PCI Journal Sept-Oct 2000 VOL. 45 NO. 5 PDF

I JI
V.
/
I/ /
1k I,.
J4/
,rIli /
r
PCI Design Award Winner
for Best Retail Building
.
U
4
UO
O
@
O
I
e
S
O
I
aW0S(
?
0
6
1.OBS 009 I.
Ii
III.
11
This state of the art concrete transporter features full 6-yard

capacity for maximum efficiency. Its maneuverable, quick and
easy to operate. The T-630 is built from all new, readily
available parts so that if you ever need repairs, youll
experience less down time. At Tuckers Machine and Steel
Service, Inc. were proud to build what most professionals in
the Pre-stress Industry consider the ultimate concrete transport
vehicle.., because it saves time and money on every jobl
Articulated steering
6-yard Concrete
Transport Vehicle
Long, high reach
Highly maneuverable
Completely sealed hydraulics
Massive frame, axle and tires

9
Easy to maintain
14 auger for faster delivery less auger wear
1
The Thckerbilt 630 design,
manufacturing and sales by
-
P. H Wit
74 T-O!
P0. Box 492810
MACHINE & STEEL SERVICE, INC

352-787-3157 FAX 352-365-1732
Leesburg, Florida 34749-2810
Visit our website: vvww.tuckerbilt.com
a--
For a Strong Start...

To a Great Finish,
Rely on
Master Builders
From structural to architectural, Master Builders offers a wide variety of products and systems for
the precast industry. Our technical expertise and knowledgeable staff can help you build precast
E NIACO
Mortars
components that offer high performance and keep you competitive.

Start with our quality admixtures that improve strength and durability,
long-term structural integrity and enhanced surface finish.

Our grouts provide consistent, nonshrink
Rheobuild
r
Admlxtures
performance. A wide variety of repair products and
epoxy adhesives offer low-cost solutions for all

types of concrete repair.
Finish with our Preco line for creating architectural
precast concrete, including Preco
retarders, reinforcing accessories, elastomeric
formliners, architectural stains, and protective coatings and sealers.

For the right solutions to your next precast projects,
call Master BuildersYour partner in precast.
In
II!
Master Builders
Technologies
Master Builders Inc.. 237(X) Chagrin Blvd.. Cleveland, Ohio 44122-5554. -MBT-Fax: 21&831 -3470
Preco line. 800-645-1258
Website: http://www.masterbuilders.com
Building Tomorrow Together
SKWMRT
september-october 2000
vol. 45, no- 5
JOURNAL
Editor-in-Chief
George D. Nasser
Assistant Editor
Karen Banasiak
STREST SED CONCRETE INSTITUTE
DEPARTMENTS
6
Coming Ahead
13 Chairmans Message
Proposed PCI Education
Foundation
14 2000 PCI Design Awards

Program The Winning
Entries
128 Problems and Solutions
Formwork Issues by
John E. Dobbs and
B. A. (Skip) Plotnicki
TABLE OF CONTENTS
Precast Concrete Panels Give Scale and Grandeur to
Lazarus Department Store
Gar Muse and Anthony Di Giacomo
Presents the design-construction highlights of this multi-million dollar
department store in Pittsburgh, Pennsylvania.
Central ArterylTunnel Project: PrecastlPrestressed

Structures Span the Big Dig
Keith Donington, Paul Towell, and Vijay Chandra
In Part 3 of this multi-billion dollar project, the authors discuss the use of
precast, prestressed concrete in two bridges and a marine pier.
130 Reviews of Technical
Publications
20
30
Proposed Revisions to 1997 NEHRP Recommended

Provisions for Seismic Re9ulations for Precast Concrete
Structures: Part 2 Seismic-Force-Resisting Systems
Nell Hawkins and S. K. Ghosh
131 Reference Cards
In Part 2 of this series of articles, the authors discuss specific

recommendations for precast concrete that are intended to be part
of the 2000 NEHRP Provisions.
135 Industry News

CNL Center
150 Calendar of Coming Events
34
Design Criteria for Headed Stud Groups in Shear:

Part 1 Steel Capacity and Back Edge Effects
151
Index to Advertisers
Nell S. Anderson and Donald F. Meinheit

In this PCI-funded research project on headed studs, the authors present
results and conclusions from the first part of the investigation.
46
Influence of Flexure-Shear Cracking on Strand

Development Length in Prestressed Concrete Members
Robert J. Peterman, Julio A. Ramirez, and Jan Olek
Based on a comprehensive test program, the authors recommend specific

changes to the AASHTO-ACI requirements for strand development length.
76
Strand Development and Transfer Length Tests on High

Performance Concrete Box Girders
P. Benson Shing, Daniel E. Cooke, Dan M. Frangopol, Mark A. Leonard,
Michael L. McMullen, and Werner Hutter
Presents the results of an investigation on strand transfer and development
length in high performance concrete prestressed box girders.
96
ON THE COVER
Precast concrete played a prominent role in
the construction of this prestigious Iuxuiy
department
store
in
September-October 2000
Nabll F. Grace
Pittsburgh,
Pennsylvania. For more details on this
project, see p. 20.
Transfer Length of CFRP/CFCC Strands for Double-T Girders
Presents the results of a study on variations in transfer length in two types of

carbon fiber reinforced polymer (CFRP) prestressing strands.
110
5
COMING AHEAD
Highlights of PCIs 46th Annual
Convention & Exhibition and
PCI!FHWA/fib International
Symposium on High Performance
Concrete
Design-Construction of The
Terraces at Riverfront Recapture
Design-Construction of Cruise
Terminals for Port of Miami
PCI DIRECTORS, TAC & STAFF

BOARD OF
DIRECTORS
TECHNICAL ACTIVITIES
COMMITIEE (TAC)
Chairman
of the Board
William E. Whitcher
Chairman
C. Douglas Sutton
Vice Chairman
of the Board
Saul Shenkman
Reliability of I8 in. Anchor Stud

Welds Subject to Tensile Loads
Secretary-Treasurer
Ron Schlerf
Evaluation of a High Performance

Concrete Box Girder Bridge
Directors
Charles B. Baker
Brian J. Conway
Thomas J. DArcy
Gerald E. Goettsche
Degan G. Hambacher
Marvin F. Hartsfield
Fred W. Heldenfels, IV
R. Wayne Lyle
Robert S. McCormack
Richard L. Mogel
David N. Nesius
Milo J. Nimmer
Charles P. OLeary
Michael E. Quinlan
Christopher N. Quinn
H. W. Reinking
Stanley J. Ruden
William F. Simmons
C. Douglas Sutton
Scott M. Waldron
Richard L. Wells
Continuous CFRP Prestressed

Concrete Bridges Under Static and
Repeated Loadings
JOURNAL
PRECAST/PRESThESSED CONCRETE PNSTTUTE
EDITORIAL DATA
The PCI JOURNAL (ISSN 0887-9672)
is published bimonthly by the Pre
cast/Prestressed Concrete Institute,
209 W. Jackson Boulevard, Ste. 500,
Chicago, Illinois 60606, (312) 7860300. Copyright 2000, Precast/Pre
stressed Concrete Institute.
Original manuscripts and Reader
Comments on published articles
accepted on review by the PCI
Technical Publications Review Board.
No payment is offered. The Precast!
Prestressed Concrete Institute is not
responsible for statements made
by authors of papers or claims made
by advertisers in the PCI JOURNAL.
Advertising rates available on request.
Direct all correspondence to:
The Editor, PCI JOURNAL
Precast/Prestressed Concrete Institute
209 W. Jackson Boulevard, Ste. 500
Chicago, Illinois 60606
Tel.: (312) 786-0300
Fax: (312) 786-0353
e-mail: info@pci.org
Subscription Rates: United States
$38.00 per year, 3-year rate $90.00.
Foreign $53.00 per year, 3-year rate
$131.00. Single copy $7.00.
Ex-Officlo
William C. Richardson, Jr.
William N. Avard
PCI JOURNAL STAFF

Editor-in-Chief
George D. Nasser
Secretary
Phillip J. Iverson
Kenneth Baur
Ned M. Cleland
Thomas J. DArcy
S. K. Ghosh
Simon Harton
Michael W. Lanier
Donald R. Logan
Todd G. McCoy
Guillermo Mecalco
Frank A. Nadeau
George 0. Nasser
Michael G. Oliva
Andrew E. Osborn
Chuck Prussack
Donald C. Raths
A. Fattah Shaikh
Clark Weber
Ex-Officio
FIB Representative
Charles W. Wilson
PCI STAFF
President
Thomas B. Battles
Technical Director
Phillip J. Iverson
Information Officer
John A. Lishamer
Research Director
L. S. (Paul) Johal
Architectural Director
Sidney Freedman
Assistant Editor
Karen Banasiak
Administration
and Finance Director
Gary H. Munstermann
EditorIal Assistant
Susan Bowden
Marketing Director
Brian Goodmiller
Layout Artist
Karen Marie Rokos
Structures Director
John S. Dick
Cover Design
Leader Graphic Design, Inc.
Plant Certification Director

David S. Jablonski
POSTMASTER: Please send address changes to PCI JOURNAL, 209 West Jackson Boulevard, Suite 500, Chicago, Illinois 60606.
Periodicals postage rates paid at Chicago and additional mailing office.
PCI JOURNAL
WHATS NEW
AT JVI?
7550 North Linder Avenue

Fax 847/675-0083 1-800-742-8127
Skokie, Illinois 60077 USA 847/675-1560

E-mail: jvi-inc.com http://www.jvi-inc.com
BMH SYSTEMS Custoifi Concrete

1ts
All BMH technologies and products have been developed along with the evolving
of the industry. Color dispensing systems, color blending, twin mixer systems for core
mix and face mix that prevent any possible contamination are only a few examples
of what BMH can offer its customers. In either vertical plants or low-profile installa
tions, BMH Systems delivers customized plants that fit any machine, any capacity,
and any building requirement.
Concrete Products Plants

Precast Plants
Block Plants
Paver Plants
Pipe Plants
1 I
Using only the best materials, we directly control each and every step of the project,
from design and specifications to manufacturing, installation and commissioning of
the plant. For more than 20 years, this is what has made the difference between
simply fair results and exceptional performance.
BMH SYSTEMS
www.bmhsy,temi. cm
r,i. (450) 3494770/f- (450) 449-4898
Emih bmhf4bmh.cytcms.com
PCI JOURNAL
:
w
How to Save Money When You

Produce Your Next Parking Garage
Your customers count on you for reliability, value,
and aesthetics, the same benefits you get from Producer
Solutions accessories and embedments.
For a typical parking garage project with three elevated
levels and 175,000 square feet, the following Producer
Solutions products would be used:
Cost Comparison Per Piece
$46 53
Tt[
900 Double Tee Stem Blockouts

155 Spandrel-to-Column Connector
Sleeves arid Closure Caps
320 Grouted Connection Tubes
1,200 Swift Lift Covers
ER
hiflinhi
Do.vble Toe
Spaevleel-ln-Colomn
Greeted
Stem Bluckout
Coenector Sleeve Connection Tube
Swift-Lift
Cover
HIGH CONCRETE ACCESSORIES
on
125 Denver Rood
Denver, PA 17517
(608) 508-2583
FAX (717) 336-9301
entoil. kiddengs@t1.96.nel
www.producerscdutions.com
lnclodcs Coot of Field Palchrng

eulneltebon Cost of Field Welding
[]Producer Solutions
A AotOn olhi6, cotuem.
Alternatives
1
Q UA L...LT...Y.-
..MO NEY
b%i
Ratec provides everything that makes a
modern precast concrete plant more efficient. Intelligent, tried-and-tested systems
and components that effectively rationa

lise your siderail technique and produc
tion runs. Contact us for your copy of
our Ratec catalog today or arrange a per

sonal appointment. This will be a decisi
on that is sure to prove worthwhile.
As clever as concrete!
the
Europ3e: RATEC GmbH + Co. Karisruher Strasse 32 68766 Hockenheim/Germany

