed in specimen CP.
The same observations and considerations made for speci-
mens GS4 and GS4B apply to specimen GS4U. The addi- • The curvature values were consistently higher for all
tion of an unbonded length in the reinforcement inside the of the grouted sleeve specimens than for the cast-in-
grouted steel sleeves further reduced the column damage. place and pocket foundation connections. This effect
The constant strain distribution in the unbonded region led once more demonstrates that a larger localized strain
to a rigid rotation at the column base, which reduced the occurred in the grouted sleeve specimens than in the
reinforcing bar strain required to reach the same column top cast-in-place and pocket foundation specimens.
displacement compared with specimens GS4 and GS4B.
• Among grouted steel sleeve connections, specimen
After reaching 5% drift, specimen GS4U was retrofit- GS4 had higher curvature values than GS4B because
ted and became specimen GS4UR. Figure 5 compares of bar slippage.
specimen GS4U with its retrofitted connection specimen
GS4UR. The cyclic response of the two specimens is simi- • The unbonded connection of specimen GS4U showed
lar up to 3% drift. At 3.5% drift, a bar failed in specimen the highest curvature values due to the elongation
GS4UR. This failure was likely due to low cycle fatigue associated with the constant strain distribution in the
of the reinforcing bar, considering the number of cycles unbonded region.
experienced by the specimen during both tests.
Regarding the energy dissipation for each test in the drift
From the moment-curvature envelope curves for each test range of 1% to 5% (Fig. 8), no significant differences were
(Fig. 7), the following observations were made: recorded among specimens CP, PF, GS4, and GS4B, while
specimens GS4U and GS4UR showed less energy dissipa-
• In the high level (Fig. 2), all of the specimens behaved tion because of the rotation at the column base associated
as linear elastic. with the constant strain in the unbonded region.
• In the intermediate level, specimen PF was the only

specimen to behave inelastically; this is in accordance
with the crack pattern in Fig. 6.

CP PF
GS4 GS4B
GS4U GS4UR 3.5%
Figure 6. Damage at 5% lateral drift. Specimen GS4U was repaired and renamed GS4UR; the latter experienced the failure of one reinforcing bar at a drift of 3.5%.

Considerations
on the unbonded length
As shown by the experimental results, providing an un-

bonded length in the longitudinal reinforcement inside the
grouted steel sleeves led to a better seismic performance of
the column-to-foundation connection compared with tradi-
tional connections. The longitudinal reinforcement buckling
is inhibited by the confinement provided by the grouted steel
sleeves, and the use of an unbonded length of reinforcement
leads to an increase in the column displacement capacity.
Furthermore, the damage caused by the superimposed cyclic
displacement history is limited to the grout layer existing at
the base, allowing easier connection repair after a seismic
event compared with the other connections analyzed.
Having proved the effectiveness of the unbonded length, the

next step is to determine how to take it into account in the
design process and how to evaluate the column curvature φ
and the curvature ductility μφ for different degrees of lateral
displacement. The value of μφ can be directly implemented
in some design procedures based on displacement control.7–9
To take the unbonded length into account, it is sufficient

to add its value to the plastic hinge length Lp, that is, the
length over which strain and plastic curvature are con-
sidered constant and equal to the maximum values at the
column base, according to a bilinear idealization of the
curvature distribution (Fig. 9). The new plastic hinge
length, which accounts for debonding, is calculated by add-
ing the unbonded length Lunb to a plastic hinge formulation
provided by Priestley et al.,9 leading to Eq. (1).
Lp = 0.2 f - 1 p Lc + Lsp + Lunb $ 2Lsp + Lunb (1)

fu
fy
where
fu = steel ultimate stress
fy = steel yield stress
Lc = column contraflexure length, corresponding in this

case to the column height H
Figure 7. Experimental moment and average curvature response for three dis-
placement transducer positions: low, intermediate, and high. Note: B = column
Lsp = strain penetration length, which takes into account the
cross-section height.
splice reinforcement tension strain distribution into
the foundation9
The approximations of the yield Δy and ultimate Δu col-
= 0.15 fy db if fy is in ksi umn displacements are obtained by double integration of
the idealized curvature distribution. In Δy evaluation, the
= 0.022 fy db if fy is in MPa rotation associated with the constant strain distribution in
the unbonded region needs to be added. In the case of Δu,
where this rotation is taken into account directly by the plastic
hinge length calculation proposed in Eq. (1).
db = reinforcing bar diameter

3
CP
PF
GS4
2.5
GS4B
GS4U
GS4UR
Ehysteretic / Eelastic
1.5
0.5
0
1.0 1.0 1.0 1.5 2.0 2.5 2.5 2.5 3.0 3.5 4.0 4.5 5.0
Drift, %
Figure 8. Dimensionless energy of the column-to-foundation connections. Note: Eelastic = elastic energy; Ehysteretic = hysteretic energy.
Beam Roof element
Column H
Mat foundation
LP
LSP
Фy Фp
Figure 9. Column-curvature idealization. Note: H = column height; Lp = plastic hinge length; Lsp = strain penetration length; øp = column plastic curvature;
øy = column yield curvature.

100 22.5
GS4U
2 GS4B
zy H
Δy = + zy Lunb H (2) Drift yield (GS4U)
Drift yield (GS4B)

3 75
Horizontal load, kN
Drift ultimate (GS4U)
Drift ultimate (GS4B)
where
50 0
φy = column yield curvature
Δu = Δy + Δp (3) 25
where
0 -22.5
Δp = column displacement associated with plastic hinge 0 2 4 6 8
rotation Drift, %
= (φu − φy) LpH (4) Figure 10. Yield and ultimate drift for the grouted sleeve solution with and
without debonding.
where
φu = column ultimate curvature curs in the column outside the base section. Due to the
damage localization observed, the postseismic column
Equations (2) to (4) were applied to find the yield and repair for the grouted sleeve connection is simpler than
ultimate drift associated with specimens GS4B and GS4U, that for cast-in-place concrete or pocket foundation con-
that is, the grouted steel sleeve connection with and with- nections and can be effectively done by replacing the base
out the unbonded region. Figure 10 shows the performance layer of crushed grout with new high-strength grout. To
enhancement in terms of displacement capacity increase. further reduce the column damage and increase the column
The adoption of an unbonded length of reinforcement leads lateral displacement capacity, an unbonded length of steel
to a reduction of the curvature demand at a given imposed reinforcement can be used within the grouted sleeves. The
lateral displacement and therefore to a reduction of the proposed equations make it possible to take into account
column damage that is directly related to the curvature the unbonded length of reinforcement in the design process
demand. and to evaluate the column yield and ultimate displacement
when unbonded reinforcing bars are used.
In the case of debonding, the ultimate reinforcing bar
capacity is ensured, providing an appropriate develop- Acknowledgments
ment length in addition to the unbonded length. The
use of grouted steel sleeves contributes to reducing the The experimental program reported in this paper was part
development length due to the sleeves’ ability to restrain of a research program on precast concrete column-to-foun-
the splitting cracks in the grout rather than providing pas- dation connections financed by Moretti SpA in Erbusco,
sive confinement. Although the aim of the research is not Italy.
the determination of the bond strength associated with
grouted steel sleeves, the use of the sleeves, especially with The cooperation of Cristian Ratti, Andrea Zini, and Nicola
debonding, resulted in significant column lateral displace- Barzi in setting up the reaction frame and conducting the
ment without reinforcement bar failure. experimental tests is acknowledged.
Conclusion References
The experimental program showed that grouted steel 1. fib (International Federation for Structural Concrete)
sleeves are suitable as column-to-foundation connections Commission 7. 2003. Seismic Design of Precast
in seismic regions. The high ductility of the grouted steel Concrete Building Structures, International Federa-
sleeve connections is related to the confining effect of the tion for Structural Concrete. Bulletin 27. Lausanne,
corrugated steel sleeves on the grout. Furthermore, the Switzerland: fib.
presence of a highly confined grout prevents buckling of
the longitudinal reinforcement. 2. PCI Connection Details Committee. 2008. PCI Con-
nections Manual for Precast and Prestressed Concrete
In such connections, the damage was localized to the Construction. 1st ed. Chicago, IL: PCI.
20 mm (0.8 in.) grout layer between the precast concrete
column and the foundation. As a result, little damage oc-

3. Elliott, K. S. 2005. Precast Concrete Structures. Ox- P = axial load
ford, Great Britain: Elsevier Butterworth-Heinemann.
P-Δ = 2nd order load-displacement effects
4. Blandón, J. J., and M. E. Rodríguez. 2005. Behavior
of Connections and Floor Diaphragms in Seismic- Δp = column displacement associated with plastic
Resisting Precast Concrete Buildings. PCI Journal, V. hinge rotation
50, No. 2 (March–April): pp. 56–75.
Δu = column ultimate displacement
5. Martins, R., F. Canha, M. K. El Debs, K. de Borja Jag-
uaribe Jr., and A. L. Homce de Cresce El Debs. 2009. Δy = column yield displacement
Behavior of Socket Base Connections Emphasizing
Pedestal Walls. ACI Structural Journal, V. 106, No. 3 μφ = curvature ductility
(May–June): pp. 268–278.
φ = column curvature
6. Osanai, Y., F. Watanabe, and S. Okamoto. 1996. Stress
Transfer Mechanism of Socket Base Connections with φp = column plastic curvature
Precast Concrete Columns. ACI Structural Journal, V.
93, No. 3 (May–June): pp. 1–11. φy = column yield curvature
7. Priestley, M. J. N. 2002. Direct Displacement-Based φu = column ultimate curvature

Design of Precast/Prestressed Concrete Buildings.
PCI Journal, V. 47, No. 6 (November–December): pp.
66–79.
8. Priestley, M. J. N., and D. N. Grant. 2005. Viscous

Damping, in Seismic Design and Analysis. Journal
of Earthquake Engineering, V. 9, Special Issue 2: pp.
229–255.
9. Priestley, M. J. N., G. M. Calvi, and M. J. Kowalsky.

2007. Displacement-Based Seismic Design of Struc-
tures. Pavia, Italy: IUSS Press.
Notation
B = column cross-section height
db = reinforcing bar diameter
Eelastic = elastic energy
Ehysteretic = hysteretic energy
fu = steel ultimate stress
fy = steel yield stress
H = column height
Lc = column contraflexure length
Lp = plastic hinge length
Lsp = strain penetration length
Lunb = unbonded length

About the authors The results demonstrate that grouted sleeves ensure
a ductility and energy dissipation capacity similar
Andrea Belleri, PhD, is a post- to those of traditional connections. The confinement
graduate fellow at the Department provided by the grouted sleeves inhibits buckling
of Design and Technologies at of the longitudinal reinforcement and increases the
the University of Bergamo in compressive strength of the grout. As a consequence,
Bergamo, Italy. the damage associated with this kind of connection is
localized to the column base, allowing easier postseis-
mic column repair compared with traditional connec-
Paolo Riva, PhD, is a full profes- tions. The comparison of the experimental response
sor in the Department of Design of a grouted sleeve connection with partly unbonded
and Technologies at the University reinforcement within the sleeves and the same speci-
of Bergamo. men after retrofitting is also included.
Keywords
Connection, debonding, grouted sleeve, pocket founda-

Abstract tion, seismic retrofit.
The paper investigates the suitability of grouted sleeve Review policy

connections as column-to-foundation connections for
precast concrete structures in seismic regions. Al- This paper was reviewed in accordance with the
though grouted sleeves are commonly used, the typical Precast/Prestressed Concrete Institute’s peer-review
advantages from this connection have not been fully process.
addressed in the literature.
Reader comments
Experiments on the cyclic behavior of column-to-
foundation subassemblies compared the response of Please address any reader comments to journal@pci
grouted sleeve connections with cast-in-place and .org or Precast/Prestressed Concrete Institute, c/o PCI
pocket-foundation connections. Journal, 200 W. Adams St., Suite 2100, Chicago, IL
60606. J

Structural behavior
of an innovative all–precast
concrete dual system
for residential buildings
T
he performance of a structure is mainly linked to its
strength, stiffness, and ductility, although function-
ality, aesthetics, speed of construction, and econo-
my are also significant.
The following functional requirements should be taken

into consideration:
• Long spans with floor space uninterrupted by columns

or structural walls give users maximum flexibility.
• A maximum floor-to-floor height allows adequate

spaces and a reduced overall building height.
• Open-space apartments or offices (adaptable floors)

■ A new all–precast concrete system was used in Romania to can accommodate future tenant alterations with maxi-
construct a residential building. mum speed and minimum disruption or cost.
■ To validate the quality and performance of this type of structure, It is well known that the economic performance of a struc-
an extensive program of theoretical analyses and structural tural system greatly depends on the quantities of materi-
tests (including shake table tests) was conducted. als per square meter (for example, concrete, reinforcing
bars, and strands), but the real construction cost can also
■ Due to its functional and technical advantages, high productiv- depend on speed of construction, local market conditions,
ity, and low cost, the building system has been used in more availability of labor and specialized equipment, and other
than 500 flats in Transylvania, Romania. factors.