Phone +49 (0) 6205-9407-29 Fax +49 (0) 6205-9407-30 E-mail: info@ratec.org Internet: www.ratec.org
North America: RATEC LLC. 250 Julia Circle North St. Petersburg, FL 33706
Phone (877)33 RATEC or (727) 363-7732 Fax (727) 363-7463 E-mail: infous@ratec.org Internet: www.ratec.org
u!ir@j1i YOU OAN
ON
i
P7i
PUT CONCRETE WHERE YOU WANT IT! WHEN YOU WANT IT

The T-630 is a Strategic Concrete Placement Vehicle that can pour
into even the most complex forms without spillage. The T-630
Concrete Delivery Vehicle is gaining worldwide acceptance in the
prestress/precast industries.
PRODUCTS YOU CAN DEPEND ON

WEVE GOT YOU COVERED
NOW IS
THE TIME
to start thinking
about all of that money
escaping from your
beds this winter.
Catch the savings with
P.S..
CURING
COVERS
IPI
PRESTRESS SUPPLY INC.
1804 W. Lake Parker Drive Lakeland, Florida 33805 863-683-4492

1-800-328-8036 FLORIDA 1-800-282-7321 Fax 863-683-2886
LESPQWR IPI
for Prestressing
The PSI Hercules Single Strand Tensioning unit is over

designed to assure many years of reliable service. Many
Hercules Jacks are still in service after 20 years. With a
working capacity of 52,000 lbs., the Hercules Jack will pull from
4mm up to .600 strand.
NEW FEATURES INCLUDE

Retractable wheel dolly for improved handling
Foam filled tires, no more flats
English / Spanish user manuals
Gauges read in U.S. and Metric
Optional loadcell with readout recorder instantly
measures actual prestressing force
-
4?
Dfl
-/ALI1
WLL\
UL *p4hucks
SuperChucks are made of special high tensile alloy

steel. The unique manufacturing process insures
dimensional accuracy and high resistance to wear.
Jaw assemblies are also from specially formulated
steel. They are tested for hardness and tooth design
plus tensile tested to assure the highest quality of
product.
IPI
PRESTRESS SUPPLY INC.
1804 W. Lake Parker Drive Lakeland, Florida 33805 863-683-4492

1-800-328-8036 FLORIDA 1-800-282-7321 Fax 863-683-2886
P01 Deiign Ilcindbook

flfth Edition
his indispensable publication provides the designer with comprehensive and
efficient procedures for the safe design of both architectural and structural pre
T cast and prestressed concrete products. The design information and recom
mendations contained in the Fifth Edition of the Handbook are based not only on the
latest research but also on a consensus of engineers in practice. The topics in the
Handbook are illustrated with numerous design examples that are in conformance
with AC! 318-95.
The Fifth Edition also includes new
design recommendations on:
* Structural Integrity
* Headed Stud Connection Design
* Pocketed Spandrel Panels
* Seismic Design
* Slender Walls
PRE CAST/PRESTRES S ED
CONCRETE INSTITUTE
and much more!
209 W. Jackson Blvd

Chicago, IL 60606
Phone: 312/786-0300
Fax: 312/786-0353
http://wwwjpci.org
Also available is the Background and

Discussion of the PCI Handbook, Fifth
Edition reprinted from the PCI JOUR
NAL. This valuable resource is avail
able for just $7.00 ($14 for non-PCI
members.)!
$85.00 for PCI member,

Shipping Information
$95.00 for non-member Engineeri
Name
Handbook(s) will be shipped as soon as

possible
Book Rate by U.S. Postal
Service unless otherwise requested.
Shipping and handling charges are as fol
lows (check one):
Firm
Address
(If P0. Box, also give street address for shipping.)
City
Shipment
Charge
To U.S. addresses:
J Book Rate
ci pays
First class/UPS
$10/book
Li Next day service.. $30/book
4-6 weeks
ASAP
Overnight
Outside U.S.:
Li Surface rate
LiAir Mail
2-4 months
ASAP
....
$30/book
$75/book
Zip
State
Country
Delivery
Telephone
All payments must be U.S. denominated

and drawn on a U.S. bank.
Fax
(required)
Form of payment(Payment must accompany order):
Visa/MasterCard(on/y)
Credit Card #
Exp. Date
# of books_____
Price per book
# of copies of Commentary_____
Price per copy
Illinois residents add 8.75% sales tax
Shipping and handling charges
Total
Check
CHAIRMANS MES SAGE
William E. Whitcher
Executive Vice President and
General Manager
Coreslab Structures (Miami) Inc.
Medley, Florida
Proposed PCI Education Foundation

hen I assumed the PCI chairmanship a year ago, I
expressed both concern and hope that we could
address a major interest of mine, namely, reaching
out today to Americas youth that will emich and enhance our
industry tomorrow.
Earlier this year, as I outlined my goals for my tenure as
Chairman of PCI, I placed a special emphasis on our need to
bring students into the mainstream of the precastlprestressed
concrete industry. We have now taken a significant step toward
the achievement of that goal in the planned establishment of the
PCI Education Foundation.
Spurred on by our Associate Member Jim Voss of JVI, Inc.,
the new foundation will focus on several goals, all aimed at
making students become more aware of the advantages that
precast/prestressed concrete offers in the design and construc
tion of a wide variety of structures.
First, the foundation seeks to influence educators so that they
create and promote courses in the curricula of their respective
engineering schools emphasizing precast concrete. Far too
often, our industry gets only cursory attention from educators,
many who have little or no knowledge of our products and in
stead focus their educational efforts on competing materials.
A second goal is to encourage and nurture student participa
tion in our industry. Our hope is that this endeavor will lead
them to consider positions within the precast concrete industry
after they complete their schooling.
A third goal will be to support and subsidize selected areas in
the education arena to stimulate the study and enthusiastic ac
ceptance of precast concrete as a material to be considered in
their future professional careers.
The new foundation will in no way conflict with present PCI
educational activities. We seek instead to enhance efforts that are
already in place and work within the Institute. It remains my sin
cere hope that the PCI Student Education Committee continues its
high level of positive activity within the educational community.
Chairman Al Ericson, working with his Student Education
Committee, has produced a new $25,000 Engineering Competi
tion called the Big Beam Contest, co-sponsored by Sika Corpo
ration. Student teams will be encouraged to work with local
precasters to build and test a prestressed concrete beam. Our

RMDs will be involved and prizes will be awarded regionally
and nationally. In addition, Al has been working with our PCI
Marketing Department in producing a Career Paths Brochure
now in the development stages. He and his committee are tak
ing this initial draft design and are tweaking it to meet the needs
of the student, the committee and PCI.
Professor Norm Lach, from Southern Illinois University, is
starting a PCI Student Chapter and the Board at their recent meet
ing voted to amend the PCI Bylaws to allow this membership
category. Other important education outreach programs include
educational subsidies of free Design Handbooks, Hollow-Core
Slab, Bridge Design and Architectural manuals being disthbuted
to students taking courses in architectural and prestressed con
crete.
Currently, there is a joint education project with our Geor
gia/Carolina region where PCI is planning to hold a PCI Educa
tor/Practitioner Conference next summer at the University of
North Carolina.
Also, Professor Eric Steinberg has a student team from Ohio
University working with PCI staff to build an image bank, i.e.,
a library of photos and other artwork taken from Ascent, the
PCI JOURNAL and the PCI headquarters library for use by Re
gional Marketing directors and members.
Finally, our PCI Marketing and Student Education commit
tees have created a business card-sized CD ROM. In this small
30-meg. CD, a fully animated and musically scored presenta
tion offers students an overview of the excitement of the pre
castJprestressed industry. It will be distributed nationwide along
with the new Career Paths brochure.
It is gratifying to witness how the PCI committees and staff
have responded to our goal of reaching out to the youth of
America that will enrich our industry. To all who have partici
pated, I offer my sincere thanks and a hearty.. .WeIl Done!
13
2000 PCI Design Awards Program

The Winning Entries
ver 100 entries were judged for
the 2000 PCI Design Awards
Program, held May 19 in
Chicago, Illinois. In the end, 17 entries
received top honors. Entrants needed
to show how they creatively used pre
cast, prestressed concrete, glass fiber
reinforced concrete and/or architec
tural precast concrete in their struc
ture. Structures must have been com
pleted within the past three years to be
eligible in the program and have been
built in the United States, Canada and
Mexico.
There were five categories, each
with several subcategories. The
Commercial Buildings Category in
cluded office buildings, retail stores,
mixed-use structures, public build
ings, institutional buildings, schools
and hospitals.
Warehouses, manufacturing facili
ties, food processing facilities, distri
bution centers and special industrial
structures were included in the Indus
trial Structures Category.
The Housing Structures Category
covered single-family homes, multi
family structures, hotels, motels, con
dominiums, retirement facilities, dor
mitories and assisted living centers.
The Specialized Structures Category
included parking structures, stadiums,
arenas, sports centers, tanks, justice
facilities, jails, prisons and correc
tional facilities.
The Transportation Structures Cate
gory is also divided into subcategories to
recognize the diversity of design possi
ble using precast and prestressed con
crete. The eight subcategories of this
category are: soundwall projects, piling
projects, rail depots, bridges with spans
less than 65 ft (20 m); bridges with span
lengths between 65 to 135 ft (20 to
41 m); bridges with span lengths greater
than 135 ft (41 m); rehabilitated bridges;

and non-highway bridges.
As in the past, there are no first
place winners. Instead, each winning
project is given equal recognition in
order to acknowledge the diversity of
design solutions used by architects and
engineers.
In addition, two entries received the
Harry H. Edwards Industry Advance
ment Award, which is given for those
ideas and concepts that hold the poten
tial to move the industry to the next
generation of technology. A separate
jury judged the projects considered for
this award.
PCI extends its gratitude to each
member of the 2000 Jury of Awards.
Jury members for the Commercial
Buildings, Industrial Structures,
Housing Structures and Specialized
Structures categories were: (Co
Chairwoman) Donna Robertson,
Dean of the College of Architecture,
Illinois Institute of Technology; (Co
Chairman) Eliseo Temprano, Imme
diate Past President, RAIC; Degan G.
liambacher, president, FDG, Inc.;
and Chris Sullivan, editor-in-chief,
Building Design & Construction.
Jury members for the Transporta
tion Structures category were: (Chair
man) David H. Densmore, Chief of
Bridge Division, Federal Highway

Administration; Richard A. Miller,
Associate Professor, University of
Cincinnati; and William N. Nickas,
State Structure Design Engineer,
Florida Department of Transporta
tion.
Jury members for the Harry H. Ed
wards Industry Advancement Award
were: (Chairman) Harry A. Gleich,
Vice President of Engineering,
Metromont Prestress Co.; and Gerald
E. Goettsche, President, The Con
sulting Engineers Group, Inc.
Award winners received a special
recognition at PCIs 46th Annual
Convention & Exhibition in Orlando,
Florida, during the Design Awards
Dinner/Dance on Tuesday, September
26. Summary descriptions of the winfling projects appear in the Fall 2000
issue of ASCENT Magazine. In addi
tion, winning entries will receive ex
tensive national publicity through
major industry and architectural pub
lications. The PCI JOURNAL will
feature articles about some of the win
ning entries and other noteworthy en
tries throughout the year.
The credits for the winning projects
are listed on the following pages. PCI
extends congratulations to all those in
volved with these exemplary projects.
he purpose of the Design Awards Program is to recognize design

excellence in the use of precast, prestressed concrete and/or arch i
tectural precast concrete. The annual program is open to all registered
architects and engineers practicing professionally or in government
agencies in North America. Winners are chosen on the basis of excep
tional achievement in aesthetic expression, function and economy, and
ingenuity in the use of materials, methods and equipment. Please visit
our website at www.pci.org for more in formation.
PCI JOURNAL
2000 Design Awards Program Judges
David H. Densmore
Gerald E. Goettsche
Donna Robertson
William N. Nickas
Degan G. Hambacher
Chris Sullivan
Harry A. Gleich
Richard A. Miller
15
Winner
Winner
Correctional Facility (Low-Rise)
Winner
Custom Solution
Correctional Facility (High-Rise)
David L. Moss Criminal

Justice Center
Tulsa, Oklahoma
Federal Metropolitan
Detention Center
Philadelphia, Pennsylvania
The Terraces at Riverfront

Recapture
Hartford, Connecticut
Entrant and Architect:

BKL, Inc./HDR Architecture, Inc.
Tulsa, Oklahoma
Entrant, Engineer and Architect:

Ewing Cole Cherry Brott
Entrant and Engineer:

Brecher Associates P.C.
Engineer:
BKL, Inc.
Tulsa, Oklahoma
Precast Concrete Manufacturer:

High Concrete Structures
Denver, Pennsylvania
General Contractor:
Keating Building Corp.
Bala Cynwyd, Pennsylvania
Owner:
U.S. Department of Justice
Federal Bureau of Prisons
Washington, D.C.
Architects:
Carol R. Johnson Associates
Cambridge, Massachusetts

Coreslab Structures (Oklahoma), Inc.
Oklahoma City, Oklahoma
General Contractor:
Hensel Phelps Construction Co.
Cape Canaveral, Florida
Owner:
Tulsa County Criminal Justice Authority
Tulsa, Oklahoma
Goddes Bicher Quails & Cunningham, Inc.

Blakeslee Prestress, Inc.
Branford, Connecticut
General Contractor:
0 & G Industries, Inc.
Torrington, Connecticut
Project Manager:
Connecticut Department of Transportation
Newington, Connecticut
Owner:
Riverfront Recapture, Inc.
Hartford, Connecticut
Winner
Winner
Hotel
Office Building (Low-Rise)
Winner
Office Building (High-Rise)
Hilton Boston-Logan Airport

Boston, Massachusetts
Merrill Lynch Facility

Englewood, Colorado
IJL Financial Center

Charlotte, North Carolina*

Cambridge Seven Associates, Inc.

Thompson, Ventulett, Stainback &
Associates
Atlanta, Georgia

Smaliwood, Reynolds, Stewart,
Stewart & Associates, Inc.
Atlanta, Georgia
Engineer:
W.P. Moore and Associates, Inc.
Atlanta, Georgia
Engineer:
Stanley D. Lindsey & Associates Ltd.
Atlanta, Georgia

Rocky Mountain Prestress, Inc.
Denver, Colorado

Metromont Prestress Co.
Greenville, South Carolina
General Contractor:
Huber, Hunt & Nichols, Inc.
Branch Burg, New Jersey
General Contractor:
Beers Construction Co.
Atlanta, Georgia
Owner:
Merrill Lynch Denver Holdings, Inc.
Plainsboro, New Jersey
Owner:
Bank of America
Charlotte, North Carolina
Engineer:
Weidlinger Associates, Inc.
Precast Concrete Manufacturers:
Bton Prfabriques Du Lac, Inc.
Alma, Quebec, Canada
Northeast Concrete Products LLC
Plainville, Massachusetts
General Contractor:
Beacon Skanska Construction Co.
East Boston, Massachusetts
Owner:
Hilton Hotels Corporation
Beverly Hills, California
16
*A detailed article on this project was

published in the May-June 2000
PCI JOURNAL.
PCI JOURNAL
Co-Winner
Co-Winner
Parking Structure
Winner
Parking Structure
Public Building
MacArthur Center North and

South Parking Structures
Norfolk, Virginia
Sun Trust Plaza Parking

Structure
Winter Park, Florida
Center of Science &

Industry (COSI)
Columbus, Ohio

Hobbs & Black Associates, Inc.
Ann Arbor, Michigan

Rich & Associates, Inc.
Southfield, Michigan
Entrant and Lead Architect:

NBBJ Architects, Inc.
Columbus, Ohio
Design Architect:
BrownlMcDaniellBhandari, Inc.
San Francisco, California
Design Architect:
RTKL Associates, Inc.
Baltimore, Maryland
Design Architect:
Arata Isozaki & Associates
Tokyo, Japan
Engineer:
E & S Mechanical Services
Pasadena, Maryland
Engineer:
Rich and Associates, Inc.
Tampa, Florida
Engineer:
Korda-Nemeth Engineering, Inc.
Columbus, Ohio

Tindall Corporation Virginia Division
Petersburg, Virginia

Finfrock Industries, Inc.
Orlando, Florida
Gate Concrete Products Company

Oxford, North Carolina
General Contractor:
Jack Jennings & Sons
Orlando, Florida

Concrete Technology, Inc.
Springboro, Ohio
Marietta Structures Corp.
Marietta, Ohio
General Contractor:
Sordoni Skanska Construction Co.
Parsippany, New Jersey
Owners (Joint Venture):
Taubman Realty Group
Bloomfield Hills, Michigan
Owner:
Rollins College
Winter Park, Florida
Construction Manager:
Ruscilli Construction Co.
Columbus, Ohio
Owner:
Center of Science & Industry
Columbus, Ohio
The Conroy Development Company

Greenwich, Connecticut
Winner
Winner
Retail Building
Winner
Stadium
Warehouse
Lazarus Department Store

Pittsburgh, Pennsylvania*
Princeton University Stadium

Princeton, New Jersey
Park Fletcher Building #40

Indianapolis, Indiana

Cooper Carry
Atlanta, Georgia

Raphael Vifloly Architects
Minneapolis, Minnesota

Ratio Architects, Inc.
Structural Engineer:
Structural Engineering Corp.
Pittsburgh, Pennsylvania
Engineer:
Thoruton-Tomasetti Engineers
New York, New York
Engineer:
Smith Roberts & Associates

Modern Mosaic Ltd.
Niagara Falls, Ontario, Canada
General Contractor:
Turner Construction Co.
Pittsburgh, Pennsylvania
Owner:
Federated Department Stores
Cincinnati, Ohio

General Contractor:

Spancrete Industries, Inc.
Waukesha, Wisconsin
General Contractor:
Duke-Weeks Realty Corp.
Owner:
Princeton University
Owner:
Duke-Weeks Realty Corp.
*Adetailed article on this project appears

in this issue of the PCI JOURNAL.
tFor more information on these projects,

see articles in the September-October and
May-June 1999 issues of the PCI JOURNAL,
respectively.
17
Winner
Bridge Span
Less than 65 ft (20 m)
Winner
Winner
Bridge Span
of 65 to 135 ft (20 to 41 m)
Bridge Span
Greater than 135 ft (41 m)
U.S Route 3A over

Newfound River
Bristol, New Hampshire
King Avenue over the

Olentangy River
Columbus, Ohio
H-3!Lower Halawa Valley

Viaduct; Unit 1, Phase 1 B
Aiea, Hawaii

New Hampshire Department of
Transportation
Concord, New Hampshire

HNTB Corporation
Kansas City, Missouri

T.Y. Lin International
San Francisco, California
Engineer:
Eriksson Engineering
Columbus, Ohio

Con-Fab California Corp.
Lathrop, California

Tecspan Concrete Structures, Inc.
Grove City, Ohio
General Contractor:
Kiewit Pacific Co.
Vancouver, Washington
General Contractor:
C.J. Mahan Construction
Grove City, Ohio
Owner:
Hawaii Department of Transportation
Honolulu, Hawaii

Northeast Concrete Products LLC
General Contractor:
R.S. Audley Construction
Bow, New Hampshire
Owner:
New Hampshire Department of
Transportation
Concord, New Hampshire
Winner
Owner:
Franklin County, Ohio
Columbus, Ohio
Winner
Non-Highway Bridge
Hotel
Metropolitan line B Stations

Mexico City, Mexico
Route 7 Bridge over Route 50

Fairfax County, Virginia
Entrant, Engineer and Architect:

Rioboo, S.A. de C.V.
Mexico City, Mexico

Wilbur Smith Associates
Falls Church, Virginia

Pretencreto
Mexico City, Mexico

Bayshore Concrete Products Corp.
Cape Charles, Virginia
General Contractor:
Ingenieros Civiles Asociados (ICA)
Mexico City, Mexico
General Contractor:
Corman Construction
Annapolis, Maryland
Owner:
Secretaria de Obras de Distrito Federal
Mexico City, Mexico
Owner:
Virginia Department of Transportation
Fairfax, Virginia
PCI JOURNAL
Harry H. Edwards Industry Advancement

Award Winners
Public Building
Stadium
Center of Science &

Industry (COSI)
Princeton University Stadium

Princeton, New Jersey*
Columbus, OIiio*

Raphael Violy Architects
Minneapolis, Minnesota
Entrant and Lead Architect:

NBBJ Architects, Inc.
Columbus, Ohio
Design Architect:
Arata Isozaki & Associates
Tokyo, Japan
Engineer:
Korda-Nemeth Engineering, Inc.
Columbus, Ohio
Precast Concrete Manufacturer
Springboro, Ohio
Marietta Structures Corp.
Marietta, Ohio
Engineer:
Thornton-Tomasetti Engineers
New York, New York
General Contractor:
Owner:
Princeton University
Construction Manager:
Ruscilli Construction Co.
Columbus, Ohio
Owner:
Center of Science & Industry

Columbus, Ohio
*For more information on these projects,

see articles in September-October and
May-June 1999 PCI JOURNAL,
respectively.
19
COVER FEATURE
Precast Concrete Panels Give

Scale and Grandeur to Lazarus
Department Store
Gar Muse, MA
Principal
Cooper Carry Inc.
Atlanta, Georgia
Precast concrete played a prominent role in building the new multimillion dollar Lazarus Department Store in dowtown Pittsburgh,
Pennsylvania. The glitzy department store is four stories high with
an underground three-level parking structure. Highly articulated
precast concrete panels with deep reveals give the building scale
and character, and also complement the nearby historic buildings.
In addition, the use of granite at the base provides richness and
depth of color. These features enhance the faades verticality and
contemporary look. To speed construction, structural precast
concrete columns are used in the parking structure. This article
presents the conceptual and architectural design features of the
building, design considerations, and erection highlights of the project.
ittsburghs downtown Fifth

Avenue shopping district is
undergoing a major revital
ization, spearheaded by the citys
own development activities and the
deep involvement of a variety of re
tailers. A centerpiece in this reno
vation program is the Lazarus De
partment Store, a new four-story,
) retail cen
2
247,000 sq ft (23000 m
ter and three-level underground
parking facility. The new structure
blends very well with the nearby
historic buildings while offering a
distinctly contemporary feeling in
Anthony Di Giacomo
Vice President
Modern Mosaic Ltd.
Niagara Falls, Ontario
Canada
20
its character and tone (see Fig.1).