Flat-slab structures in seismic zones mainly consists of long precast concrete columns—a unit
is from 5.40 m (18 ft) to 11.40 m (37.4 ft) long—and pre-
Performance characteristics related to building functional- cast concrete flat-slab panels with spans of 3.60 m (12 ft)
ity, aesthetics, and economy can be encountered in the case to 6.00 m (20 ft).
of structures made of slabs and columns, the so-called typi-
cal flat-slab structures. These structures carry vertical and The promotion of such a system in seismic areas was
lateral forces and achieve their lateral stability by frame conditioned by the use of a limited number of structural
action generated by the interaction of columns and slabs. walls in the structure, leading to a dual structural system of
the frame-wall type (Fig. 1). Figure 2 shows this innova-
In many cases, such as multistory buildings or structures tive precast concrete structure during erection in the city of
in seismic areas, pure flat-slab structures have to be altered Deva, Romania.
by the addition of vertical elements, such as shear walls or
steel bracing systems, leading to dual systems. The equiva- In Europe, Romania is considered to have a high level
lent frames, comprising slabs and columns, mainly ensure of seismic risk, where earthquakes over 7 on the Richter
the transfer of gravity loads and a portion of the seismic scale occur quite frequently. Bucharest is the most vulner-
loads, while the structural wall system provides the lateral able capital in Europe.3 The seismic response to the accel-
resistance and stiffness of the building1 so that such a dual erograms recorded in Bucharest during the 1977 Vrancea
structure can be successfully used even in areas with high earthquake (magnitude of earthquake on the moment
seismic activity. magnitude scale Mw equal to 7.5 with 0.2g peak ground
acceleration, where g is acceleration due to gravity) shows
To improve the competitiveness of such a structure type uncommonly high displacement demands. Extensive non-
and considering constructability as a powerful tool for re- structural damage during this earthquake was reported due
ducing the cost and the time of construction, the designers’ to large lateral displacements. For this reason, the Roma-
interest has focused on a standard precast concrete struc- nian Seismic Design Code P 100-924 limits the allowable
ture using fewer elements with standard dimensions that interstory drift to 0.35% for buildings with infill walls
can be produced in large numbers on an industrial basis. interacting with the frame structure when the structure is
checked in the serviceability limit state (SLS). The new
Objectives Romanian Seismic Design Code P 100-1/20065 introduced
a more relaxed value of 0.5% for allowable interstory drift
In an attempt to solve the difficult problem of the con- for verifications in the SLS, comparable to that specified
struction of residential buildings in Romania, taking into by FEMA 356/20006 for shear wall structures when imme-
consideration the previously mentioned factors and require- diate occupancy structural performance level is considered.
ments, a design team from the Design Institute of Hunedo-
ara in Romania, led by Eusebiu Tripa, proposed a new all– Precast concrete columns
precast concrete system of a dual flat-slab structure type.
Columns are precast concrete elements with variable
A complex program, including theoretical studies and lengths, a function of the structure height. Columns may
many experimental tests,2 preceded the national-scale extend over one or more levels depending on technologi-
promotion of this system. Academics from Romanian cal, strength, and cost criteria. Typically, the joints between
civil engineering faculties in Jassy, Cluj, Timis̹oara, and columns and slabs are not filled with concrete, and the
Bucharest; researchers from INCERC-Bucharest (National longitudinal reinforcement in the columns is continuous
Building Research Institute) and from the INCERC depart- (Fig. 3). The reinforcement resists the loads generated
ments of Timis̹oara, Jassy, and Cluj; architects; and civil during the transport, handling, and assembling opera-
engineers participated in this program. tions when the concrete is missing in the joint zone of the
column.
This paper presents the main characteristics of the new build-
ing system, the principal features of the experimental program The analytical and experimental studies performed by the
that preceded its launch on the market, and the conclusions authors2 on up to 14.00-m-long (46 ft) precast concrete
concerning the structural performance of the system. columns have shown that the most efficient solution lies in
the precasting of 9.00-m-long (30 ft) columns, which cor-
Precast concrete respond to three-story buildings.
dual flat-slab structure
In taller buildings, more members are assembled to form
Structure characteristics a column. The connection between the precast concrete
members is of a bayonet type: the reinforcing bars from
The structural system presented in this paper is dedicated the upper member are inserted into the sheaths at the end
to residential buildings with up to nine stories, and it of the lower column (Fig. 3).

Figure 1. Types of elements in the precast concrete dual flat-slab structure.
Precast concrete floors
The floors are made of large precast concrete flat-slab pan-

els supported at the corners by mechanical devices fixed on
the columns.
The spacing between the columns, usually ranging from

3.00 m (10 ft) to 6.00 m (20 ft), dictates the span of precast
concrete slab panels; the floor panel is of the ribbed flat-
slab type (Fig. 4). The continuity of the precast concrete
floor is achieved by narrow cast-in-place concrete strips re-
inforced by longitudinal bars and horizontal loops (Fig. 4).
Figure 2. All–precast concrete dual structure during erection.

Figure 3. Connection between column members. Note: 1 mm = 0.0394 in.
Precast concrete shear walls panels spanning between adjacent columns. Through their
horizontal connections, the panels provide the structural
The shear wall comprises large reinforced precast concrete wall continuity along its height. Through their vertical
Figure 4. Precast concrete flat-slab panel and connections between the panels.

Figure 5. Vertical column–wall panel connection.
connections (column-panel joints) they make up a dual between precast concrete members should be determined
structural system, combining frames with structural walls by analyses or by tests.”7
(Fig. 2 and Fig. 5).
In addition, section 21.2.1.5 says that a precast concrete
Experimental program structural system “may be used for earthquake resistance if
it is demonstrated by experimental evidence and analysis
Objectives of the program that the proposed system will have strength and toughness
equal to or exceeding those provided by a comparable
The structural behavior of a system that incorporates pre- monolithic reinforced concrete structure.”7
cast concrete members may differ substantially from that
of a comparable monolithic cast-in-place system. The pre- To assess the structural quality and the performance of this
cast concrete members should be assembled and connected structural system designed for seismic zones, the research
to produce a structural system capable of resisting vertical team conducted an extensive analytical and experimental
and lateral forces due to shrinkage, creep, temperature study. The research program2 included tests on column-
changes, elastic deformation, wind, or earthquake. to-column joints, experimental tests on a precast concrete
structural wall subassemblage, postelastic computer analy-
The design and detailing of the connections to transmit sis of precast concrete structural walls, tests on a full-scale
such forces are especially important to performance. assemblage of columns and precast concrete flat slabs
Connections are essential to the integrity of the structural subjected to gravity forces, shake table tests on a struc-
system. ture prototype (scale 1:4), and in-place dynamic tests on
a real five-story building. In this large scientific program,
Taking into consideration the novelty of the proposed researchers from three universities and four institutes
system and the overall behavior of the precast concrete and structural designers from two design institutes were
structure, “the adequacy of connections to transfer forces involved.

Figure 6. Full-scale precast concrete structure under gravity loads.
Flat-slab structure ally good behavior of the structure under vertical loads and
under gravity loads underlines the large load-bearing capacity of the structure.
The full-scale test of an assemblage of precast concrete The shape of the gravity load–vertical deflection P–Δv
columns and slabs was performed to observe the behavior diagrams reveals an elastic behavior of the structure with
of the precast concrete flat-slab panels, the connections 26.80% to 39.70% permanent irreversible deformations
between panels, and the column-to-slab connections under and 60.30% to 73.20% elastic recovery (Fig. 7).
gravity loads. The behavior up to the failure of the entire
floor system and the problem of slab punching were also The maximum vertical displacement Dvmax of 7.3 mm
studied. (0.29 in.) does not exceed the allowable vertical displace-
ment Dva of L/350 equal to 9.3 mm (0.36 in.) given by
The assemblage comprised eight square precast concrete the Romanian design code for concrete structures STAS
slab panels of 3.60 m (12 ft) each and 15 square columns 10107/0-90.11 The monolithic connections between the
of 350 mm (14 in.) (Fig. 6). slab precast concrete panels, which have 8 mm loops at
150 mm (0.3 in. loops at 6 in.) spacing and longitudinal
The authors used the equivalent frame method accepted reinforcing bars, ensure complete integrity of the precast
by many design codes, including the American Concrete concrete floor, including its flexural and shear integrity.
Institute (ACI 318-08),7 European (EN 1992-1-1:2004),8
and Romanian standards (SR EN 1992-1-1:20049 and Dual precast concrete
STAS 10107/1-90).10 The designers calculated theoretical structural walls’ subassemblage
values for cracking load P crd of 5.50 kN/m2 (115 lb/ft2) and under lateral loading
ultimate gravity load P ud of 7.70 kN/m2 (160 lb/ft2).
The experimental program focused on the general behav-
The strains in the concrete and reinforcing bars, the verti- ior of the precast concrete structural walls under static
cal displacement of the precast concrete panels, and the alternate-reversed loadings in all ranges—elastic, crack-
displacements of the connections between the slab panels ing, postelastic, and failure—as well as the behavior of the
were recorded. Figure 7 shows a typical applied load-dis- vertical connections, horizontal connections, and joints
placement diagram recorded during the test (Fig. 6) at the between precast concrete column members.
midspan of the panel.
The tested subassemblage was designed to model at scale
The first cracks developed at 4.40 kN/m2 (92 lb/ft2) in the 1:2.5 the structural behavior of a real precast concrete
longitudinal edge connections and at 9.90 kN/m2 (210 lb/ structural wall with eight stories. The unit’s components
ft2) in the middle of the panel. The experimental ultimate are monolithic concrete foundations under each shear wall,
gravity load P uexp of 15.90 kN/m2 (332 lb/ft2) was reached precast concrete columns of two or three stories for each
in the last loading stage and corresponded to the initiation member, precast concrete shear wall panels on the height
of slab punching in the vicinity of columns. The ratio of of a floor, and precast concrete slab panels (Fig. 8).
2.10 between the experimental and the theoretical (design)
values of the ultimate load ( P uexp /P ud ) confirms the gener- The loading of the model was accomplished by gravity

Figure 7. Gravity load–vertical displacement P–Δv recorded at the middle of the panel. Note: Δ vmax = experimental maximum vertical deflection.
1 mm = 0.0394 in.; 1 daN/m2 = 0.208 lb/ft2.
loads Ng (dead load) so that the same compressive stress The large subassemblage, measuring 2.02 m × 1.62 m ×
as in a real eight-story shear wall could be reached at each 5.95 m (6.62 ft × 5.31 ft × 19.50 ft), was loaded vertically
level. and horizontally (Fig. 8). The principal observations and
conclusions include the following:
The vertically applied forces N1 (Fig. 8) compensate in
the model for the effect of the reduction of the upper three • The cracking extended only along the height of three
floors of the real structural wall. These adjustments from stories. Although it was all–precast concrete, the
the real structure to the model did not alter the results structural wall subassemblage (column–shear wall
of the experiment. On the lower third of the shear wall, panels–column) behaved at cracking as a monolithic
the part of maximum interest, the bending moments are structural wall, the cracks of the columns being contin-
practically equal and the shear forces are generally slightly uous through connections and through precast concrete
higher than in the real structure. panels. The cracks are typical of bending.
Previous tests (Maniu et al.)12 had shown that the cracking • The main cracks that finally led to failure developed
does not extend beyond the lower third of the structural only after the imposed 3Δy cycle (Fig. 9). In the
wall. The horizontal load H was applied in alternate cycles monolithic area (vertical connections), the cracks were
following the principle of imposed displacements, where narrow, having a width under 0.20 mm (0.0078 in.)
Δ is the top lateral displacement of the experimental model throughout the test (Fig. 10).
and Δy is the top lateral displacement at the first yield of the
steel reinforcement (Fig. 9). • The column-to-column joint behaved well. The cracks
were insignificant in two of the joints and totally absent
in the other two joints.