To achieve that combination, the
design team took advantage of the
many benefits offered by architec
tural precast concrete panels to clad
the structure and precast concrete
structural columns to support the
parking structure. The result of their
efforts was not only an impressive
and successful addition to the down
town area but the building won the
award for the Best Retail Facility in
the 2000 Design Awards Competi
tion sponsored by the Precast/Pre
stressed Concrete Institute.
PCI JOURNAL
Fig. 1.The
Lazarus
Department
store in
downtown
Pittsburgh
features a
three-layer
look on its
faade and a
strong
entrance
tower aligned
with the
intersection.
The project was designed and built

following the requirements of two
distinct clients. The Pittsburgh Eco
nomic Industrial Development Cor
poratiori (PEIDC) owns the site and
served as the impetus for the con
struction. PEIDC in turn leased the
building to Federated Department
Stores to operate a Lazarus retail cen
ter in the building. As a result, both
groups provided input into the de
sign, with the PEIDC focused primar
ily on the structure and the under
ground parking facility, and Lazarus
working with the team on the store
design. Figs. 2 and 3 show a plan, el

evation and various sections of the
building.
The site selected by PEIDC is
within the citys main downtown
shopping district and required the de
molition of eight existing buildings to
clear the site for construction of the
new building. The availability of this
site convinced Federated Department
Stores to agree to build a new down
town store; however, once site work
began, demolition and foundation
preparation showed that a retaining
wall was required before the new
construction could begin, causing an

unforeseen delay that threatened to
slow down the project substantially.
Timing was a critical issue for the
retailer, as the project had to be com
pleted before the Christmas holiday
season, when the company makes a
significant percentage of its annual
sales and profit. Because the demoli
tion and preparation work delayed the
time when the site could be turned
over to the general contractor to
begin construction, the contractor
suggested converting the under
ground parking structures structural
21
Fig. 2. North and south plan of perimeter of building showing location of precast panels.
2,
KEY PLAN
CD
Di
ooo .ieqopo-ieqwedes
C,
C-)
-1
0
z
C
D
U,
Di
-o
CD
Di
C
CCD
Di
(1)
C)
I
C
CD
Di
U,
Di
CD
U,
T
\
1,
I-
><
/<<___
I
I I I .._I I I I
I
r
a)
m
C)
0
1
C,)
-l
0
C
C
-Ti
-n
I
0
0
11
I
C
7
I
0
0
0
0
I
11
0
0
0
0
-Ti
0
0
0
Fig. 4. Architects rendering of building faade showing three layers and entry tower.
Fig. 5. Rustication and textures on the precast panels help the luxury store fit in with
its neighboring buildings, built in the 1 900s.
24
components to precast concrete

rather than using cast-in-place con
struction. The parking facility con
tains 192,700 sq ft (17920 m
) of ve
2
hicular space on three floor levels.
Using precast concrete columns
saved approximately one month out
of the projects construction sched
ule, because the precast scheme al
lowed for the implementation of an
up-down construction approach.
Originally, the plan was to use castin-place columns, install the floor on
the upper level of parking and build
the retail store with a steel frame
work on top of that base.
With the change to precast con
crete columns, the underground park
ing structure could be built in mod
ules, with construction on the retail
center progressing as each parking
bay was completed. Erection of the
precast columns was performed with
the crane sitting on a 30 ft (9.14 m)
deep foundation and working its way
backward across the site.
One end of the site was secured
with a retaining wall, with the foun
dations for the precast columns set
and the precast columns lifted into
PCI JOURNAL
place on the footings. Then, the crane

moved backward across the hole to
set more columns. Once the bay was
completed, steel crews began erecting
the steel framework above. Thus,
while the final parking levels were
being completed, the initial portions
of the retail store above were already
underway.
Although their primary concern was
the structure and parking facility, city
planners had considerable influence on
the overall design of the retail portions
as well. In particular, they wanted to
ensure the buildings exterior look was
appropriate for its setting in the retail
area and complemented the surround
ing older buildings. The design team
worked closely with the citys Design
Review Committee and Urban Plan
ning Department to find a design that
fit those needs.
In fact, the designers spent consider
able time working with city planners
and the head of the PEIDC, who pro
vided a tour of the downtown neigh
borhoods to explain the ambience and
design goals they wanted to reach. The
designers did not actually begin creat
ing the buildings final look until they
had been working on the project for
three months. That lag time gave the
designers time to sort out all the vari
ables involved in the design process
and think through how best to meet the
varying needs of the participants.
In determining the final approach,
the designers took note of what the
area stood for and the statement made
by other buildings surrounding the site.
These buildings included the citys
courthouse, the Koppers Building, and
others built at the turn of the century
using heavily rusticated stone. The de
sign teams goal was to provide a
faade that blended with these designs
without incurring the high cost that
using real stone would have meant.
This goal was the reason why architec
tural precast concrete panels with stone
accents were specified.
Often, for this type of retail design,
Cooper Carrys designers furnish a
faade incorporating an alternate insu
lating finish system in conjunction
with precast concrete panels. The ar
chitect has worked on a variety of pro
jects for Federated Department Stores,
and the retailer had in fact suggested to
the PEIDC that its planners interview

the architect, leading to the commis
sion for this project.
In this instance, however, an all-pre
cast design was recommended to take
advantage of the contemporary sophis
tication of an established urban context
and blend it closely with the environ
ment. The goal was to have the build
ing appear to be crafted much as the
stone buildings around it had been.
With the precast concrete panels, the
architect could design scaled-down
modules with rustication joints, so the

retail center would feel like a building
made up of components rather than one
monolithic structure (see Figs. 4 and 5).
Connections for the panels were
fairly conventional, with loads handled
by the first-floor retaining wall along
the parking structure and the steel
frame used for bracing on the gravity
connections (see Figs. 6 and 7). Panel
sizes for the most part corresponded to
floor-to-floor heights. This meant that
first-floor panels were 20 ft (6.10 m)
1fl
1fl
4
-
.4
z:::zzz J
W 14 CCL.
PLAN
Fig. 6. Plan of typical panel connection with steel column.
SECTION
Fig.
7.
PLAN
Section and plan of typical panel connection with steel frame.

25
Fig. 8. The outer plane of panels is lighter in tone and serves

to frame the tall window bays.
high while panels for upper floors

were 18 ft (5.49 m) tall.
Within those sizes, the panels could
vary by 1 ft (0.305 m) in cross section
to achieve the thickness of the plane
desired. The top component at the
tower consists of 5/2 ft (1.68 m) tall
panels offset by 212 ft (0.76 m) to meet
the projection of the interior plane. The
size made this precast component
slightly more difficult to cast with uni
form consistency and erect at the top
of the structure.
The scaling of the rustication al
lowed the faade to address the other
buildings in the area. A twist, however,
was added to that design with the lay
ering of three planes of panels across
the faade and two textures achieved
within one mix to allow the planes to
stand out from one another. The three
layers help emphasize the verticality of
the four-story structure without adding
significant depth to the faade.
26
Fig. 9. The second plane of panels features a light sandblast

that contrasts with the other layers.
Each plane is about 1 ft (0.305 m)

deep, adding only 3 ft (0.91 m) from
front to back. But the ways in which
they weave together the three planes
and emphasize the buildings architec
tural features add considerable visual
interest while maintaining the classic
design look desired to match with the
surrounding buildings.
The outermost layer of panels frame
the tall bands of windows along both
streets, four along one side and five
along the other (see Fig. 8). These
panels and tall windows provide the
vertical emphasis that was desired.
This outer layer uses Cedar Lake ag
gregate and limestone in the mix and
the panel was given a smooth, acidetched finish that emphasizes the
lighter shade.
The middle plane of panels connects
the ones behind and before it by using
a darker shade of panels. This was
achieved by lightly sandblasting the
panels but not changing the cement

mix used. The third plane of panels,
which forms the backdrop and base
for the projecting two layers in front,
connects to the adjoining entrance
tower and again features a lighter
shade with heavy rustication to break
up its mass.
The second planes darker shade
also was used as the main color and
texture for the entrance tower, which
was stepped back from the three-plane
main facades, giving it a distinctive
flair (see Fig. 9). The whiter tone
again was used to frame the tall band
of vertical windows stretching to the
top of this section.
The towers verticality was
emphasized by aluminum arms that
protrude above the window bands at
the top of each tower side, as well as
by the inclusion of added height with
a grille design to make it a focal point
(see Fig. 10). The entrances presence
PCI JOURNAL
at the street level was enhanced by

setting the doorways at the corner of
the intersection facing each street.
All three planes are tied together by
the rustication in the panels, which
match up from layer to layer and pro
vide continuity among the dimen
sional pieces. Shaping these precast
panels proved to be difficult due to
the depth of the planes. Some of the
panels required multi-stage casting to
achieve the illusion of planes of stone
set into each other and protruding
from the one behind while in fact
they were portions of the same panel.
To ensure the proper texture and
colorations, samples were produced
as a guide during the bidding process.
These then were used by the precaster
to produce strike-off panels for in
spection. The first iteration was deli
cately adjusted and two samples of a
final control panel were produced.
The precaster retained one sample to
ensure quality control while the other
panel was shipped to the site for com
parison with arriving panels.
Erecting the architectural precast
panels presented numerous chal
lenges for the erector. The major
problem concerned the need to per
form much of the work at night. Try
ing to use transits and levels to shoot
elevations and set the precast panels
to these lines and elevations was dif
ficult with these complicated panel
configurations. Often, what looked
fine at night would have to be read
justed once the day shift came onto
the job site and saw what it looked
like in daylight (see Fig. 11).
Another major challenge was that
the three sides of the building being
clad with architectural precast panels
faced busy streets in the downtown
area (see Fig. 12). This was compli
cated by the fact that the building had
not been back-filled. As a result, the
erector had to work around a hole sur
rounding the structure about 30 ft deep
and 15 ft wide (9.14 x 4.57 m).
In order to set up the crane, crews
had to have falsework engineered and
installed each time the crane had to be
moved to another location. When each
shift finished, crews had to take off
the outriggers, pick up the mats and
reinstall the fences so they would not
interfere with traffic the next day.
Fig. 10. The towers verticality was emphasized by adding aluminum projections at
the top of the building.
In addition, because the precast

components were being delivered
from the precasters plant in Niagara
Falls, Ontario, Canada, the trailers de
livering the pieces could not be
dropped and left at the site, because
their erection could not be timed that
precisely. Instead, a storage yard was
established where the panels could be
left and shuttled to the site as needed.
If a truck still had pieces remaining
on it at the end of the day, it had to be
shuttled to the storage yard and
brought back again for the next shift.
Construction of the project began
in June 1997 and was completed in