• Failure of the subassemblage occurred because of the
concrete spalling and crushing in compression and
because of the failure of a few column reinforcing
bars that, after 26 cycles, buckled locally between
the hoops (Fig. 10). The predicted ultimate capacity
was 2% smaller than the experimental one, 360 kN-m
compared with 354 kN-m (266 kip-ft compared with
261 kip-ft).
• The measured lateral load–top displacement hysteretic

loops (Fig. 11), by their convex aspect, showed a be-
havior typical of long shear walls mainly subjected to
the bending moment. There was no pinching in the last
cycles, demonstrating the integrity of the connection
and, practically, the absence of slipping between the
precast concrete elements (column to panel).
• The maximum values of the joint slipping attained in

the cycle 4Δy did not exceed 0.30 mm (0.012 in.) in the
vertical column-to-panel joint and was only 0.10 mm
(0.0039 in.) in the horizontal panel-to-panel joint.
• The area enclosed by the hysteretic loops indicated large

energy absorption. The loops were stable. The stiffness
reduction, after three or four cycles under the same im-
posed displacement, did not exceed 10% (Fig. 11).
• The experimental displacement corresponding to

the design code force of 52.8 kN (11.9 kip) was in
good agreement (5% higher) with the theoretical one.
The experimental ultimate top lateral displacement Figure 8. Experimental model of loading conditions. Note: H = horizontally
applied load on model; N1 = vertically applied forces on model; Ng = gravity load
recorded in the last failure cycle Dexpmax
of 112.3 mm (dead load) on model.
(4.421 in.) was 26% smaller than the design ultimate
top lateral displacement Ddmax of 153.0 mm (6.02 in.).
This ratio indicates that in reality, the lateral rigidity of The dual precast concrete structure consisted of eight shear
the precast concrete model, after 27 cycles of alternate walls and 24 frame columns, interconnected by five precast
loads, is greater than the stiffness assumed in design. concrete flat slabs (Fig. 12).
• The lateral stiffness of the experimental specimen The dynamic response of this precast concrete dual system
was remarkable. After 23 cycles of loading, the lateral was established during four static tests, five tests to deter-
interstory drift measured at the uppermost level in the mine the dynamic characteristics of the structure, and 17
4Δy cycle was only 1/205. dynamic tests.
• The ultimate displacement ductility factor Δu/Δy (where The dynamic tests took 900 seconds in all and had about
Δu is ultimate top lateral displacement) attained dur- 70 acceleration peaks a0max of more than 1 m/sec2 (3 ft/
ing the test was 5.25. The structural subassemblage sec2) from those 1200 peak accelerations applied. The real
behaved well from this point of view, too. precast concrete structure and the experimental prototype
were designed for a seismic intensity of VIII on the modi-
Shake table tests fied Mercalli intensity scale (MM scale).
on a five-story prototype
The experimental program included dynamic tests with an
The most comprehensive experimental program was the artificial Housner-Jennings type earthquake (a0max of 1.108
testing of a 1:4 scale model of a five-story building on a ÷ 2.632 m/sec2 [3.635 ÷ 8.635 ft/sec2]); 1977 Vrancea,
140 kN (315 kip) shake table (Fig. 12). The model struc- Romania, earthquake (a0max of 2.775 ÷ 3.885 m/sec2 [9.104
ture was tested in the transverse direction. ÷ 12.75 ft/sec2]); 1940 El Centro, Calif., earthquake (a0max
of 2.775 m/sec2 [9.104 ft/sec2]); artificial Jassy, Roma-
nia, earthquake (a0max of 4.163 ÷ 6.383 m/sec2 [13.66 ÷

Figure 9. Loading history. Note: Δy = top lateral displacement at first yield. 1 mm = 0.0394.
20.94 ft/sec2]); and harmonic type earthquake (a0max of The most significant results and conclusions are as follows:
1.943 m/sec2 [6.375 ft/sec2]). The shake table tests in-
creased in intensity from V to X on the MM scale. • The static tests intermingled among the dynamic tests
did not produce damage in the experimental model.
Figure 13 presents lateral displacements recorded for
each floor, where A corresponds to the elastic stage,
B was conducted after cracking, C corresponds to
the elastic-plastic stage when the columns and shear
panels have open cracks, and D was performed in the
presence of plastic zones in the structure. By compar-
ing the static deformations recorded at different stages
of structural behavior (A, B, C, and D), a lateral stiff-
ness degradation of 12.50% was observed in the final
failure stage (Fig. 13).
• Cracks appeared, propagated, and widened only in the

vertical elements (columns and shear panels) at the
bottom of the first floor. Horizontal cracks appeared
at intensity level VIII on the MM scale, reaching their
maximum width of 1 mm (0.04 in.) in the last test,
which corresponds to the seismic intensity degree of X
on the MM scale. The lateral interstory drifts recorded
during the tests are: 1/600 to 1/400 at the seismic
intensity of VIII, 1/275 to 1/125 at IX, and 1/122 to
1/80 at X on the MM scale. It must be emphasized
that the prototype designed for the seismic intensity
of VIII on the MM scale satisfies the requirements of
the Romanian Seismic Design Code P 100-1/20065 in
Figure 10. Precast concrete shear wall after failure.

Figure 11. Lateral load–top lateral displacement envelope curves. Note: H = horizontal applied load on model; Δc = cracking displacement; Δy = top lateral displace-
ment at first yield. 1 mm = 0.0394 in.; 1 kN = 0.225 kip.
the ultimate limit state, where the allowable interstory spans of 3.60 m (11.8 ft), ten bays of 3.90 m (12.8 ft), and
drift is 2.5%, even for earthquakes of X on the MM five stories of 3.00 m (9.84 ft) high.
scale, where the measured maximum interstory drifts
are of 0.82% to 1.25%. Comparisons have been made between the dynamic char-
acteristics of a precast concrete structure comprising only
• The tests to determine the dynamic characteristics structural columns, panels, and walls (the first stage) and
within five different stages of structural behavior, from
the elastic to the plastic stage, showed that the natural
frequency of the model diminished from 6.50 Hz to
5.00 Hz in the transverse direction and from 6.00 Hz
to 5.03 Hz in the longitudinal direction. At the same
time, the critical damping ratio increased from 1.38 to
3.10 in the transverse direction and from 1.00 to 3.02
in the longitudinal direction. Despite reductions of fre-
quency by 23% transversally, 16% longitudinally, and
29% in torsion, the stiffness degradation after a series
of high-intensity seismic actions was only 12.50%,
which shows that the structure can be consolidated and
its structural integrity can be restored.
In-place dynamic test

on a real structure
Working with experts from INCERC-Bucharest, the au-

thors coordinated an in-place investigation to determine the
dynamic characteristics of a real structure that has three
Figure 12. Five-story precast concrete structure (model scale 1:4) during shake
table tests.

those of the second testing stage, when the structure also
had nonstructural walls (partitions) mounted.
A large volume of experimental data related to the struc-

tural response under five different types of dynamic excita-
tions was recorded, and detailed conclusions are presented
in the author’s PhD thesis.2
The main observations are as follows:
• The nonstructural walls significantly increased the

stiffness and damping capacity of the building. Thus,
the natural period of vibration decreased from 0.63 sec
to 0.32 sec in the longitudinal direction and from
0.40 sec to 0.32 sec in the transverse direction.
• The damping capacity of the building increased due to

the contribution of nonstructural elements, and, conse-
quently, the critical damping ratio increased from 0.01
to 0.05 in the horizontal longitudinal direction.
• The general torsion of the structure is beneficially in-

fluenced when nonstructural elements are introduced;
the structure becomes stiffer, and the natural period
Figure 13. Lateral displacements–static horizontal load diagrams. Note: Test A
of vibration decreases significantly from 0.44 sec to
corresponds to the elastic stage, test B was conducted after the cracking, test C
corresponds to the elastic-plastic stage, and test D was performed in the pres- 0.25 sec.
ence of plastic zones in the structure. F = static horizontal load = 6 kN.
1 mm = 0.0394 in.; 1 kN = 0.225 kip. The natural periods of vibration in place determined on
this new type of dual precast concrete structure are simi-
Figure 14. Residential condominiums

lar to the values determined in Bucharest on reinforced hibited good behavior during the 1977 Vrancea earthquake
concrete frame structures with infill masonry panels; those (Mw of 7.5).
structures exhibited good behavior during the March 4,
1977, Vrancea earthquake (Mw of 7.5). Due to its functional and technical advantages, high pro-
ductivity, and low costs, the constructive system has been
Conclusion used in more than 500 flats in Transylvania, Romania.
Three hundred apartments are in five- to nine-story build-
In an attempt to solve the difficult problem of the construc- ings (Fig. 14), and 200 are in residential condominiums
tion of residential buildings in Romania, the design team that have to five stories.
proposed a new all–precast concrete system of a dual flat-
slab type. References
To assess the quality and performance of this system, an 1. Crainic, L. 2003. Reinforced Concrete Structure. Cluj-
extensive program of theoretical analyses, structural tests, Napoca, Romania: Napoca Star.
and applied studies to improve the construction technology
were conducted. These studies maximized the effect on 2. Tripa, E. 2002. Contributions to the Study, Design and
construction productivity and minimized the total cost. Erection of Buildings with Precast Concrete Mixed
Structure. [In Romanian.] PhD thesis. University Po-
A full-scale test on an assemblage of precast concrete litehnica from Timis̹oara (UPT), Timis̹oara, Romania.
columns and precast concrete flat slabs confirms the gener-
ally good behavior of the structure under gravity loads and 3. Lungu, D., C. Arion, S. Demetriu, and A. Aldea. 2003.
underlines the large load-bearing capacity of the structure. Probabilistic Zonation of Vrancea Seismic Hazard.
The ratio between the experimental and the theoretical Eurocode 8, Representation of Design Actions. In Pro-
values of the ultimate load is 2.10. ceedings of the International Conference: Construc-
tions 2003, July 16–17, 2003. Cluj-Napoca, Romania:
The large subassemblage of two precast concrete, five- Napoca Star and Argonaut.
story structural walls; four precast concrete columns; and
five precast concrete flat slabs was tested up to failure. The 4. Ministry of Public Works and Territory Management
loading consisted of constant vertical forces and horizontal (MLPAT). 1992. Seismic Design Code (P100-92). [In
forces applied in alternate cycles following the principle Romanian.] Bucharest, Romania: MLPAT.
of imposed displacements. The test on this model (scale
1:2.5) confirmed its structural performance with respect to 5. Ministry of Transportation, Construction and Tourism
the load-bearing capacity, stiffness, ductility, and cracking. (MTCT). 2006. Seismic Design Code—Part I: Provi-
The experimental values regarding the ultimate capacity and sions for Buildings (P 100-1/2006). [In Romanian.]
lateral displacements are in good agreement with the theo- Bucharest, Romania: MTCT.
retical ones. The hysteretic loops (H–Δ) are stable, and the
prototype demonstrated a large energy absorption capacity. 6. Federal Emergency Management Agency (FEMA).
2000. Prestandard and Commentary for the Seismic
The types of connections proved suitable, resulting in a Rehabilitation of Buildings. FEMA 356. Washington,
behavior for the precast concrete subassemblage that was DC: FEMA.
similar to that of the monolithic solution.
7. American Concrete Institute (ACI) Committee 318.
The dynamic tests carried out on the 140 kN (315 kip) 2008. Building Code Requirements for Structural
shake table at INCERC-Jassy, where a five-story precast Concrete (ACI 318-08) and Commentary (ACI 318R-
concrete structure (scale 1:4) was tested, showed that 08). Farmington Hills, MI: ACI.
the lateral stiffness degradation of the structure was only
12.5% in the final failure stage. The tests took 900 seconds 8. European Committee of Standardization (CEN). 2004.
and had 70 acceleration peaks of more than 1 m/sec2 EN 1992-1-1: 2004. In Eurocode 2: Design of Con-
(3 ft/sec2). The interstory drifts recorded during the tests crete Structures—Part 1-1: General Rules and Rules
satisfy the requirements of the Romanian Seismic Design for Buildings. Brussels, Belgium: CEN.
Code P 100-1/2006.5
9. Romanian Standards Association (ASRO). 2004. SR
In-place tests on a real structure erected in the city of Deva EN 1992-1-1: 2004. In Eurocode 2: Design of Con-
indicated that the natural periods of vibration of this new crete Structures—Part 1-1: General Rules and Rules
type of precast concrete structure are similar to the values for Buildings. [In Romanian.] Bucharest, Romania:
determined in Bucharest on reinforced concrete frame ASRO.
structures with infill masonry panels. These structures ex-

10. Romanian Standards Institute (IRS). 1990. Reinforced Δv = vertical displacement (deflection)
and Prestressed Concrete Floors. General Design
Specifications. In Civil and Industrial Buildings. STAS Dva = a llowable vertical displacement (deflection) speci-
10107/1-90. [In Romanian.] Bucharest, Romania: IRS. fied by design code
11. IRS. 1990. Design and Detailing of Concrete, Rein- Dvmax = experimental maximum vertical deflection
forced and Prestressed Concrete Structural Members.
In Civil and Industrial Buildings. STAS 10107/0-90. Δ c = cracking displacement
[In Romanian.] Bucharest, Romania: IRS.
Δu = ultimate top lateral displacement
12. Maniu, H., A. M. Ioani, and E. Tripa. 1992. Structural
Investigation on a Seismic Performance of Precast Δu/Δy = displacement ductility factor
Mixed Shear Walls. In Proceedings of the Tenth World
Conference on Earthquake Engineering, Madrid, July Δy = top lateral displacement at first yield
19–24, 1992. Rotterdam, Netherlands: Balkema.
Notation
a0max = maximum values of accelerations (acceleration

peaks) applied during dynamic tests
F = static horizontal load applied during shake table

tests
g = a cceleration due to gravity = 9.81 m/sec2

(32.2 ft/sec2)
H = horizontal applied load on model
L = length (span) of precast concrete slab panels
Mw = magnitude of earthquake on the moment magnitude

scale (MMS)
N1 = vertically applied forces on model
Ng = gravity load (dead load) on model
P = gravity load applied on the full-scale assemblage
P crd = design (calculated) cracking load for the full-scale

assemblage
P ud

= design (calculated) ultimate gravity load for the
full-scale assemblage
P uexp = experimental (measured) ultimate gravity load ap-

plied on the full-scale assemblage
Δ = top lateral displacement
Ddmax = d esign (calculated) maximum (ultimate) top lateral

displacement
Dexp
max
= e xperimental maximum (ultimate) top lateral dis-
placement

About the authors analyses and structural tests (including shake table tests)
was conducted. The tests demonstrated good behav-
Adrian M. Ioani, PhD, FACI, ior of the precast concrete dual system (frame-wall)
FAICPS-Romania, is a professor of adapted for a classic flat-slab structure. The connections
structural mechanics at the resulted in a similar behavior of the precast concrete
Technical University of Cluj-Nap- system to that of the monolithic solution. Due to its
oca in Romania and former functional and technical advantages, high productivity,
director of INCERC-Cluj (National and low cost, this system has been used in more than
Building Research Institute). 500 flats in Transylvania, Romania.
Keywords
Eusebiu Tripa, PhD, P.E., FA-
ICPS-Romania, is a senior Connection, dual system, flat-slab panel, loading,
structural engineer at Tripexpert seismic.
Design Office in Deva, Romania.
Review policy
This paper was reviewed in accordance with the

Precast/Prestressed Concrete Institute’s peer-review
Abstract process.
This paper discusses a new all–precast concrete system Reader comments

used in Romania to construct a residential build-
ing. Designed for constructability, a new all–precast Please address any reader comments to journal@pci
concrete system comprising columns, flat slabs, and .org or Precast/Prestressed Concrete Institute, c/o PCI
structural walls was proposed by the design team. To Journal, 200 W. Adams St., Suite 2100, Chicago, IL
validate the structural quality and performance of this 60606. J
type of structure, an extensive program of theoretical

Architectural precast
concrete panel systems used
for lateral-force resistance
J. Paul Hobelmann, Macarena Schachter, and Matthew C. Cooper
E
xterior enclosure systems for buildings often
consist of elements that are supported by the build-
ing structure. For high-rise buildings, support of
the exterior enclosure by the building structural frame is
necessary, but for low-rise buildings, self-support of the
exterior enclosure may be possible and could result in cost
savings in the structural frame due to the reduced demand
on its members. For self-supported systems, the structural
frame typically provides lateral support at each floor and
the foundation system supports the enclosure gravity loads
directly.
For both frame-supported and self-supported enclosure

systems, deformation compatibility between the enclosure
system and the structural frame must be considered to
prevent performance problems or failure of the exterior
enclosure system. For self-supported enclosure systems,
deformation compatibility presents a unique challenge
because in-plane deformation of the cladding system is
■ This paper presents a method of using a building’s enclosure typically much different from that of the structural frame.
system of precast concrete panels as the lateral-force-resisting Where enclosure systems are self-supported, it may be
system for a steel framed structure. possible to use their in-plane strength to offer force resis-
tance to the structural frame and eliminate the problem of
■ While the design, fabrication, and installation of the enclosure deformation compatibility. These systems are referred to
system are more complex and expensive, the costs are gener- here as self-supporting lateral-frame-resistant enclosure
ally less than the savings in the structural frame costs. (SSLFRE) systems.

Precast concrete cladding systems are good candidates for should align through to the foundation. Typically, the
lateral force resistance where appropriate panel arrange- bearing between panels will occur only at these vertical
ments are available. elements by use of high-density plastic or steel shims. It is
efficient to keep vertical loads off of the horizontal panel
Design considerations members to avoid the resulting flexural stresses. For this
reason, the horizontal joints between adjacent vertical
Where enclosure systems can be self-supported and can elements should not be grouted such that vertical loads are
serve as the lateral-force-resisting system for the building transferred to the horizontal members.
(an SSLFRE system), significant savings can be realized.
For one specific project, implementing an SSLFRE system Foundation design for SSLFRE systems should take
increased precast concrete costs about $900,000, while into consideration that the support points for the precast
the savings in structural steel were estimated at about concrete may not be known until the precast concrete
$2,500,000, a net savings to the client of $1,600,000. subcontractor is retained. For the best precast concrete
panel performance, uniform bearing at the bottom of the
System suitability considerations precast concrete is not desirable. To complete the design
documents for the foundations, it should be assumed that
Not all projects are suitable for SSLFRE systems. High- the precast concrete will impart concentrated point loads
rise buildings are not suitable for SSLFRE systems because below each of the continuous vertical elements. It is pos-
of material limitations related to the height and weight sible that the panels will be supported at each end only,
of the enclosure. Precast concrete SSLFRE systems are even if there are more vertical elements within each panel.
typically 10 stories or fewer, though higher systems are To proceed, foundation design for both of these conditions
conceivable. could be considered (that is, support below each vertical
within the panel or support at each end of the panel only).
For precast concrete SSLFRE systems, a good panel Also, it may be possible to refine the foundation design
configuration is a story-tall panel with punched windows. after the precast concrete subcontractor is retained, if ac-
Panels with this configuration have vertical-load-carrying ceptable to the project team.
elements at the jambs of the windows and horizontal mem-
bers, which can be rigidly connected to the jamb members, Where the precast concrete enclosure is self-supported but
and can provide the lateral resistance for horizontal loads. does not serve as part of the lateral-force-resisting system,
Panels in a U-shape or an inverted U-shape can also be deformation compatibility between the structural frame
suitable. Strip or ribbon window enclosure systems, as and the precast concrete in the plane of the wall becomes
well as strip panel configurations with column cover panel a challenge. Self-supported precast concrete cladding
systems, consist of horizontally spanning elements that are systems that also serve as the lateral-force-resisting system
not capable of carrying any vertical load. Thus they are not will match the structural frame deformations, and compat-
self-supported and are not suitable for SSLFRE systems. ibility will not be a concern.
It is best to have redundancy in the SSLFRE system. Typi- When designing a precast concrete SSLFRE system, it is
cally, all or nearly all of the precast concrete enclosure is desirable to avoid or limit structural connections between
part of the SSLFRE system. This way stresses in individual adjacent horizontal panels (at vertical joints). If enough
components are low and the system is not compromised if panels are used for the lateral-force-resisting system, de-
one or two panels are damaged. Also, it is best if most of sign stresses will be limited such that the vertical stacks of
the panels are similar to provide uniform force distribution panels can act independently. This independence signifi-
and improve the redundancy performance. If an SSLFRE cantly reduces temperature performance effects, compared
system is proposed, the architect and the owner or user with connecting all of the panels together, and simplifies
should be aware and in agreement because changes during the design of the panels for vertical and lateral loads. If ad-
the design or in the future may be restricted. jacent panel connections are necessary to achieve appropri-
ate stress limits in the panels, one should consider connect-
Also, if a panel needs to be removed at some point in an ing only two or three vertical stacks so that temperature
SSLFRE system, the removal is more difficult than for deformations have regular relief along the length of the
systems in which panels are individually connected to the wall. Stresses should be limited to avoid cracking of the
structural frame. precast concrete under normal load conditions. Depending
on the project, cracking under seismic conditions may be
Structural design considerations acceptable.
For gravity design of precast concrete SSLFRE systems, For temperature performance reasons, it is suggested that
the weight of the system will be carried by the vertical each vertical stack (or stacks, if connected together later-
elements within the panels; therefore, the vertical elements ally) should possess only one lateral shear force connec-

for Structural Concrete (ACI 318-05) and Commentary
(ACI 318R-05)2 chapter 21, while intermediate precast con-
crete shear walls, having higher ductility values, are required
to comply with the seismic specifications. ACI 318-05 chap-
ter 21 also includes the definition and additional require-
ments for special precast concrete structural walls; however,
this system is not listed in ASCE/SEI 7-02 Table 12.2-1.
The difference in design between an ordinary and an

intermediate precast concrete wall system is the design of
the connections. For intermediate precast concrete wall
systems, the connection should be designed to ensure
ductile behavior of the system (that is, yielding of the wall
reinforcement prior to fracture of the connections).
ASCE/SEI 7-02 recognizes that a different performance is

expected if the lateral system is a bearing wall (that is, one
supporting additional gravity load from the structure) or
if it is a frame system. More ductile behavior is achieved
with a frame system because the axial force on vertical
members is reduced. An SSLFRE precast concrete wall
system designed as a frame requires a backup frame (steel
or concrete) that will carry the gravity loads. In a frame
system, because of the reduced gravity forces, it is likely
that the system will experience net uplift under lateral
loads. Vertical members should be designed and properly
anchored to the foundation. If intermediate precast con-
Figure 1. Architectural schematic of one bay of typical precast concrete panels. crete walls are used, appropriate seismic detailing should
be implemented. The seismic demand will usually exceed
the tie force requirements of ACI 318-05 chapter 16.
tion to the structure at each floor. If more connections are
provided, the thermal stresses between these connections Documentation considerations
should be considered because they can be overwhelming.
On most projects, the details for the enclosure system are
Although only one shear connection is provided at each typically indicated on the architectural drawings. For SSL-
floor, multiple connections for each panel are necessary at FRE systems, some details should be indicated on the struc-
each floor to resist out-of-plane lateral forces and ensure tural drawings. It is important that the contractor be aware
stability. These additional connections should not resist that the enclosure system is self-supporting and serves as the
vertical loads or in-plane lateral loads but should only lateral-force-resisting system for the structure.
provide lateral out-of-plane resistance.
Because the design of enclosure systems is typically
For seismic design, American Society of Civil Engineers’ performance based, the criteria specific to the SSLFRE
(ASCE’s) Minimum Design Load for Buildings and Other system should be clearly identified on the contract docu-
Structures (ASCE/SEI 7-02)1 Table 12.2-1 includes values ments. These criteria include the following:
for response modification factor R, overstrength factor Ω,
and deflection amplification factor Cd for ordinary and • each SSLFRE panel type and location
intermediate precast concrete shear walls. Ordinary precast
concrete wall systems are allowed only for buildings in • panel-to-structure shear connection locations for each
seismic design category B. Intermediate precast concrete panel and the forces for which these connections
wall systems are allowed without restrictions for buildings should be designed
in seismic design categories B and C and in buildings up
to 40 ft (12 m) in height in seismic design categories D, E, • panel-to-panel shear connection locations and the
and F. forces for which these connections should be designed
Consequently, ordinary precast concrete shear walls do not • foundation support design assumptions
have to comply with seismic specifications of the American
Concrete Institute’s (ACI’s) Building Code Requirements • forces (or alternately the reinforcement) at the panel joints

Column
Column
column
Double
Brace Brace
Rigid
Rigid
Rigid
Beam Beam
Rigid
Rigid
Beam Beam
Rigid
Rigid
Rigid
Column
Column
column
Double
Rigid
Rigid
Rigid
Beam Beam
Figure 2. Detail of the structural analysis model showing supports for the panel-to-panel connections.
Example analysis and design The panels spanned one bay horizontally (between two
building columns) and were one story tall. For the panels
The following description represents a recently completed to perform as a lateral-force-resisting system, the inter-
project that used a precast concrete SSLFRE system. The nal panel forces were transferred vertically between the
exterior enclosure for this multibuilding project consisted stacked panels, and each panel was connected to the struc-
of precast concrete panels with punched windows. The ture so that diaphragm forces were transferred to the pan-
buildings ranged from four to six stories tall and had con- els. Therefore, the panel-to-panel and panel-to-structure
siderable repetition in the exterior enclosure. For confi- connections were critical. Because the panels did not carry
dentiality purposes, the location and name of the project is gravity load from the structure, uplift forces were present.
withheld at the request of the owner.
The performance of the panels under service wind loads
General description of the panels was important. Because the panels served as the facade of
the building, cracking of the precast concrete under service
The precast concrete panels were typically 30 ft wide × loads was not allowed.
14 ft tall (9.1 m × 4.3 m). The windows were about 13 ft
wide × 8 ft tall (4 m × 2.4 m) with two windows in each In addition to the typical panels (Fig. 1), there were solid
panel. The panel profile reflects the architectural require- panels without windows, panels with louvers, and panels
ments (Fig. 1). with doors. Panels that were not suitable for use as SSL-
FRE elements were hung from the structural frame.
Structurally, the panel was considered as a weak pier-
strong spandrel type of wall, with three columns and two The shear wall panels were self-supported and designed to
beams. The edge columns were about 12 in. × 12 in. stack on top of each other. Only two supports were consid-
(300 mm × 300 mm), while the center column was about ered, one at each end of the panel. The tension or compres-
24 in. (610 mm) wide × 12 in. thick. The spandrel beams sion force at the panel piers resulted from the combination
were 8½ in. (220 mm) thick with ½-in.-deep (13 mm) of the weight of the panels and the vertical seismic or wind
reveals and a border element 12 in. deep and 12 in. thick. loads.
Structurally, the spandrels were modeled with a constant
cross-sectional thickness of 8 in. (200 mm). The design High-density plastic shims placed between panels trans-
compressive strength of the concrete was 5000 psi (35 MPa). ferred the compression load to the panel below, while a

Figure 3. Three-dimensional view of numerical model of one of the buildings. Precast concrete panels are shown at the perimeter and internal steel K-braces are
used for lateral resistance in the perpendicular axis.
vertical-shear connection transferred the tension load. The IBC 2003 does not specify seismic design coefficients and
numerical analysis model reflected the lack of moment factors for a lateral-force-resisting system based on precast
transfer between panels at the mechanical connection. concrete shear walls. However, UFC 3-310-04 lists a response
modification factor R of 4, an overstrength factor Ω of 2.5,
Just one connection at one end of each panel transferred and a deflection amplification factor Cd of 4 for ordinary
the shear force from the diaphragm to the panel (the panel- precast concrete shear walls in the building frame system cat-
to-structure connection). egory. An ordinary precast concrete shear wall in the bearing
wall systems category has the following design coefficients: R
This configuration allowed the panels to expand under equal to 3, Ω equal to 2.5, and Cd equal to 3. The provisions
temperature increases without being constrained by the of UFC 3-310-04 are shared by ASCE/SEI 7-02.
structure. This single panel-to-structure shear connection
resulted in a concentration of stresses in the diaphragm at The project is located in an area with a seismic design
the connection location. category of B. The buildings are four to six stories high,
and the total seismic base shear in the applicable direction
Seismic properties is about 4% of the weight of the building.
of precast concrete shear walls
The buildings have, in general, a rectangular shape with
The project was designed in accordance with Interna- dimensions of about 120 ft wide × about 210 ft long (36 m
tional Building Code 20033 and Department of Defense’s × 64 m) (seven bays at 30 ft [9 m]). Precast concrete shear
Seismic Design for Buildings (Unified Facilities Criteria walls were used along both longitudinal sides of each
[UFC] 3-310-04).4 building except at loading docks and entrances.