November 1998, right on schedule
for the holiday season. The total shell
cost for the retail facility was $16
million, while the underground park
ing structure cost $11 million. The ar
chitectural precast concrete contract
totaled $1.1 million and included 300
components, including 255 architec
tural precast panels.
The resulting project proved to be
well worth the time and extra care
caused by the unique design. The use
of tall expanses of glass not only em
phasize the verticality of the building
27
Fig. 11. Erection of the panels required both a day and night crew, with major
challenges arising from having to work in the busy downtown area.
but add a distinctive and built-in mar

keting element. These openings allow
passersby to peek into the store
throughout the day, while at night,
the light shining out through the glass
from inside creates excitement along
the neighboring streets and sidewalk
and provides a beacon of activity that
can be seen from several blocks away
(see Fig. 13).
PCIs Design Award judges were
impressed by the design, too. In award
ing the project the 2000 Award for
Best Retail Facility, they said This
project is a good example of a faade
that blends precast elements in a very
appealing way with metal and glass. It
also is very respectful of the fact that
the building is located in a downtown
streetscape with a need for a human
scale. It shows that precast concrete
can be used to create very delicate
members that allow for a lot of open
ness and visual penetration in a build
ing elevation. It is a big box done
right.
In retrospect, the design creates a
centerpiece for Pittsburghs shopping
district that illustrates how contempo
rary design can show its respect to the
classical past. The faade harmonizes
with the surrounding buildings while
bringing a new sophistication to an es
tablished urban context. This could
only have been achieved by using pre
cast concrete components in the ways
that were done, both to save precious
time in the underground parking struc
ture and to create a distinctive and
complementary stone-like design for
the faade that met the schedule and
budget constraints.
CREDITS
Fig. 12. The towers entrances align with the streets at the intersection where the
structure is located.
28
Owner: Federated Department Stores,

Cincinnati, Ohio
Architect: Cooper Carry Inc., Atlanta,
Georgia
Engineer: Structural Engineering
Corp., Pittsburgh, Pennsylvania
General Contractor: Turner Construc
tion Co., Pittsburgh, Pennsylvania
Precast/Prestressed Concrete Manufac
turer: Modem Mosaic Ltd., Niagara
Falls, Ontario, Canada
Precast Concrete Erector: Alpha Steel
Erectors Inc., Worthington, Pennsyl
vania
PCI JOURNAL
.
C.
Fig. 13. Bright interior lights highlight the entrance tower of the Lazarus Department Store.
29
PART 3
Central Artery/Tunnel Project:

Precast/Prestressed Structures
Span the Big Dig
Keith Donington, P.E.
Senior Bridge Engineer
Parsons Brinckerhoff Quade &
Douglas, Inc.
Precast, prestressed concrete is used at many

locations for a multitude of functions throughout
the multi-billion dollar Central Artery/Tunnel
Project in Boston, Massachusetts. This article
describes two bridges and a marine pier where
precast, prestressed concrete was selected as the
ideal material to span these structures.
he Broadway and Dorchester Avenue Bridges are

local street bridges located in South Boston, Mas
sachusetts, that utilize butted box beam construction,
while the Spectacle Island Pier in Boston Harbor uses but
ted double T beams. The latter two structures are located
in a marine environment where corrosion resistance and in
herent low-predicted maintenance costs were major factors
in selecting precast, prestressed concrete construction. This
article discusses the design-construction highlights of these
three structures.
Paul Towel!, P.E.

Senior Bridge Engineer
Douglas, Inc.
BROADWAY BRIDGE
Vijay Chandra, P.E.

Senior Vice President
Douglas, Inc.
New York, New York
30
The Broadway Bridge (see Fig. 1) is a nine-span continu

ous structure that carries local traffic to South Boston over
Amtrak railway tracks, a Massachusetts Bay Transportation
Authority railroad yard, and the Fort Point Channel. The
new structure opened to traffic in January 1999. It replaced
an existing historic, but obsolete, swing bridge on an align
ment south of the old alignment.
The 760 ft long x 75 ft wide (230 x 23 m) structure pro
vides two travel lanes in each direction and two sidewalks.
The substructure comprises multi-column cast-in-place
PCI JOURNAL
Fig. 1. Completed view of the Broadway Bridge, Boston, Massachusetts.
concrete bents supported on 14 in.

(0.36 m) square precast, prestressed
concrete piles, except for certain sup
ports founded on 3 and 4 ft (0.9 and
1.2 m) diameter drilled shafts where
space limitations prohibited the use of
driven piles with a typical footing.
The superstructure is comprised of
precast, prestressed concrete butted
box beams with a 5 in. (125 mm) castin-place concrete wearing course. This
topping is reinforced with steel mesh
and additional reinforcing bars over
the piers to provide continuity for live
loads.
Spans vary between 56 and 105 ft
(17 and 32 m). A typical span has six
teen 48 in. wide x 42 in. deep (1.2 x
1.1 m), concrete box beams placed
side-by-side with a mortar joint be
tween the beams. For the spans over
the railroad tracks, 33 in. (0.8 m) deep
concrete beams were used.
Transverse prestressing strand ties,
at the ends and quarter points of the
span, provide nominal transverse pre
compression of the mortar joints. To
provide a classic appearance, precast
concrete light supports topped with

double luminaires line both sides of
the bridge, as well as an architec
turally enhanced steel railing system.
The total cost of this bridge was $8
million.
DORCHESTER AVENUE
BRIDGE
The Dorchester Avenue Bridge (see
Fig. 2) is an eight-span structure
crossing Fort Point Channel, provid
ing access to Bostons main post of
fice. The new structure replaces an
existing structure on an alignment
east of the old alignment. The 550 ft
(167 m) long bridge varies in width
from 73 to 135 ft (22 to 41 m), pro
viding two travel lanes in each direc
tion and two sidewalks. Piers are
skewed at 45 degrees. The total cost
of this bridge was $7.2 million.
The bridge is very unusual in sev
eral aspects; for example, several
bridge columns are located above the
walls of the 1-90 immersed tube tun
nel. The first concrete immersed tube
sections were floated out from the

casting basin in January 2000 to rest
on drilled shafts that have been in
stalled in bedrock.
Other bridge columns located away
from the tunnel are 4 ft (1.2 m) in di
ameter, supported on 6 ft (1.8 m) di
ameter single drilled shafts. Ovalshaped columns supported on the
immersed tubes are all 2.7 ft x 4 ft (0.8
x 1.2 m). Furthermore, to accommo
date an S-curve alignment at the south
end of the bridge, the superstructure
almost doubles in width to 135 ft (41
m). The excess deck area created by
the 45-degree skew and S-curve is am
chitecturally treated and will be land
scaped due to its function as a pedes
than outlook to the channel.
Spans vary from 62 to 78.7 ft (19 to
24 m) and use a combination of 36 and
48 in. wide x 27 or 33 in. deep (0.9 and
1.2 x 0.7 or 0.8 m) precast, prestressed
concrete box beams. The eight-span
structure is made continuous for live
loads over one two-span and two
three-span units by reinforcing the 5
in. (125 mm) cast-in-place topping.
31
4(4
OOCHIIYfl #$AWCHfl
IARBOH WALI
CABOT COVE
PARK
MBTA
h.
.-.-
ENUEBRIDGE
.-
--
-FUTURE HARBOR WALK

GILLETTE
FORT POINT CHANNEL
I=72-O+i-
Fig. 2. Plan of Dorchester Avenue Bridge, Boston, Massachusetts.
The cross section of a typical span

is comprised of fifteen 48 in. wide x
33 in. deep (1.2 x 0.8 m) box beams
and four 36 in. wide x 33 in. deep (0.9
x 0.8 m) box beams placed side by
side with transverse post-tensioning
located at pier diaphragms and at
quarter span points.
Neoprene bearings are provided
under the ends of each beam and lon
gitudinal restraint is provided at inter
mediate piers by extending the cast-in-
32
place pier diaphragms into recesses

cast into the piers. Lateral restraint is
provided by shear blocks.
SPECTACLE ISLAND PIER

Spectacle Island is located in Boston
Harbor and has been filled and recon
figured with 3.7 million cu yd (2.8 mil
lion m
) of tunnel spoil from the Cen
3
tral Artery/Tunnel Project, becoming
the highest point in Boston Harbor.
The island is in the process of being

transformed from a landfill wasteland
to a public parkland as part of the pro
jects extensive environmental mitiga
tion program. A Visitors Center and 5
miles (8 1cm) of pathways will be con
structed, and landscaping will include
the planting of 28,000 trees and
shrubs.
To provide access to the island, the
Spectacle Island Pier (see Fig. 3) is a
marine trestle type structure designed
for the dual function of initially dock
ing construction barges and, finally, as
a public recreational boat dock after
the island park has been completed.
The structure is L-shaped, comprising
a 500 ft long x 40 ft wide (152 x 12
m) trestle that extends out from the
shore to a 270 ft long x 50 ft wide (82
x 15 m) boat dock.
As shown in Fig. 4, cast-in-place
pier caps are supported by 14 in. (0.36
m) square prestressed Concrete piles
with an 80-ton capacity in compres
sion and a 40-ton capacity in tension
(for barge impact forces). The caps are
made composite with the precast, pre
stressed concrete double-T superstruc
ture, which has a 4 in. (100 mm) castin-place mesh reinforced topping. The
beams are integral with the caps so as
to raise the caps as far above high tide
as possible.
PCI JOURNAL
Eight 6.25 ft (1.9 m) wide double-T

beams are provided for the docking
area. The stems vary from 6 in. (150
mm) at the bottom to 9 in. (230 mm) at
the underside of the 4 in. (100 mm)
thick top slab. The stems are pre
stressed using pairs of 0.5 in. (13 mm)
diameter strands; the lower strands are
debonded 8 ft (2.4 m) from the ends.
For the trestle, a similar design uses six
6.7 ft (2 m) wide beams.
Originally, the pier was designed as
an all timber structure. However, the
contractor hired consulting engineers
to completely redesign the structure in
precast concrete as part of a value en
gineering proposal. The construction
savings derived from switching from
timber to precast construction totaled
$285,000 for the $4 million structure.
To address corrosion concerns in a
marine environment, the contractor
hired a specialized corrosion consulting
firm. Corrosion mitigation measures in
cluded using pozzolanic additives and
epoxy coated reinforcing bars in all
concrete, and calcium nitrate in the
cast-in-place piers and superstructure.
The maximum water-cementitious
ratio was limited to 0.35 for the precast
piles and 0.40 for the other concrete
components.
ACKNOWLEDGMENTS
Client: Massachusetts Turnpike Au
thority, Boston, Massachusetts
Management Consultant:
Fig. 4. View of the precast, prestressed concrete double-T superstructure of the

Spectacle Island Pier, Boston, Massachusetts.
Bechtel/Parsons Brinckerhoff,
Section Design Consultants:
Maguire
Broadway
Bridge,
Harris/Lin Associates, Boston, Mas
sachusetts
Dorchester Avenue Bridge, Gannet
Flemming, Boston, MA
Spectacle Island Pier, Universal En
gineering, Boston, Massachusetts
Contractors:
Broadway Bridge, Keiwit/Cashman,
Dorchester Avenue Bridge Modern
Continental, Boston, Massachusetts
Spectacle Island Pier, CashmanlKei

wit/Perini/Atkinson, Boston, Mas
sachusetts
Precast Producers:
Broadway Bridge, Northeast Con
crete Products LLC, Plainville,
Massachusetts
Dorchester Avenue Bridge, North
east Concrete Products LLC,
Spectacle Island Pier, Northeast
Concrete Products LLC (piles),
Unistress Corporation (double T
beams), Pittsfield, Massachusetts
33
Proposed Revisions to 1997