Figure 4. Rear elevation of a typical panel in storage. Figure 5 shows the connection details.
Description of numerical model Results from numerical model
The precast concrete pier-spandrel panel system was mod- The models were first subjected to the lateral forces from
eled in a structural analysis program with frame elements. service wind load (designed wind velocity of 90 mph
[150 kph]) combined with the self-weight of the panels
Minimum reinforcement for columns and deep beams (about 1 kip/ft [15 kN/m]). All of the panels were sub-
was required for these elements. The frame elements were jected to a combination of axial compression and moment.
modeled at their center of gravity. Rigid pinned elements Tension forces were not present under wind loads alone.
connected two panels at the end columns, and rigid pinned Cracking was limited by the following equation:
braces were added for lateral stability. Shear and tension
or compression forces could then be transferred with this fr = 7.5 fcl = 7.5 5000 = 530 psi (3660 kPa)
configuration while moments would not (Fig. 2). ACI 318-05 Eq. (9-10)
The panel-to-structure shear connections were located at where

the bottom of each panel except at the panel that spans
between the fifth floor and the roof, which was connected fcl = compressive concrete strength
to both the fifth floor and the roof. In the numerical model,
these connections were modeled by applying lateral loads fr = cracking stress of concrete
at each of these connection points. When the full structure
was modeled, all of the nodes were connected by a rigid Analysis of all buildings showed that the stress in most
diaphragm such that the load was transferred to the panels panels was below this limit. In a few cases the stress slight-
according to their relative stiffness. Figure 3 shows a ly exceeded this limit, which was considered acceptable.
three-dimensional view of the numerical model of one of
the buildings. Global values for the design of the various connections
were obtained. The upper two floors always remained in
For service wind analysis, the uncracked properties of the compression. Thus, it was decided that the upper two pan-
sections were used, but for the seismic design of the panels would not require panel-to-panel tension connections.
els, the cracked section properties were assumed. For the bottom levels, however, tension in the panels varied

Grid
Cast-in-place concrete
Interior finish slab to be blocked out at
connection locations and
concrete filled after
precast concrete
erection
Top of slab
Metal decking
Sheet metal
pour stop
blocked out at
W-beam
connection
locations
Figure 5. Section depicting the in-plane horizontal-shear-resisting connection.
with the height and ranged from 25 kip to 40 kip (110 kN walls were used at panels with punched windows, while
to 180 kN). provisions for minimum reinforcement of walls—includ-
ing boundary elements—were used at panels with no
The horizontal shear connection between each panel, as windows. The required reinforcement was specified on the
well as at the base, was established for each building. For structural drawings.
the upper two floors, shear values ranged from 20 kip to
35 kip (90 kN to 160 kN). For the lower floors, maximum Although each panel behaves like a strong beam–weak
base shear values ranged from 45 kip to 75 kip (200 kN to column moment frame, the global system (collection of
330 kN). stacked panels) does not. There is no moment continuity
at the columns along the height of the structure or at the
Finally, the panel reinforcement was checked. Mini- beams along the length of the building. The overturning
mum axial and shear reinforcement were determined to moment from seismic and wind demand is resisted through
be adequate to resist the imposed loads. Provisions for axial forces only at the ends of each panel, like boundary
minimum reinforcement of beams and columns rather than elements in shear wall systems.

foundation
Figure 6. Panel-to-structure shear connection at base.
Documentation considerations to anchorages and plate thicknesses. The subcontractor

also modified the precast concrete–to–structure connection
The analyses’ results and connection designs were incorpo- details and combined some of the connections to limit the
rated into the design documents in a simple way. The dif- overall number of connections to the panels.
ferent panel types were identified on the plans. Elevations
for each of these panels were provided on the structural Connection design
drawings. Sections, location of connections, and main
openings were also shown. Figure 4 shows the elevation of After obtaining the shear load values for connection design
the typical panels. from the structural analysis program, connections were
designed. The cast-in-place concrete topping and edge
Connection details, reinforcement, and tables with con- deck angle were blocked out to install the connection
nection design values were indicated on the structural (consisting of threaded reinforcing bars fastened into the
drawings. During the shop drawing process, the precast precast concrete embedment) and the composite slab was
concrete fabricator made adjustments to the connections filled after precast concrete erection. These blockouts were
to attain efficiencies. These adjustments included changes required to be large enough to ensure that reinforcing in

Grid
Interior finish
Slab on grade
foundation
Figure 7. Panel-to-structure tension connection at base.
the cast-in-place concrete topping could lap sufficiently. Shear connection

Figure 5 shows the typical structure-to-panel horizontal-
shear and panel-to-panel vertical-shear (tension) connec- The shear connection to resist horizontal in-plane load-
tion design. ing at the base was designed for the maximum base shear
of 75 kip (330 kN). Analogous to the panel-to-structure

shear connection, the slab was required to be blocked out. References
Figure 6 shows this connection concept and a photograph
of connection prior to fill. 1. American Society of Civil Engineers (ASCE). 2002.
Minimum Design Load for Buildings and Other Struc-
The tension connection from the panel to the foundation tures (ASCE/SEI 7-02). Reston, VA: ASCE.
was designed for uplift. At one end, allowing horizontal
movement was imperative because of thermal expansion 2. American Concrete Institute (ACI) Committee 318.
and contraction of the precast concrete panels relative to 2005. Building Code Requirements for Structural
the structure. To accommodate this, an angle was installed Concrete (ACI 318-05) and Commentary (ACI 318R-
between a cast-in-place concrete embed and the precast 05). Farmington Hills, MI: ACI.
concrete embed. This angle contained oversized holes on
the vertical leg for fabrication and erection tolerances. 3. International Code Council. 2003. International Build-
The angle was welded to the cast-in-place concrete embed ing Code 2003. Whittier, CA: International Confer-
and was bolted into the precast concrete embed using ence of Building Officials.
plate washers with slots to allow for in-plane horizontal
movement. These plate washers were welded to the angle 4. Department of Defense (DOD). 2007. Seismic Design
to transfer vertical shear from the threaded rods into the for Buildings. Unified Facilities Criteria (UFC) 3-310-
angle. Figure 7 shows this connection schematic. 04. Washington, DC: DOD.
Construction considerations Notation
When using an SSLFRE system, a few points should be Cd = deflection amplification factor
considered during construction. Because the SSLFRE sys-
tem only provides lateral-force resistance for the structure fcl = compressive concrete strength
in the completed state, temporary bracing requirements for
the structural frame may be increased. Temporary bracing fr = cracking stress of concrete
will be required until completion of the exterior enclosure
installation. R = response modification factor
The provision of an SSLFRE system may also have an Ω = overstrength factor

effect on the concrete placement details of a project. For
the previously discussed project, floor slab blockouts were
provided at the slab perimeter to allow for the installation
of the embedded shear connections after placement of the
floor slabs. The placement of the slabs preceded the instal-
lation of the precast concrete connections such that the
blockouts were required. The blockouts required the cast-
in-place concrete subcontractor to remobilize after precast
concrete installation to complete the slabs but reduced the
amount of field welding required during precast concrete
installation.
Acknowledgments
The authors acknowledge Quanterra Structural Associates

for their input.

About the authors steel resulting from the use of the enclosure system as
the lateral-force-resisting system can be significant.
J. Paul Hobelmann, P.E., S.E., is While the design, fabrication, and installation of the
an associate principal with enclosure system is more complex when used as the
Weidlinger Associates Inc. in building’s lateral-force-resisting system, which results
Washington, D.C. in additional precast concrete costs, the costs are gen-
erally less than the subsequent savings in the structural
frame costs.
Macarena Schachter, PhD, P.E., is Keywords

a structural engineer with Wei-
dlinger Associates Inc. in Wash- Architectural, connection, seismic, shear, wind.
ington, D.C.
Review policy
This paper was reviewed in accordance with the

Matthew C. Cooper, P.E., is an Precast/Prestressed Concrete Institute’s peer-review
engineering manager with the process.
Shockey Precast Group in
Winchester, Va. Reader comments
Please address any reader comments to journal@pci

Journal, 200 W. Adams St., Suite 2100, Chicago, IL
Abstract 60606. J
This paper presents an innovative method of using a

building’s enclosure system of precast concrete panels
as the lateral-force-resisting system for a steel framed
structure. Compared with traditional methods of lateral
force resistance systems, the savings in structural

Development of a laser-speckle
imaging device to determine the
transfer length in pretensioned
concrete members
Weixin Zhao, Kyle Larson, Robert J. Peterman, B. Terry Beck,

and Chih-Hang J. Wu
T
o evaluate the structural performance of prestressed
concrete members, it is often necessary to experi-
mentally determine the deformations in the member
due to applied forces. One typical example of this is the
determination of the transfer length in prestressed concrete
members. The transfer length is defined as the distance
required to transfer the fully effective prestress force in the
prestressing strand to the concrete.1
Transfer lengths affect structural design considerations in

two ways. First, current code provisions for shear design of
prestressed concrete members are based on the amount of
■ A rapid, noncontact method for determining transfer lengths precompression in the member. Both codes governing pre-
in pretensioned concrete members has been developed that stressed concrete design in the United States (the Ameri-
generates and digitally records laser-speckle patterns at vari- can Concrete Institute’s Building Code Requirements for
ous points along the prestressed concrete member. Structural Concrete [ACI 318-08] and Commentary [ACI
318R-08]2 and the American Association of State Highway
■ The technique was verified against results obtained using and Transportation Officials’ [AASHTO’s] LRFD Bridge
the traditional method of adhering stainless steel discs and Design Specifications3) suggest a transfer length (50db for
measuring surface strains with a mechanical strain gauge, a ACI 318-05 and 60db for AASHTO LRFD specifications,
time-consuming and tedious process. where db is the nominal strand diameter) to be used when
checking shear capacity in prestressed concrete members.
■ The new method has a higher accuracy, requires minimal The prestress force in a concrete beam has been shown
setup, and can be implemented on a production-based time to vary approximately linearly from zero at the member
frame. end to a constant value at a distance from the end of the

readings often prohibits taking these measurements before
the concrete reaches an age of one or more days, thereby
not representing typical release times as early as 14 hours
to 18 hours. Thus, the implied transfer lengths obtained
using the typical method often do not represent properties
of the concrete at the actual tension release times for many
prestressed members (when concrete is less mature and
therefore weaker).
While the conventional measurement technique does not

allow transfer lengths to be determined on a production
basis, a rapid, noncontact method to accomplish this has
been developed. The technology is called laser-speckle
imaging (LSI). The objective of this study was to develop
Figure 1. A small pretensioned concrete member was used to measure the a rapid, noncontact test method for determining transfer
transfer length. lengths in pretensioned concrete members using LSI and to
validate the method by obtaining concrete surface strains
beam equal to the transfer length.1,4–9 Therefore, significant throughout the transfer length for several pretensioned
deviations in the transfer length from the code-suggested concrete members.
50 or 60 diameters could mean inadequate performance of
the member in shear. For this reason, the transfer length is Laser-speckle imaging
often empirically determined when new mixtures or strands methodology
are employed.
A controlled calibration setup using a laboratory inter-
The transfer length can also have a significant effect on ferometer was employed to determine the accuracy of
the flexural behavior of prestressed concrete members. the current surface-strain measurement system (Fig. 2).
Prestressed concrete girders can fail suddenly when Using the interferometer as the standard for displacement
flexural cracking propagates through the transfer zone of measurement, repeated measurements were taken by an
the strand.1 Beams with debonded strand are especially experienced Whittemore strain gauge user, and it was
susceptible to this phenomenon.10 Therefore, an accurate determined that the accuracy of the experienced user was
prediction of the transfer length is an important param- about ±0.0002 in. (±0.005 mm). This corresponds to a
eter used to determine whether flexural cracks will likely strain of ±25 μє over the standard 8 in. (200 mm) gauge
propagate into this zone before the member reaches its length. Thus, it was necessary for the optical system to
design capacity. have an accuracy of ±25 μє.
Transfer lengths are determined by measuring concrete sur- Optical speckle techniques have evolved into powerful
face strains at the ends of actual members or prismatic test tools for the measurement of surface strain since digital
specimens. Metal discs called gauge points are typically image recording and processing have become widely avail-
secured to the surface of the specimens at 2 in. (50 mm) able. They have the advantage of minimal surface prepa-
spacing before releasing the tension in the strands. These ration, work with almost all kinds of rough surfaces, and
points are typically mounted using epoxy or are directly have high resolution.11–13
embedded into the concrete. They are located at the struc-
tural depth of the prestressing steel (Fig. 1). Speckle is generated by illuminating a rough surface with
coherent light (Fig. 3). The random reflected waves inter-
Distances are measured between the gauge points using a fere with each other, resulting in a grainy image (Fig. 4,
mechanical gauge with a typical resolution of about 20 μє. lower right). The speckle pattern could be thought of as
Surface strain readings are usually taken before tension a fingerprint of the illuminated area in the sense that the
release, immediately after release, and then periodically speckle pattern produced by every surface area is unique.
during the first few months after tension release. Furthermore, when the surface undergoes movement or
deformation, the speckle pattern in the image plane will
The determination of transfer lengths in pretensioned also move or deform accordingly. Thus the displacement
concrete members has been done using the current proce- or deformation information of the object surface can be
dure for more than 40 years. This method is time consum- extracted by measuring the speckle pattern movement.
ing and is subject to considerable human judgment and
possible errors because the mechanical readings are taken At the present time, no speckle strain technique had been
and manually recorded. In addition, the labor-intensive applied to prestressed concrete surface measurement, ex-
process of installing the gauge points and taking repetitive cept possibly in a very controlled laboratory setting. This