NEHRP Recommended Provisions
for Seismic Regulations for
Precast Concrete Structures
Part 2 Seismic-Force-Resisting Systems
by
Neil M. Hawkins, Ph.D.
S. K. Ghosh, Ph.D.
Professor Emeritus
Department of Civil Engineering
University of Illinois at
Urbana-Champaign
Urbana, Illinois
President
S. K. Ghosh Associates, Inc.
Northbrook, Illinois
This is the second in a series of three papers discussing

significant modifications expected to be included in the 2000
NEHRP Provisions, dealing with the design of precast,
prestressed concrete seismic-force-resisting systems. These
modifications are expected to be part of the 2003 edition of
the International Building Code.
n the May-June 2000 issue of the PCI JOURNAL, the his

tory and development of the requirements of the NEHRP
(National Earthquake Hazards Reduction Program) Rec
ommended Provisions
1 for precast and prestressed concrete
structures were discussed.
In this article, the specifics of the requirements of the pro
posed 2000 NEHRP Recommended Provisions for seismic-
force-resisting frame and structural wall systems composed

of precast and prestressed concrete elements are given arid
the basis for those provisions is documented.
The revisions discussed in this article are related to Pro
posal 4-37 and the associated commentary. That proposal
and its commentary can be viewed on the internet at
www.bssconline.org.
34
PCI JOURNAL
ORGANIZATION OF
SEISMIC-FORCE-RESISTING
SYSTEM PROVISIONS
Discussed first is the organization
of seismic-force-resisting system
provisions.
Table 1 identifies the systems and
connections covered by the propbsed
provisions and shows how the provi
sions are organized. The information
concerning the systems is also pre
sented in Fig. 1. Concrete Structure
Design Requirements will appear in
Chapter 9 of the 2000 NEHRP Provi
sions. The 2000 NEHRP Provisions
will adopt ACT 318992 as the basic
reference document for concrete struc
tures in Section 9.1.
The NEHRP Provisions then amend
ACT 318-99 by inserting additional
provisions into, or revising the exist
ing provisions of, ACT 3 18-99. Bold
face numerals in Table 1 and Fig. 1
starting with the number 9 identify
specific provisions of the NEHRP Pro-
visions. Bold face numerals starting

with the number 21 identify specific
provisions inserted into ACT 318-99.
With one exception, all the seismicforce-resisting system requirements of
the 2000 NEHRP Provisions, relating
to precast and prestressed concrete,
are placed in a new Section 21.11
added to Chapter 21 of ACT 3 18-99.
In ACI 3 18-99, the seismic risk of a
region is described as low, moderate
or high. Chapter 21 contains specific
requirements for the design of con
crete structures in regions of high and
moderate seismic risk. Structures in
regions of low seismic risk need only
meet the requirements of Chapters 1
through 18 of ACT 3 18-99. In the
NEHRP Provisions, the applicability
of Chapter 21 requirements depends
not only on the seismic risk at the lo
cation of the structure but also on the
occupancy of the structure and the
characteristics of the soil on which it
is founded. In the NEHRP Provisions,
the three considerations are combined
in terms of Seismic Design Categories

(SDC) which are assigned letters A
through F.
ACT 318-99 recognizes SDCs A and
B as being equivalent to regions of
low seismic risk and needing only de
tailing that meets the requirements of
Chapters 1 through 18. Structures as
signed to SDC C are recognized as re
quiring detailing mandated for regions
of moderate seismic risk and struc
tures assigned to SDCs D, E and F re
quire detailing prescribed for regions
of high seismic risk. The relation be
tween the ACI designation of seismic
risk and the NEHRP Seismic Design
Categories is shown in Rows 1 and 2
of Table 1.
For precast and prestressed concrete
structures assigned to SDCs D, E and
F, special moment frames, special
structural walls, and Type Z connec
tions must be used in seismic-force-re
sisting systems. For the design and de
tailing of special moment frames or
special structural walls, the designer
Tabe 1. 2000 NEHRP Provisions for Precast Concrete Seismic-Force-Resisting Systems.

Seismic Risk Category
Seismicity
High
Moderate
Low
BandA
Seismic Design Category

System
F,E,andD
Moment Frame
Emulation
Intermediate Moment Frame

21.10.7
Special Moment
Frame Strong
Connections
21.11.5
No Requirement
Ductile Connections
Wet or Dry
21.11.3.1
-
Non-Emulative
Interconnected with
Dry Connections
21.11.4.1
Structural Walls
Emulation
Special Structural
Wall
Strong Connections
N/A
Ductile Connections
21.11.4.2
As Ordinary Wall
21.1
No Requirement
Interconnected with
Dry Connections
21.11.4.2
Non-Emulative
Connections
21.11.6
Connection Type Z 21.11.6.5

l.S=l.4S
2. Anchor for 1.3 Si,,
3. Strut and tie model for connection region design
4. Confinement iff > 0.7f
5. Ductility capacity> 8.0
Connection Type Y 21.11.6.4

l.S=l.25S
2. Anchor for 1.3 S,,.
3. Anchor directly connected to main
reinforcement
4. Ductility capacity > 4.0
No Requirement
35
Fig. 1. NEHRP 2000 requirements concerning precast concrete seismic systems.
can use either emulation procedures or

procedures that recognize the unique
properties of a structure composed of
interconnected precast elements (non
emulative).
For structures assigned to SDC C,
moment frames made from precast ele
ments must utilize, as a minimum, Type
Y connections. However, they may also
have the tougher Type Z connections if
the designer so chooses. Structural
walls composed from precast elements
can be designed as ordinary structural
walls per Chapters 1 through 18 of ACT
318-99 with the requirements of Chap
ter 16 superseding those of Chapter 14
and with Type Y connections, as a min
imum, between elements.
SPECIAL MOMENT FRAMES

For precast and/or prestressed con
crete special moment frames, hinge lo
cations (nonlinear action locations),
must be selected so that there is a strong
column/weak beam deformation mech
anism under seismic effects, regardless
of whether emulative or non-emulative
36
design procedures are used.

Emulative Design
Requirements for design of precast
and prestressed concrete special mo
ment frames, utilizing procedures and
details that result in a structure with a
behavior under seismic loading emulat
ing that of cast-in-place special moment
frames, were first introduced into the
NEHRP Provisions in the 1994 Edition.
Two design alternatives were provided
and those alternatives, with minor
changes, have been carried over into the
2000 NEHRP Provisions.
One procedure allows elements to be
joined using ductile connections and the
other allows elements to be joined using
strong connections. A strong connection
is designed to remain elastic while in
elastic action takes place away from the
connection. Because a strong connec
tion must not yield or slip, its design
strength in both flexure and shear must
be greater than the bending moment and
shear force, respectively, corresponding
to the development of probable flexural
or shear strengths at nonlinear action lo

cations.
Ductile connections, on the other
hand, have adequate nonlinear re
sponse characteristics, making it un
necessary to ensure nonlinear action
locations remote from connections.
Typical connection configurations are
shown in Fig. 2. Additional informa
tion on the behavior and design of pre
cast concrete structures using emula
tive procedures is contained in Refs. 3
through 6.
Ductile Connections
Where ele
ments are joined using ductile connec
tions, the aggregate interlock that is
present at hinge locations in monolithic
construction is unlikely to exist for pre
cast concrete construction. Therefore,
to prevent shear slip when the moment
acting at the hinge location is at its
maximum probable value of Mpr, the
co-existing shear must not exceed half
the sum of the nominal shear strengths,
Connection of all the connections at the
hinging section. The nominal shear
strength, V
, of the section where the
0
connection is made must also not be
PCI JOURNAL
Connection
Connection
7/
(a) Beam-to-Column
(b) Column-to-Beam
-4.-
7
Connecti
Connection
-4,(c) Beam-to-Beam
(d) Column-to-Column
Connection
Fig. 2.
Typical precast connection
configurations.
(d) Column-to-Footing
less than the shear strengths of the

members immediately adjacent to the
connection.
Individual connections must satisfy
Type Z connection requirements.
Those connections can be either wet
or dry. A wet connection is defined
as one that uses any of the splicing
methods (mechanical, welded or lap)
specified in Chapter 21 of ACT 318-99
to connect precast members or precast
and cast-in-place members, and uses
cast-in-place concrete or grout to fill
the splicing closure. One type of ductile
wet connection widely used in emula
tive design is the splice sleeve con
8 Other connections with simi
7
nection.
jar ductility capabilities have recently

become available or are under develop
ment.
Strong Connections
Where
strong connections are used, the nonlin
ear action location (center of the non
linear action region) must be no closer
to the near face of the strong connec
tion than half the member depth. Thus,
for a frame with strong connections at
the beam-to-column interface, rein
forcement details must result in beam
hinging no closer than half the beam
depth away from the column face (see
Fig. 3). Any strong connection located
outside of the middle half of the span of
the beam must be a wet connection un
less a dry connection can be justified

by approved cyclic test results.
Non-Emulative Design
Over the last decade many advances
have been made in our understanding
of the seismic behavior of precast! pre
stressed concrete frame structures, as a
US
9
result of the NIST,
12
PRESSS
1
3 and JAPAN-PRESSS re
4
search programs. Those advances have
made possible the provisional standard
ization by ACT
5 of acceptance criteria
for concrete special moment frames
based on validation testing. That provi
sional standard, together with the re
37
Nonlinear
Action Location
Precast
Precast Column
Strong Connection
h/2
\h,2
Nonlinear
Action Region
Fig. 3. Example showing nonlinear action region and location.
search advances, has made possible the

development of criteria for the design
of frames constructed from intercon
nected precast elements.
While criteria for such frames have
existed in the NEHRP Provisions since
1994, the previous criteria were in an
Appendix and contained penalties for
the use of precast concrete elements
compared to monolithic concrete ele

ments. Those penalties are eliminated
in the 2000 Provisions and the possible
behavioral benefits of using precast
construction are recognized.
The complexity of structural sys
tems, configurations and details possi
ble with the use of precast elements
requires:
1. The selection of functional and

compatible details for connections and
members that are reliable and can be
built with acceptable tolerances.
2. Experimental and analytical ver
ification of force-deformation rela
tionships for critical connections of
the proposed seismic-force-resisting
system.
3. Design of the building using
those force-deformation relationships
and recognizing the loading effects
likely to be imposed by the anticipated
ground motions.
Traditionally, designers have had
the flexibility to widely vary connec
tion details within prescribed code re
quirements. For non-emulative spe
cial moment frame design, that
flexibility is sharply curtailed be
cause experience shows that small
design changes can have marked ef
fects on the building response in an
earthquake. Thus, ACI/ITG Ti. 1-99
requires a prior development pro
gram, including both analytical and
experimental investigations of a pro
posed seismic-force-resisting system,
before any validation testing of criti
cal details of the generic frame is un
dertaken. That ACI Provisional Stan
dard requires that:
1. A minimum of one module of
(a) Interior One-Way Joint
(b) Exterior One-Way Joint
(c) Corner Joint
Fig. 4. Illustration of connection configuration.