Figure 2. Determining the accuracy of a standard mechanical strain gauge (Whittemore type).
is because prestressed concrete surface strain measurement disrupts the in-plane displacement measurement.
has some characteristics that make it difficult for a regular
optical sensor to be used: Therefore, an optical strain sensor for the prestressed con-
crete surface strain measurement must be removable from
• The tension release process during the production of the specimen, have large dynamic range, and be insensitive
the prestressed concrete members involves the sudden to surface tilt.
release of large forces into the member. These forces
produce intense vibrations that disrupt the relative An optical strain sensor based on a five-axis freedom
position of the sensor and the concrete surface. To
protect the optical sensor from damage, the sensor
must be removed from the concrete surface after the Coherent light
source
initial reading and before the tension release process.
This prevents some optical strain sensors, designed
to be attached to the specimen surface throughout the In phase-
bright speckle
measurement process, from being used.14
• The prestressed concrete surface strain is in the range

of several hundred μє, which requires the strain sensor
to have a large dynamic range.
Out of phase-
• The concrete surface undergoes out-of-plane tilt (yaw dark speckle
and pitch) and roll (rotation to the axis perpendicu-
lar to the surface) as well as in-plane displacement
components. Among them, only in-plane displacement
components are useful in extracting the transfer length Surface Film plane
information, but the presence of the other axis move- or detector
ments generates additional speckle shifting and thus
Figure 3. Concept of laser speckle.

Figure 4. Basic sensor operation principle. Note: CCD = charge-coupled device.
movement measurement technique15,16 was developed. The camera (Fig. 4).

optical system of the sensor is configured to have large
dynamic range and is able to measure the surface dis- During the measurement, the optical strain sensor is
placement accurately without being affected by other axis mounted onto the concrete surface before tension release.
movements. The camera captures a speckle image with two side-by-
side speckle patterns in it (Fig. 4). These two patterns
Figure 4 shows a diagram of the sensor. A laser is collimated are denoted A1 and B1. After tension release, the opti-
by lenses L1 and L2 and then directed to the specimen cal sensor is mounted back onto the surface. The camera
surface at point A and point B, respectively, by a polariza- captures another image with two speckle patterns, which
tion beam-splitter B1. The reflected waves from the diffusive are denoted A2 and B2. By comparing the pair of speckle
surface are directed through polarization beam splitter B1 patterns A1 and A2 and using well-known cross-correla-
and lens L3. Right behind the lens is a nonpolarizing beam tion methods, the displacement ΔA of one pattern relative
splitter B2 that sends the laser beam to mirror M4. The light to the other can be extracted. The displacement ΔB can be
beams then go back through beam splitter B2 and are finally extracted from pattern B1 and pattern B2 in a similar way.
captured by a charge-coupled device (CCD) camera. The surface strain S between point A and point B thus can
be calculated by
Because there are two laser beams (which produce images
that are reflected from point A and point B on the object
DB - DA
surface), the camera actually captures two speckle patterns. S =
L
The analysis of the speckle images would be difficult if
the two speckle patterns overlapped each other. To prevent
this from happening, half of each laser beam is blocked where
with stop 1 and stop 2, such that only half of the areas
around point A and point B are illuminated. This results L = the gauge length
in two side-by-side speckle patterns generated by point A
and point B on the digitized images captured by the CCD = 8 in. (200 mm) for the current setup

Laboratory verifications
of LSI technique
To test the accuracy and sensitivity of the LSI methodol-

ogy and the feasibility of using this method to determine
concrete surface strains, a laboratory setup was fabricated
and used to conduct direct comparisons with conventional
strain measurement methods. Figure 5 shows the test setup
used. A concrete prism was compressed in a universal test-
ing machine and the concrete surface strains were recorded
using both the LSI technique and with surface-mounted
electrical resistance strain gauges (ERSGs). This prism,
which had ERSGs mounted on all four vertical faces, was
successfully used to assist the researchers in isolating the
longitudinal (axial) strain component from other distortions
that are inherently present due to varying degrees of bend-
ing. Excellent correlation between the two methods was
achieved in the laboratory.
The next stage in the validation procedure was to compare

surface strain results between the LSI technique and Whit-
temore strain gauge readings when a large strain change
occurred in a short time (that is, the tension release of a
prestressing strand).
To accomplish this, several pretensioned concrete members

were fabricated using different concrete mixtures. The mix-
ture proportions used in this study corresponded to self-con-
solidating concrete mixtures that were part of a larger PCI Figure 5. Comparison of laser-speckle imaging method with electrical resis-
tance strain gauges on concrete prism in a universal testing machine.
study.6 The members had a trapezoidal cross section (Fig. 6)
that was used as part of the development of a simple quality
assurance test reported previously.17 fixture was accomplished using a steel mounting plate.
The pretensioned concrete members were each 9 ft 6 in. Because the LSI technique relies on the optical pattern
(2.9 m) long. The transfer lengths were measured on one recognition of images before and after movement, changes
side of each member using both the traditional Whittemore or weathering to the concrete surface could limit the ability
strain gauge and the noncontact LSI method. To facilitate of this technique to measure long-term effects. However,
the laser-speckle measurements, an aluminum rail was
mounted to the side of the member (Fig. 7 and 8). The rail
was attached to the members using small ¼-in.-diameter
(6 mm) inserts that were cast into the sides of the preten- 8 in.
sioned concrete members.
Results
41/2 in.
Surface strain measurements for the trapezoidal specimens 6 /2 in.
1
were obtained using both the standard (Whittemore) strain-
gauge technique and LSI (optical) technique. Figure 9
shows that the LSI technique results in a smoother data
line with less scatter than that generated using the existing
surface strain measurement technique with the Whittemore 3 in.
strain gauge. The LSI technique has been validated on
members cast in both indoor and outdoor operations. Ther-
mal effects, due to temperature differences between the 21/2 in. 21/2 in.
measuring instrument and measured object, are similar to
those exhibited by the traditional Whittemore strain gauge
Figure 6. Trapezoidal cross section of the pretensioned members used to verify
method, as the coupling of the gauge points in the optical the laser-speckle transfer length measurement technique. Note: 1 in. =
25.4 mm.

Figure 7. The measurement of surface strains was accomplished using both the traditional surface-mounted method and the noncontact laser-speckle imaging method.
the LSI method works well during the first month after The LSI method offers a significant improvement in the
tension release. The peak strains varied along the length of reliability of estimating the transfer length.
the member, producing an asymmetric shape. This was due
to a slight horizontal eccentricity of the strand in the small The researchers are working to automate the process of
trapezoidal cross section, which produced biaxial bending traversing along a concrete member and capturing the cor-
in the member. responding LSIs in increments of ¼ in. (6 mm). This will
enable a near-real-time determination of transfer lengths
Conclusion through computerized postprocessing of the digital im-
ages in the field. It is envisioned that LSI will become an
The LSI technique is a viable method to measure transfer effective quality-control technique to screen out deleteri-
lengths in pretensioned concrete members. The LSI tech- ous combinations of strand and concrete mixtures and to
nique, using the current optical arrangement, is more ac- determine the effect of changes in these parameters on
curate than the existing Whittemore strain gauge technique strand bond.
because it eliminates human bias and improves repeat-
ability of measurement. Furthermore, the technique does Acknowledgments
not require extensive operator training to achieve reliable
measurements, as does the Whittemore strain gauge tech- This project was funded through the Kansas Department
nique. The accuracy of the LSI technique has been shown of Transportation (KDOT) and a PCI Daniel P. Jenny Fel-
to be less than 10 μє, compared with a 25 μє accuracy for lowship. Specifically, the authors would like to thank Brian
the Whittemore strain gauge technique. Much of the error Miller and Paul Johal of PCI and David Meggers of KDOT
associated with the use of the Whittemore strain gauge for helping to make this a successful project.
technique is due to poor repeatability, resulting in a large
random error contribution. Thus, without operator training,
it is difficult to get reliable transfer length measurements
with the traditional Whittemore strain gauge technique.

Figure 8. Similar outdoor measurements were made to investigate the thermal characteristics of the device.
The authors would also like to thank Ed Volkmer, former 4. Larson, K., R. Peterman, and A. Esmaeily. 2007. Bond
graduate student at Kansas State University, who assisted Characteristics of Self-Consolidating Concrete for
with the fabrication of the prestressed concrete members Prestressed Concrete Bridge Girders. PCI Journal, V.
and the initial data collection and reduction. 52, No. 4 (July–August): pp. 44–57.
References 5. Peterman, R. J., J. A. Ramirez, and J. Olek. 2000.

Influence of Flexure-Shear Cracking on Strand De-
1. Russell, B. W., and N. H. Burns. 1993. Design Guide- velopment Length in Prestressed Concrete Members.
lines for Transfer, Development and Debonding of PCI Journal, V. 45, No. 5 (September–October): pp.
Large Diameter Seven Wire Strands in Pretensioned 76–94.
Concrete Girders, Report number 1210-5F. Austin,
TX: Center for Transportation Research, University of 6. Peterman, R. 2007. The Effects of As-Cast Depth and
Texas at Austin. Concrete Fluidity on Strand Bond. PCI Journal, V. 53,
No. 3 (May–June): pp. 72–101.
2. American Concrete Institute (ACI) Committee 318
2008. Building Code Requirements for Structural Con- 7. Lane, Susan N. 1998. A New Development Length
crete (ACI 318-08) and Commentary (ACI 318R-08). Equation for Pretensioned Strands in Bridge Beams
Farmington Hills, MI: ACI. and Piles. FHWA-RD-98-116. Washington, DC: Fed-
eral Highway Administration.
3. American Association of State Highway and Trans-
portation Officials (AASHTO). 2007. AASHTO LRFD 8. Barnes, R. W., J. W. Grove, and N. H. Burns. 2004.
Bridge Design Specifications. 4th ed. Washington, DC: Experimental Assessment of Factors Affecting Trans-
AASHTO. fer Length. ACI Structural Journal, V. 100, No. 6
(November–December): pp. 740–748.