38
PCI JOURNAL
each characteristic configuration of in

tersecting beams and columns in the
generic moment frame be subject to
validation testing (see Fig. 4).
2. That testing be conducted at a
scale large enough to capture the full
complexities and behavior of the ma
terials and load transfer mechanisms
in the prototype frame. Test modules
must be not less than one-third scale.
3. The first loading cycle applied to
the module be within the linear elastic
response range of the module.
4. Test modules be subject to a se
quence of displacement controlled
drift cycles of increasing magnitude
that are representative of the drift cy
cles expected under earthquake mo
tions (see Fig. 5). Testing is to be con
tinued until the drift ratio (see Fig. 6)
equals or exceeds 0.035.
For acceptance of the generic frame,
the nominal strength, E, must be de
veloped before the drift raio exceeds
the allowable story drift limitation of
the governing building code (Value B
in Fig. 7); and the characteristics of
the third complete cycle for each test
module, at a drift ratio not less than
0.035, must satisfy the following crite
ria:
27
2 0
.75
0.5
x
0
I I II
Fig. 5. Cyclic deformation history for validation testing.
1. The peak force for a given load

ing direction must be not less than
75 percent of the peak lateral load
for the same loading direction
(Value A in Fig. 7).
2. The relative energy dissipation

ratio, f3, must not be less than oneeighth. That ratio equals the area
within the hysteretic loop divided by
the areas of the circumscribing paral
Fig. 6.
Definition of drift
15
ratio.
(..
HcE
Hcc
Initial Position
Final Position
Drift Ratio 0
Jh
VOL
HAO
HAL
39
lelograms defined by the initial stiff

ness for each loading direction during
the first loading cycle and the peak
resistance for that same direction dur
ing the third cycle to a drift ratio of
0.035 (see Fig. 8).
3. The secant stiffness between drift
limits of -0.0035 and +0.0035 must be
not less than 0.05 times the initial
stiffness of the module for the first
loading cycle (see Fig. 9).
The studies that lead to the specifica
tion of a limiting drift ratio of 0.035
are documented in the Commentary to
ACIIITG Ti .1-99. Conventional castin-place special moment frames con
forming to Chapter 21 of ACT 318-99
will have limiting drift ratios approach
ing 0.035 but they may be unable to
achieve that limit on a consistent basis
for the range of properties found in
11 Thus, precast/prestressed
practice.
special moment frames in the 2000
NEHRP Provisions are being held to a
standard at least as high as that for
cast-in-place special moment frames.
Some strength degradation at high
cyclic-drift ratios is inevitable and the
degree of degradation that can be ex
pected will increase with increase in
the limiting drift ratio. A strength
Fig. 7.
Quantities used in
evaluating acceptance
criteria.
degradation of 25 percent is consistent

with analytical and experimental find
ings for a drift ratio of 0.035. For a
given earthquake motion, the maxi
mum drift experienced by a structure
increases as its relative energy dissipa
tion ratio decreases.
If the relative energy dissipation
ratio is less than one-eighth, oscilla
tions may continue for a considerable
time after the earthquake and low
cycle fatigue effects can result. If the
stiffness is very small around zero
drift ratio, the structure is prone to
large displacements for small lateral
force changes following a major earth
quake and is, therefore, vulnerable to
low cycle fatigue effects in after
shocks and moderate winds.
For precast/prestressed special
moment frames, the 2000 NEHRP
Provisions add three additional items
to the ACI/ITG Tl.1-99 criteria as
follows:
1. The test modules must be shown
to be able to continue to carry the
gravity loads that act on them in the
generic frame at a 0.05 drift ratio. This
requirement was considered necessary
to document that the precast/pre
stressed frame had a toughness equiv
alent to that anticipated for a cast-inplace concrete frame.

2. Unless there was substantial ex
perimental evidence obtained in a
prior development program, the vali
dation tests of ACTJITG T1.i-99 must
be conducted at full scale and be at
least two in number for each charac
teristic configuration of intersecting
beams and columns.
While the Commentary to T1.1-99
implies that experimental evidence
should be obtained in a prior develop
ment program, the Provisional Stan
dard does not require it. Rather, Ti .1
99 requires only that, prior to the
validation testing, a design procedure
should have been developed for the
generic frame, and that procedure used
to proportion the test modules.
In the NEHRP Provisions, the num
ber of tests required in the prior devel
opment program is not specified.
However, the results for the several
different frame systems studied in the
PRESSS (Precast Seismic Structural
Systems) program suggest that five or
more tests at one-quarter scale or
greater should be made in order to
provide the range of experimental in
formation needed to develop a mathe
-
LATERAL FORCE OR MOMENT
DRIFT RATIO
OO35
DRIFT FOR LIMITING
STIFFNESS OF BUILDING CODE
40
PCI JOURNAL
matical model sufficiently accurate for

purposes of analysis.
3. The special moment frame must
be designed using the nonlinear re
sponse history analysis procedure
specified in Section 5.8 of the 2000
NEHRP Provisions, using the forcedeformation characteristics for the
nonlinear action locations obtained
from the module tests.
The 2000 NEHRP Provisions re
quires designs to be carried out using
structural analyses conforming to one
of six types. Those types are Index
Force Analysis, Equivalent Lateral
Force Analysis, Modal Response Anal
ysis, Linear Response History Analy
sis, Nonlinear Static Analysis, and
Nonlinear Response History Analysis.
The type of analysis required de
pends on the SDC of the structure, its
height and irregularity. The typical
cast-in-place special moment frame in
structures with a limited degree of ir
regularity and not more than 17 or 18
stories in height can be analyzed with
any one of the foregoing six proce
dures except Index Force Analysis.
By contrast, a precast special mo
ment frame can be designed using Non
linear Response History Analysis only.
That method requires that a mathemati

cal model be used for the structure that
directly accounts for the nonlinear hys
teretic behavior of the components of
the structure. That model is then used
to determine the response of the struc
ture, through methods of numerical in
tegration, to suites of ground motion
compatible with the design response
spectrum for the site of the structure.
Use of Nonlinear Response History
Analysis is required for non-emulative
precast concrete special moment
frames. This is because none of the
other four procedures permitted for
cast-in-place frames can realistically
capture the strength and deformation
demands placed on the structure by
the range of structural characteristics
permitted by T1.1-99.
SPECIAL
STRUCTURAL WALLS
teria for walls have been demonstrated

by the results of tests in the direction
of the structural walls of the PRESSS
five-story building.
4
The 2000 NEHRP Provisions re
quire that the substantiating experi
mental evidence and analysis for spe
cial structural wall systems meet
requirements similar to those of Tl.199 for the design procedure used for
the test modules, the scale of the mod
ules, the testing agency, the test
method and the test report. The mini
mum test module must be a stack of
wall panels at least two stories high.
Based on the work described in Ref.
16, the test module must perform sat
isfactorily under cyclic loading at a
limiting drift ratio that is a function of
the characteristics of the wall and is
given by the criterion:
1.0 zi Ih (percent) =
0.67 [h/l] + 0.5 3.0
The studies that led to the develop

ment of the acceptance criteria of
T1.1-99 for special moment frames
also catalyzed studies that have re
sulted in the development of similar
acceptance criteria for special struc
tural walls.
16 The validity of those cri
(1)
where
height of entire module
length of entire module
Criterion 1 was derived after an ex
amination of results from tests on 178
cast-in-place walls with aspect ratios
Fig. 8.
Relative energy dissipation
Ah
A, I (E
) (0, +
2
E
5
ratio.
02)
Hatched Area
C
B
I---
DRIFT RATIO
D=O,
E,
41
(h/1) between 0.25 and 3.53. Those

results clearly showed, as apparent
from Fig. 10(a), that the limiting drift
ratio at the peak load varied linearly
with the aspect ratio and varied be
tween about 0.67 and 3.0 percent. For
ductile behavior, some post-peak
degradation must be expected, al
though the acceptable degree should
be less for walls than for frames be
cause the limiting drift ratio is less for
walls.
Analysis showed the acceptable
degree of degradation should be lim
ited to 20 percent for walls as com
pared to 25 percent for frames. For
that condition, Criterion 1 combines
the drift predictions for a simple
mathematical model of a wall hing
ing at its base and the use of a maxi
mum displacement ductility factor of
eight to assess the limiting drift. The
relationship derived between dis
placement ductility and limiting drift
for varying h/l values is compared
to the 0.035 limiting drift for frames
in Fig. 10(b).
For the third cycle at a drift ratio
equaling or exceeding the value
given by Criterion 1, the 2000
NEHRP Provisions requires that the

test module exhibit:
1. A degradation in post-peak ca
pacity not exceeding 20 percent.
2. A relative energy dissipation
ratio, defined in the same manner as
in ACT ITG/T1.l-99 (see Fig. 8) that
equals or exceeds 15 percent.
3. A stiffness around zero drift that
equals or exceeds that required by the
acceptance criteria of ACT ITG/T1.1
99 (see Fig. 9).
The basis for the slightly higher
energy dissipation ratio required for
walls than for frames is also docu
mented in Ref. 16. In the five-story
PRESSS building tests, a pair of
vertically coupled precast panel
structural wall stacks were used,
with each of the stacks having cen
trally located unbonded post-ten
sioned tendons.
The results of those PRESSS tests
for the shear wall direction and ana
lytical studies of precast post-ten
sioned walls reported in Refs. 14 and
16 validate the appropriateness of the
criteria specified in the 2000 NEHRP
Provisions for structural walls con
structed from precast panels.
-
CONNEC11ONS
Dry connections for seismic-forceresisting systems are classified into
two types, namely, Type Y and Type
Z. At nonlinear action locations, dis
placements both in the direction of ac
tion of the connection, and transverse
to it, must be controlled. For example,
if a sliding shear connection is to be
provided between two precast concrete
members, then there must also be a tie
between the two members to prevent
the sliding surfaces from separating.
Type Y connections must be able to
develop, for the flexure, shear, or axial
load, or combinations of those quanti
ties expected to act on the connection,
a probable strength, S, determined
using a value of unity, that is not less
than 125 percent of the yield strength
of the connection. In essence, the con
nection must be able to strain harden.
Under cyclic loading the connection
must be able to develop a displace
ment, at Spr, that is at least 4.0 times
its displacement at yield. The anchor
age of the connection into the precast
member on either side of a joint must
be designed to develop in tension 1.3
Fig. 9.
Unacceptable hysteretic
15
behavior.
0.035
-0.0035 I
0.0035
42
0.035
DRIFT RATIO
PCI JOURNAL
times Spr, and be connected directly by

a Type 2 splice to the principal rein
forcement of the precast or cast-inplace element.
For Type Z connections, Spr must
be not less than 140 percent of the
yield strength of the connection, and
under cyclic loading the connection
must be able to develop a displace
ment at Spr that is at least eight times
its displacement at yield. The anchor
age for the connection must also meet
in both tension and compression all

the requirements for Type Y connec
tions. Equilibrium based plasticity
models (strut-and-tie models), as de
scribed in 18.13.5 of ACT 3 18-99, are
to be used for the design of the con
nection region.
Confinement reinforcement in the
form of closed hoops or spirals with a
yield force not less than 0.05 times the
compressive force and with a spacing
not greater than 3 in. (76 mm) must be
Q
3
0
provided around the anchorage where

the local compressive stress at Sj,,,. ex
ceeds 0.7 f. The connection region is
defined in the same manner as an
chorage zone in Section 2.1 of ACI
318-99.
The testing of connections and the
evaluation of results must be made in
accordance with the principles of
ACT TTG/Tl.l-99. Appropriate pro
cedures for testing connections will
be described in more detail in the
next issue of the PCI JOURNAL in
the discussion of the design provi
sions for untopped diaphragms con
tained in the Appendix to Chapter 9
of the 2000 NEHRP Provisions. Con
nections at nonlinear action locations
in modules of frames and structural
walls used for validation testing are
deemed to satisfy the provisions for
connections if the results for the test
module satisfy the acceptance crite
ria for frames or structural walls, as
appropriate.
CONCLUDING
REMARKS
0
ASPECT RATIO (hJ)

(a) Limiting Drift Aspect Ratio Relationship
for Test Wall Database
-
.5
C
I:
0
3
2
DISPLACEMENT DUCTILITY
(b) Drift-Displacement Ductility Relationship for
Cantileaver Model of Wall
Fig. 10. Basis for Criterion 1 for special structural walls.