800
Concrete surface compressive strain, μє 700
600
500
400
300
200
100 Mechanical strain gauge

Laser-speckle imaging (optical)
0
0 20 40 60 80 100
Distance along beam, in.
Figure 9. Comparison of raw (unsmoothed) strain measurements immediately after detensioning of a pretensioned specimen.
9. Oh, B. H., and E. S. Kim. 2000. Realistic Evalua- Systems for Industrial Inspection, ed. Wolfgang Osten,
tion of Transfer Lengths in Pretensioned, Prestressed pp. 30–36. Munich, Germany: Society of Photo-Opti-
Concrete Members. ACI Structural Journal, V. 97, No. cal Instrumentation Engineers.
6 (November–December): pp. 821–830.
15. Zhao, W., B. T. Beck, and J. Wu. 2004. A Novel Opti-
10. Kaar, P., and D. Magura. 1965. Effect of Strand Blan- cal Technique for Measuring 5-Axis Surface Move-
keting on Performance of Pretensioned Girders. PCI ment. In Proceedings of SPIE Optics East: Two- and
Journal, V. 10, No. 6 (December): pp. 20–34. Three-Dimensional Vision Systems for Inspection,
Control, and Metrology II, ed. Kevin G. Harding, pp.
11. Yamaguchi, I. 1981. A Laser-Speckle Strain Gauge. 66–73. Philadelphia, PA: Society of Photo-Optical
Journal of Physics E: Scientific Instruments, V. 14, No. Instrumentation Engineers.
11 (November): pp. 1270–1273.
16. Wu, C.-H., W. Zhao, T. Beck, and R. Peterman. 2009.
12. Johnson, Peder. 1998. Strain Field Measurements Optical Sensor Developments for Measuring the Sur-
in Industrial Applications Using Dual-Beam Digital face Strains in Pretensioned Concrete Beams. Strain,
Speckle Photography. Optics and Laser in Engineer- International Journal for Experimental Mechanics.
ing, V. 30, No. 5 (November): pp. 421–431. Published electronically March 27, 2009.
13. Sjödahl, M. 1999. A Whole Field Speckle Strain Sen- 17. Peterman, R. J. 2009. A Simple Quality Assurance
sor. In Proceedings of SPIE: Optical Engineering for Test for Strand Bond. PCI Journal, V. 54, No. 2
Sensing and Nanotechnology (ICOSN ’99), ed. Ichirou (Spring): pp. 143–161.
Yamaguchi, pp. 84–87. Yokohama, Japan: Society of
Photo-Optical Instrumentation Engineers.
14. Wegner, R., and A. Ettemeyer. 1999. The Miniaturiza-

tion of Speckle Interferometry for Rapid Strain Analy-
sis. In Proceedings of SPIE: Optical Measurement

Notation S = strain
db = strand diameters ΔA = displacement at point A
L = gauge length ΔB = displacement at point B
About the authors Abstract
Weixin Zhao, PhD, is a postdoc- The current experimental method to determine the
toral research associate in the transfer length in prestressed concrete members
Department of Mechanical and consists of measuring concrete surface strains with
Nuclear Engineering at Kansas a mechanical strain gauge before and after releasing
State University in Manhattan, tension. Because this is a time-consuming and tedious
Kan. process, transfer lengths are seldom measured on a
production basis. Furthermore, when transfer lengths
Kyle Larson, PhD, was a postdoc- are determined using the current method, the times to
toral research associate in the release tension of the members being measured are
department of Civil Engineering often delayed, thereby resulting in artificially higher
at Kansas State University. release strengths for the members evaluated.
A rapid, noncontact method for determining transfer

lengths in pretensioned concrete members has been
Robert J. Peterman, PhD, P.E., is a developed. The new method uses laser-speckle patterns
professor of Civil Engineering at that are generated and digitally recorded at various
Kansas State University. points along the prestressed concrete member. The
technique was verified against results obtained using
the traditional method of adhering stainless steel discs
and measuring surface strains with a mechanical strain
gauge. The new method has a higher accuracy, requires
B. Terry Beck, PhD, is a professor minimal setup, and can be implemented on a produc-
of Mechanical and Nuclear tion-based time frame.
Engineering at Kansas State
University. Keywords
Laser-speckle imaging, strand bond, transfer length.
Chih-Hang J. Wu, PhD, is an Review policy

associate professor of Industrial
and Manufacturing Systems This paper was reviewed in accordance with the Precast/
Engineering at Kansas State Prestressed Concrete Institute’s peer-review process.
University.
Reader comments
Please address any reader comments to journal@pci.org

or Precast/Prestressed Concrete Institute, c/o PCI Jour-
nal, 200 W. Adams St., Suite 2100, Chicago, IL 60606. J

What Certification Program are you betting on?
Certification is more than inspections, paperwork, and

checklists! It must be an integrated and ongoing part of the industry’s
Body of Knowledge! PCI is the technical institute

Phone: 312-786-0300
Fax: 312-621-1114
Phone: 312-786-0300
Fax: 312-621-1114
Phone: 312-786-0300
Fax: 312-621-1114
for the precast concrete structures industry and as such, PCI Certification
is an integrated and ongoing part of the industry’s body of knowledge.
Specify PCI Certification

and hold the winning hand.
To learn more, visit www.pci.org/certification or contact

200 West Adams Street I Suite 2100 I Chicago, IL 60606-5230 Dean Frank, P.E., PCI director of quality programs,
Phone: 312-786-0300 I Fax: 312-621-1114 I www.pci.org at dfrank@pci.org or (312) 583-6770.
Professional Membership Application
MAIL TO:
PCI MEMBER SERVICES
200 WEST ADAMS ST., SUITE 2100
CHICAGO, ILLINOIS 60606
PHONE: 312-360-3214 • FAX: 312-361-8081
EMAIL: memberservices@pci.org
WEB: www.pci.org
Check One:
Professional Member: Any person who is a licensed professional engineer or
architect, or who meets the education and experience requirements deemed by the LICENSED ENGINEER
Board of Directors to be equivalent to those of professional licensure. LICENSED ARCHITECT
APPLICANT NAME APPLICATIoN DATE CrEDENTIALs (check all that apply)

o AIA o rA o PE o sE
❏ Mr. ❏ Ms. ❏ Dr. ❏ Prof. o PhD o other: ________________
ADDrEss for rECEIPT of MAIL
CITY, ToWN or PosT offICE, sTATE, AND ZIP CoDE
PHoNE fAX
EMAIL
orGANIZATIoN JoB TITLE
CoMPANY ADDrEss (If oTHEr THAN ABoVE) NUMBEr AND sTrEET
CITY, ToWN or PosT offICE, P.o. NUMBEr If APPLICABLE, AND ZIP CoDE
NATUrE of CoMPANY’s BUsINEss AND/or ProDUCTs
WEB ADDrEss
I HoLD LICENsE NUMBEr ___________________________________________________________________________________ IN THE sTATE of _________________________________________________
PrIMArY oCCUPATIoN (check all that apply)

o Academic o Architect o Bridge Engineer o Consultant o Contractor o Design/Build o Governmental o Precast Manufacturer o structural Engineer o other_______________
PrIMArY EXPErTIsE (check all that apply)

o Concrete Materials o Drafting o Design o research o Codes/standards o Education o Hollow Core o QA/QC o Management o Marketing/sales
o Disaster Mitigation (fire, seismic, blast, etc.) o specialty Products (poles, piles, soundwalls, ties, etc.) o other________________________________________________________
ANNUAL FEE = $110 (NON-U.S. = $140) (U.S. ACADEMIC EMPLOYEE = $50) (NON-U.S. ACADEMIC EMPLOYEE = $65).
PAYMENT IN U.S. DOLLARS DRAWN ON A U.S BANK MUST ACCOMPANY THIS APPLICATION.
Fee is not deductible as charitable contributions for federal income tax purposes but may be deductible by members as a business expense.
This certifies that the company with which I am affiliated does NOT qualify for Producer membership or
Associate membership as described in the Bylaws of the Precast/Prestressed Concrete Institute. Applicant’s initials
NON-PRODUCER APPLICANTS PRODUCER APPLICANTS
PAYMENT REQUIRED FOR NON-PRODUCER PROFESSIONAL APPLICATIONS PRODUCER MEMBER EMPLOYEES

NO FEE — PRODUCER AUTHORIZED SIGNER
o MasterCard o Visa o Check Amount Enclosed: _______________ REQUIRED
Card Number: | | | | | | | | | | | | | | | | | Exp. Date:____________________ Authorized Signer Name
Card Holder’s Name as it Appears on Card: _______________________________________

Authorized Signature
signature: ________________________________________________________________
Your Next Step…
Join the PCI Associate Professional Network
RECENT GRADUATES:
Your first year of PCI membership is FREE!
You’ve invested a great deal in getting to where you are now—contributing thousands of hours, financing tuition fees, and
producing piles of homework. It all adds up to dedication. You’ve proven that.
Now it’s time to start reaping the benefits of your studies and hard work. When you become an Associate Profes-
sional Member of PCI, you’ll gain valuable opportunities to develop your career and network with more than 1,800
design professionals. And you’ll become part of a vital and growing industry that’s looking for talented professionals
like you.
Download an application at WWW.PCI.ORG

Classifieds
Sales Representative Advertisement

The Tindall Corporation has an immediate need for a Sales Representative in our newly established Texas territory. The suc-
cessful candidate would report to the Vice President of Sales for Tindall. A brief list of responsibilities includes; developing
support relationships with architects, engineers, general contractors and owners to promote the design and use of precast/
prestressed concrete. Candidate will be responsible for project research to locate opportunities, obtaining drawings, prebid
contact and support, assisting to put together bids and proposals, making bid presentations, delivering proposals and quota-
tions, follow up activities and negotiating final sales.
An Architectural/ Engineering undergraduate degree as well as business development and three to five years of
prior sales experience is preferred.
Tindall is a richly diverse industry leader in the design, manufacture and erection of precast, prestressed concrete
systems for mixed-use, educational, commercial, industrial and parking structures. We are also a leading sup-
plier of underground structures and fully upfitted modular prison cells. We are privately owned with headquarters
located in Spartanburg, SC. We have manufacturing locations in Spartanburg, SC; Petersburg, VA, Conley, GA,
Moss Point, MS, in San Antonio, and Texas.
Please fax, email, or mail your resume and salary requirements to H. David Britt PO BOX 1778, Spartanburg, SC.
29304 or email to davidbritt@tindallcorp.com
PCI Level III and II Quality Control Technician/Inspectors

Join Rotondo Weirich’s talented staff of dynamic QC professionals. RW provides specialty precast modular
concrete housing. We are looking for individuals with technical competence, entrepreneurial drive and passion for
doing a job right. Please contact our HR department at 215-256-7940. www.rotondoweirich.com
Manufacturer’s Reps Wanted

Nox-Crete, an established manufacturer of concrete construction chemicals, is looking for Independent Manufac-
turer’s Sales Agents to represent our products to precast and ready-mix concrete producers in the US and Canada.
This is an excellent opportunity for Sales Agents familiar with the industry.
For consideration, email careers@nox-crete.com.
Support the
PCI Foundation
From these humble beginnings emerged the spark for a "grand plan"
to provide for the Industry's future success by helping ensure the continued
availability of talented people and new technology...
HOW TO PURCHASE:
Contact Skip Francies at sfrancies@tampabay.rr.com or
call at 813-230-8266. If there is no answer, please leave a
detailed voice mail stating the following:
• How many you would like

• Where you would like them shipped
Make check payable to PCI Foundation and mail to:
Skip Francies
6136 Kestrelridge Drive
Lithia, FL 33547
Once the check is received the PCI Foundation wristbands

100% Silicone Wristband
will be mailed to you. Option: purchase the wristbands
directly from Skip, cash (receipt issued) or check payable
to the PCI Foundation. All proceeds will be directly
forwarded to the PCI Foundation.
Purchase them for yourself, your colleagues or your

entire team…
1 to 49 ..................................................$5.00 each
50 to 99 ................................................$4.50 each
100 plus ................................................$4.00 each
Coming ahead Board of Directors
Durability Chairman Gregory F. Force
Vice chairman Tom H. Kelley
• Transverse Post-tensioning Arrangement for Side-by-
Secretary-treasurer Dean Gwin
Side Box-Beam Bridges
Immediate past chairman Donna S. Reuter
• Simplified Transverse Post-tensioning Construction
PCI president James G. Toscas
and Maintenance of Adjacent Box Girders
• Spalling of Partial-Depth Precast, Prestressed Bridge Keith G. Allen Mason Hardaway Lampton
Deck Panel Investigation Using Ground Penetrating
Michael Baez Paul J. Lautenbach
Radar
• Proposed Evaluation and Repair Procedures for Jay M. Cariveau Charles R. Lowe
Precast, Prestressed Concrete Girders with End Zone
Todd Culp Marianne Methven
Reinforcement
Francesco J. Genoese Alexander G. Mihaylov
Also
Harry A. Gleich Christopher J. Pastorius
• Torsion Tests on Spun-Cast Prestressed Concrete
Poles Leon H. Grant Chuck Prussack
• Effect of Environmental Conditions on Field Welding
Dean Gwin Larbi Sennour
of Precast Concrete Connections
• The MnDOT Inverted-T System: Parametric Studies Dan Juntunen Rita Seraderian
for Preliminary Design
Daniel L. Kennedy C. Douglas Sutton
Edward S. Knowles Keith E. Wallis Jr.
Technical Activities Daniel P. Kolb Peter Yurkiw
Council Bohdan Kusznir
Chair Larbi Sennour

Secretary Jason Krohn
Ex-Officio, fib Representative Thomas J. D’Arcy
Ex-Officio, ANIPAC Representative Mario E. Rodriguez
Ex-Officio, Code Representative S. K. Ghosh
Craig Barrett Wayne Kassian
Ned M. Cleland Karen A. Laptas
Robert Frosch Brian Lee Lawrence
Pat Hynes Jason P. Lien
Phil Iverson Clay J. Naito
David Jablonsky Stephen J. Seguirant

Georgia/Carolinas PCI
Index of advertisers Peter I. Finsen
Phone: (678) 638-6220
The Consulting Engineers Group Inc. .............................................4 Fax: (678) 638-6221
www.cegengineers.com Email: peter.finsen@gcpci.org
Hamilton Form Co. ..........................................................Back Cover www.gcpci.org
www.hamiltonform.com
JVI Inc. ....................................................................................................1 Mid-Atlantic Precast Association (MAPA)
www.jvi-inc.com Greg Winkler
Midwest Structure Engineering Inc. .............................................. 27 Phone: (856) 761-8885
www.midweststructure.com Email: gwinkler@mapaprecast.org
Prestress Supply Inc. .......................................................................... 12 www.mapaprecast.org
www.pci.org/host_sites/PSI/index.html
Tucker’s Machine & Steel Service Inc. ........................................... 10 PCI Midwest
www.tuckerbuilt.com Mike Johnsrud
Phone: (952) 806-9997
Fax: (952) 806-9998
Email: mike@midwestprecast.com
www.midwestprecast.com
PCI Central Region

Phil Wiedemann
Phone: (937) 833-3900
Fax: (937) 833-3700
Email: phil@pci-central.org
www.pci-central.org
Regional Directors PCI Northeast

Rita Seraderian
Phone: (888) 700-5670
Central Atlantic Bridge Associates
Fax: (617) 489-5810
Heinrich O. Bonstedt
Email: contact@pcine.org
Phone: (610) 395-2338
www.pcine.org
Email: info@caba-bridges.org
www.caba-bridges.org
PCI of Illinois & Wisconsin (PCI-IW)
Marty McIntyre
Colorado Prestressers Association
Phone: (708) 386-3715
J. D. Schafer
Fax: (708) 386-5922
Phone: (303) 880-3843
Email: martymci@pci-iw.org
Email: jdschafer@stresscon.com
www.pci-iw.org
Florida Prestressed Concrete Association
Precast Concrete Manufacturers’ Association
Joe Lord
of Texas (PCMA of Texas)
Phone: (813) 579-7232
Chris Lechner
Fax: (813) 315-6026
Phone: (210) 633-6743
Email: info@fpcaweb.org
Email: lechner@pcmatexas.org
www.fpcaweb.org
www.pcmatexas.org
recast/Prestressed Concrete Manufacturers

P
Association of California (PCMAC)
Doug Mooradian
Phone: (818) 247-6177
Fax: (818) 240-3041
Email: doug@precastconcrete.org
www.precastconcrete.org

PCI staff phone directory
James G. Toscas, president jtoscas@pci.org (312) 360-3204
Jeffrey Appel, controller jappel@pci.org (312) 360-3208
Roger Becker, research and development director rbecker@pci.org (312) 360-3213
Kim Brown, shipping and receiving associate kbrown@pci.org (312) 428-4946
Chris Bulvan, technical activities administrative assistant cbulvan@pci.org (312) 360-3218
Jessica Burnett, quality programs manager jburnett@pci.org (312) 583-6774
Rebecca Coleman, executive office administrator bcoleman@pci.org (312) 583-6783
Brenda Davis, staff accountant bdavis@pci.org (312) 360-3211
Dean Frank, quality programs director dfrank@pci.org (312) 583-6770
Sidney Freedman, architectural systems, industrial

sfreedman@pci.org (312) 360-3203
operations, and safety director
Jason J. Krohn, technical activities managing director jkrohn@pci.org (312) 583-6771
Megan Lanning, events manager mlanning@pci.org 312-583-6781
Brian Miller, business development managing director bmiller@pci.org (312) 360-3216
Alex Morales, educational activities director amorales@pci.org (312) 360-3219
William Nickas, transportation systems managing director wnickas@pci.org (312) 583-6776
Kirstin Osgood, sales and member development manager kosgood@pci.org (312) 360-3206
Jennifer Peters, administrative assistant jpeters@pci.org (312) 583-6780
Whitney Stephens, communications manager wstephens@pci.org 312-428-4945
Wallace G. Turner, transportation systems staff engineer wturner@pci.org (312) 583-6776
Cindi Ward, membership services administrator cward@pci.org (312) 360-3214
You can find the contact information for the PCI Journal and PCI Publications department staff in
the table of contents.

Meet Doug Sutton
Industry educator Sarah Fister Gale
When Doug Sutton sits down in a PCI meeting, it is He received his PhD in 1972,
rare that he doesn’t recognize at least one of his former along with a National Science
students. Sutton, who retired in 2007, taught structural Foundation Fellowship, and has been
engineering at Purdue University for 46 years and intro- formally recognized many times with
duced many of today’s industry leaders to the structural awards for his teaching, advising, and
and aesthetic benefits of precast concrete. service.
“My focus has always been on teaching,” Sutton says, “It was a career I obviously had
though he admits that he didn’t set out to be a professor. something of an aptitude for,” he says.
Sutton received his BSCE degree in 1957 from the University of Maine Along with teaching, Sutton
and took a position at Charles A. McGuire, an engineering consulting firm became an active member of PCI
in Rhode Island. He says that when he took that job he was certain of only in the early 1970s. He was an active
one thing, “I was never going to work in structural design.” member of the Bridge Committee,
The firm’s owners had other plans. On his first day, instead of assigning and in 1989 he became the Bridge
him to highway design, his area of experience, they put him in the bridge Committee’s liaison to the Technical
group. “I’ve been a structural engineer ever since,” he says, laughing. Activities Council (TAC). He was
At McGuire, Sutton helped shape the U.S. interstate highway system, design- chair of TAC from 1999 through
ing bridges that are still in use today. However, in 1958, at the urging of mentor 2004, chaired the Research and
Gordon Archibald, he took a leave of absence to pursue his master’s degree at Development Committee from 2004
Purdue. to 2009, and became chair of the
While there, he took a position as a teaching assistant to help pay the Education Activity Council in 2010.
bills, filling in for a semester for a full-time instructor who left to pursue his He was recently elected to a second
PhD. At the end of that semester, Sutton was ready to get back to the world term, and will hold this role until
of engineering practice. In 1960, he received his master’s degree, returned to 2013.
McGuire, and married his wife Ellie. For all his hard work and commit-
“It was a very competitive environment in those days, and my graduate ment, Sutton was named one of the
education gave me many opportunities,” he says. 50 Titans of the Precast/Prestressed
Precast concrete was growing in popularity, and Sutton was asked to Concrete Industry in 2004, and
design and even proof test many of the precast concrete beams that were in 2009 he was given the Medal of
coming into fashion. Honor, PCI’s highest award.
After only a couple of years, academia called to him again. In 1961, Purdue “PCI has been a hugely positive
invited Sutton to return as a full-time faculty member while pursuing his own thing in my life,” Sutton says. “The
PhD. familylike structure of this organiza-
“It wasn’t an easy decision,” he says. He liked designing bridges, but he tion has helped move the industry
also recognized that the world was changing. Sputnik had gone up in 1957, forward.”
spurring renewed interest in engineering. It also helped him stay connected
Sutton says, “We realized we had to get moving as a nation.” to the world of precast concrete and
So Sutton took his wife and newborn baby back to Indiana and began a made him a better educator.
new career as an academic that lasted until May 2007. “Sitting on committees, I had a
During his time at Purdue, Sutton taught at least 10 different courses at view of the engineering world that
all levels in structural analysis, structural design, mechanics, and experimen- I couldn’t get in academia,” he says.
tal mechanics, including a course on prestressed concrete, which he taught “This connectivity to the real world
from the early 1980s until he retired. has been very important to my
career.” J

2012 Call for Papers and Student Posters is Now Open!
Abstracts due FebRuARy 6, 2012
2012 PCI Convention and National Bridge Conference

September 29–October 3
Gaylord Opryland, Nashville, Tennessee
PCI is accepting abstracts for technical papers
and student posters to be presented at the 2012 PCI Submission Requirements
Convention and National Bridge Conference in
Nashville, Tennessee. Abstracts and papers will be Abstracts should be submitted electronically. Visit
peer-reviewed and accepted papers will be published www.pci.org and click on the Call for Papers and
in the conference proceedings. Student Posters button to access the submission site.
The PCI Convention and National Bridge Conference Contact:

is the premier international venue for the exchange of Buildings:
Suite 2100 Chicago, IL 60606 Suite 2100 Chicago, IL 60606
200 West Adams Stree
Phone: 312-786-0300 Phone: 312-786-0300 Phone: 312-786-0300
ideas and state-of-the-art information on precast Brian Miller, P.E., LEED AP, bmiller@pci.org
Fax: 312-621-1114
www.pci.org
Fax: 312-621-1114
www.pci.org
Fax: 312-621-1114
www.pci.org
concrete design, fabrication, and construction. The Transportation:
event attracts an average of 1,000 participants each William Nickas, P.E., wnickas@pci.org
year and provides an outstanding opportunity for Student Posters:
networking, education, and sharing of ideas. Don’t Alex Morales, EIT, M. Ed., amorales@pci.org
miss out on this excellent opportunity to share your
knowledge–submit your abstract today!

H a m i lt o n F o r m C r e at e s F u n C t i o n
16 Case study “When I call Hamilton Form, I know I’ll get

forms that are easy to set-up, cast in and strip.
I’ve come to depend on Hamilton From for
selF-stressing Jumbo tee highly complex, highly efficient formwork.
tHe sHoCkey PreCast grouP Plus, they deliver on time and within budget,
which makes for a winning combination.”
Greg Elliott
Vice President, Manufacturing
The Shockey Precast Group
Precast/prestressed concrete is a highly engineered, complex product that

requires precision fabricated, high quality, durable formwork. That’s why,
when The Shockey Precast Group needed to cast a unique double tee
product, they called on Hamilton Form Company.
The “Jumbo Tee” has sixty inch stems with a highly stressed bulb
configuration at the stem bottoms. Hamilton Form built the self-stressing
formwork in sections for both shipping and stripping considerations.
The stems have minimal draft requiring the sideforms to be pull back for
stripping. The 118-foot long, twelve foot wide product is casting efficiently
and strips easily, keeping productivity high and delivery on time.
The next time you need formwork, from the simple to the complex,
call Hamilton Form to deliver: 817 590-2111 or sales@hamiltonform.com
Hamilton Form Company, Ltd.

7009 Midway Road • Fort Worth, Texas 76118
Custom forms. Custom equipment. Practical solutions.

Pci Journal - Winter 2012 Vol. 57 No. 1 PDF

Uploaded by

Document Information

Original Title

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

Pci Journal - Winter 2012 Vol. 57 No. 1 PDF

Uploaded by

Copyright:

Available Formats

 Winter 2012

 Volume 57, Number 1

Precast/Prestressed Concrete Institute

Perry Adebar Brian Lee Lawrence

Preview of PCI’s Japan Earthquake Reconnaissance Team Report 47

Observations from the February 27, 2010, Earthquake in Chile 52

Effects of Loading Conditions on the Behavior of Unbonded Post-tensioning 76

Seismic Performance and Retrofit of Precast Concrete Grouted Sleeve Connections 97

Structural Behavior of an Innovative All–Precast Concrete Dual System 110

Postmaster: Please send address changes to PCI Journal, 200 W. Adams

Development of a Laser-Speckle Imaging Device 135 Departments

Note of Appeciation Inside Front Cover President’s Message  6

From PCI Headquarters 14

Coming Ahead 149

135 Board of Directors 149

Index of Advertisers 150

The precast concrete PricewaterhouseCoopers office build-

Journal Advisory Committee Amir Fam George D. Nasser Stephen Seguirant

PCI Journal | Wi n t e r 2012 3

THE C O N S U LT I N G ENGINEERS GROUP INC.®

making all the

Main Offices: Other Offices:

San Antonio, TX Albuquerque, NM Lake Mary, FL Ft. Lauderdale, FL

PCI Journal | Wi n t e r 2012 5

Potential advantage 1: High-performance systems

• Beautiful: this is self-explanatory (just ask the beholder).

6 W int e r 2 0 1 2 | PCI Journal

Potential advantage 2: Advanced materials

PCI Journal | Wi n t e r 2012 7

The next material for us to capitalize on might be ultra-high-performance concrete

Our opportunity is now

8 W int e r 2 0 1 2 | PCI Journal

Now is the time to bend minds—starting with our own. J

PCI Journal | Wi n t e r 2012 9

Theme-related peer-reviewed papers include a paper from Romania on precast concrete

PCI Journal | Wi n t e r 2012 11

Comments? Naaman sends thanks

PCI Journal | Wi n t e r 2012 13

Former PCI technical director celebrates 90th birthday

Distinguished Educator Award call for nominations

14 W int e r 2 0 1 2 | PCI Journal

Young Educator Award call for nominations

Engineering Student Design Competition:

PCI Journal | Wi n t e r l 2012 15

New sandwich wall panel Designer’s Notebook released

16 W int e r 2 0 1 2 | PCI Journal

Second-edition Level III personnel

PCI Sustainability Manager

PCI Journal | Wi n t e r l 2012 17

PCI-certified plants reach significant milestones

PCI Convention and National Bridge Conference

18 W int e r 2 0 1 2 | PCI Journal

New Zealand Engineering Society touts

Solicitation for Jenny Research Fellowship proposals

PCI Journal | Wi n t e r l 2012 19

PCI Foundation auction raises nearly $25,000

Becker nominated for ACI Board of Direction

20 W int e r 2 0 1 2 | PCI Journal

PCI manuals listed as referenced standards

Leadership PCI call for nominations

PCI Journal | Wi n t e r l 2012 21

The Student Research Poster Session

22 W int e r 2 0 1 2 | PCI Journal

Winter 2012

Volume 57, Number 1

Note of Appeciation Inside Front Cover President’s Message 6