The 2000 NEHRP Provisions are

expected to include expanded seismic
design provisions for precast, pre
stressed concrete lateral-force-resist
ing systems. Emulative design, where
parity with the seismic performance of
cast-in-place reinforced concrete
structures is sought, and non-emula
tive design, where the unique proper
ties of precast concrete construction
are sought to be taken advantage of,
are both permitted to be used. Emula
tive design procedures are prescribed
for special moments frames as well as
for special structural walls (suitable
for use in Seismic Design Categories
D,EandF).
Under the special moment frame
provisions, designers may utilize
strong connections which remain
elastic as inelastic action takes place
away from those connections, or
they may utilize ductile connections
in which seismic energy dissipation
is allowed. Only the ductile connec
tion option is available under the
special structural wall provisions.
Non-emulative design procedures
are also prescribed for special mo
ment frames as well as for special
structural walls.
43
ACKNOWLEDGMENT
The original proposal leading to the reported changes from the 1997 to the 2000
Edition of the NEHRP Provisions was prepared by a PCI Fast Team consisting of
Ned Cleland, Thomas DArcy, Robert Fleischman, S.K. Ghosh, Neil Hawkins,
Phillip Iverson, Michael Oliva and Richard Sause. The contribution of the members
of this team is gratefully acknowledged.
REFERENCES
1. Federal Emergency Management
Agency, NEHRP Recommended Pro
visions for Seismic Regulations for
New Buildings and Other Structures,
1997 Edition, FEMA 302, February
1998; and Commentary, FEMA 303,

Washington D.C., February 1998.
2. ACT Committee 318, Building Code
Requirements for Structural Concrete
(318-99) and Commentary (3l8R99), American Concrete Institute,
Farmington Hills, MI, June 1999.
3. Restrepo, J. I., Park, R., and
Buchanan, A. H., Tests on Connec
tions of Earthquake Resisting Precast
Reinforced Concrete Perimeter Frames
of Buildings, PCI JOURNAL, V. 40,
No. 4, July-August 1995, pp. 44-61.
4. Restrepo, J. I., Park, R., and
Buchanan, A. H., Design of Connec
tions of Earthquake Resisting Precast
Reinforced Concrete Perimeter
Frames, PCI JOURNAL V. 40, No.
5, September-October 1995, .
77
68
pp.
5. Yoshioka, K., and Sekine, M., Exper
imental Study of Prefabricated BeamColumn Subassemblages, ACT SP
123, Design of Beam-Column Joints
for Seismic Resistance, American
Concrete Institute, Detroit, MI, 1991,
pp. 465-492.
,
44
6. Kurose, Y., Nagami, K., and Saito, Y.,

Beam-Column Joints in Precast Con
crete Construction in Japan, ACT SP
123, Design of Beam-Column Joints
for Seismic Resistance, American
Concrete Institute, Detroit, MI, 1991,
pp. 493-5 14.
7. Yee, A. A., Design Considerations
for Precast Prestressed Concrete
Building Structures in Seismic Areas,
PCI JOURNAL, V. 36, No. 3, MayJune 1991, pp. 40-55.
8. Warnes, C. E., Precast Concrete Con
nection Details for All Seismic
Zones, Concrete International, V. 14,
No. 11, November 1992, pp. 36-44.
9. Cheok, G. S., and Lew, H. S., Perfor
mance of Precast Concrete Beam-toColumn Connections Subject to Cyclic
Loading, PCI JOURNAL, V. 36,
No.4, July-August 1991, pp. 56-67.
10. Cheok, G. S., Stone, W. C., and Kun
nath, S. K., Seismic Response of Pre
cast Concrete Frames with Hybrid
Connections, ACI Structural Journal,
V. 95, No. 5, May 1998, pp. 527-539.
11. Stanton, J. F., Stone, W. C., and
Cheok, G. S., A Hybrid Reinforced
Precast Frame for Seismic Regions,
12. Priestley, M. J. N., Overview of

PRESSS Research Program, PCI
JOURNAL, V. 36, No. 4, July-August
1991, pp. 50-57.

13. Priestley, M. J. N., The PRESSS Pro
gram
Current Status and Proposed
Plans for Phase III, PCI JOURNAL,
V. 41, No. 2, March-April 1996, pp.
22-40.
14. Priestley, M. J. N., Sritharan, S., Con
ley, J. R., and Pampanin, S., Prelimi
nary Results and Conclusions from the
PRESSS Five-Story Precast Concrete
Test Building, PCI JOURNAL, V.

44, No. 6, November-December 1999,
pp. 42-67.
15. ACT Innovation Task Group 1 and
Collaborators, Acceptance Criteria
for Moment Frames Based on Struc
tural Testing, American Concrete In
stitute, Farmington Hills, MI, 1999.
16. Seo, S-Y., Lee, L-H., and Hawkins,
N. M., The Limiting Drift and En
ergy Dissipation Ratio for Shear
Walls Based on Structural Testing,
Journal of the Korean Concrete Insti
tute, V. 10, No. 6, December 1998,
pp. 273-311.
PCI JOURNAL, V. 42, No. 2, MarchApril1997, pp. 20-32.
PCI JOURNAL
ARCHITECTURAL
PRECAST CONCRETE
Color and Texture Selection Guide
For the first time, a visual guide is available to assist architects in the initial
selection of color and texture for architectural precast concrete. The Guide
is an extension of the information included in the architect-oriented
Architectural Precast Concrete Manual, published in 1989.
Color and Texture Selection Guide
Architectural Precast Concrete
illustrates more than 430 colors and textures for enhancing the aesthetics of
precast concrete panels. There are 236 6i x 11 in. color plates, with the
majority of the plates having two finishes on the same sample. The
photographs are numbered and arranged from light to dark colors. In
addition, there are six photographs of buildings that illustrate brick, tile,
terra cotta, granite, limestone and marble veneer-faced precast concrete
panel applications.
The variables considered in developing the color plates were cements.

pigments, coarse and fine aggregates, and texture or surface finish with
various depths of exposure. Since aggregates were collected from all parts
of the United States and Canada, designers are able to determine the
available aggregate colors. Identification of materials used to produce the
samples is included in the back of the Guide.
Who should have a copy?
U Architects
U Producers
CTG-92, First Edition,

260 pp., 8%xll in.
U Specifiers of precast concrete

U Owner/developers
SEND FOR YOUR COPIES TODAY!
custom 3-ring binder copies of
es, please send me
Color and Texture Selection Guide (CTG-92).

Architectural Precast Concrete
per
copy. (Illinois residents add 8.75% sales tax)
I understand the price is $40.00
Payable in U.S. Dollars, drawn on a U.S. bank, or use

your Visa or MasterCaid. Payment must accompany
your order. For addresses outside the United States,
add $25.00 per copy for airmail delivery.
Amount Enclosed:
Visa/MasterCard #
Cardholders Name
Expiration Date
Mail to:
name
company
address
p.o. box, also give Street address)
city
state
country
Precast!Prestressed Concrete Institute

209 W. Jackson Blvd.
Phone: 312-786-0300 Fax: 312-786-0353
zip code
PCI-SPONSORED RESEARCH STUDY
Design Criteria for Headed

Stud Groups in Shear:
Part 1 Steel Capacity and
Back Edge Effects
Neal S. Anderson, P.E., S.E.

Consultant
Wiss, Janney, Elstner Associates, Inc.
The Precast/Prestressed Concrete institute sponsored a cornprehen

sive research program to assess the shear capacity of headed stud
group anchorages. This program was initiated in response to new
provisions introduced into the 2002 AC! 318 Building Code. The
proposed AC! provisions are based on extensive data dominated
by post-installed anchor tests. Tests of headed stud anchorages, as
used in precast construction, are not prevalent in the literature.
The test program, conducted by Wiss, Janney, Elstner Associates,
Inc. (WJE), examined headed stud connections loaded toward a
free edge, a free edge near a corner, parallel to one free edge, par
allel to two free edges, away from a free edge, and in-the-field of a
member, such that edge distance was not a factor. The in formation
reported herein addresses the steel capacity failure mode. Test data
were obtained when the shear force was directed away from a free
edge, in-the-field testing, and from other edge distance tests where
steel failure governed the capacity.
eaded stud anchorages are
used extensively in the con
crete industry in both cast-inplace and precast construction. Weld
ing studs to steel plates provides an
easy and economical means of em
bedding and providing a ready to
complete structural connection. Such
a connection has substantial versatil
ity by allowing large variations in
construction dimensions.
Donald F. Meinheit, Ph.D.,

P.E., S.E.
Senior Consultant
46
Headed stud anchorages in precast

concrete members can be found in col
umn corbels, spandrel beams, dapped
end members, wall panels, tee beams,
and other components. Commonly,
studs in precast members are 3 to 8 in.
(76 to 203 mm) long and form multistud group connections. The load capac
ities of these connections are generally
affected by stud spacings, edge dis
tances, and member depth or thickness.
PCI JOURNAL
In the past, the design of stud an

chorages usually followed procedures
set forth in the PCI Design Handbook
or the nuclear structures code, devel
oped by ACT Committee 3492 Until
now, stud anchorage design has not
been codified within the widely ac
cepted Building Code Requirements
for Structural Concrete,
3 as prepared
by ACT Committee 318. However, an
approach for the design of anchorages
to concrete has been approved as Ap
pendix D
4 of the upcoming ACI 31802 Building Code. The ACI 318 Ap
pendix D method is based on the
Concrete Capacity Design (CCD)
model proposed by Fuchs, Eligen
hausen, and Breen.
5
The ACT design approach necessar
ily must consider all types of cast-inplace and post-installed anchors. The
design procedure in ACT 318-02 Ap
pendix D is calibrated using a
database heavily dominated by postinstalled anchors. Anchorages used in
precast concrete construction fall into
a relatively narrow range of those con
sidered in ACI 318-02 Appendix D.
For headed stud anchorages, the
ACT design approach shows signifi
cantly different capacity under certain
conditions than the approach used in
the current PCI design model. The
concrete break-out capacity calculated
with the proposed ACT approach is
typically lower than that predicted by
PCI design procedures, particularly
when edge and spacing distance ef
fects on stud groups are considered.
The differences in capacity prompted
the PCT to undertake a research pro
gram with the ultimate objective of im
proving design criteria for headed stud
anchorage groups, a connection type
commonly used by the precast concrete
industry. The research project included
an experimental program to provide the
background information for modifying
the ACT or PCI design approaches or to
justify and refine the PCI design ap
proach as currently published in the
PCI Design Handbook.
This paper represents Part 1 of four
parts included in this research pro
gram. The work reported herein pre
sents findings on two of the six pri
mary variables evaluated in this shear
capacity study. The variables in the
shear testing program included condi
-1
PLAN
hf
hI
,.
dh
,.
SEC TION THRU SLAB

Fig. 1. Member geometry and edge distance notation after PCI definitions.
tions that loaded the headed stud

groups:
1. Toward a free edge (de3).
2. Toward a free edge at a corner
(de3 and del simultaneously).
3. Parallel to one free edge (del).
4. Parallel to two free edges (del
and de2 simultaneously)
5. Away from a free edge (de4).
6. In-the-field of the member such
that edge distance was not a fac
tor in the failure.
Schematic representations of the
del, de2, de3, and de4 edge distances
are provided in Fig. 1.
Stud anchorage behavior when the
connection is loaded away from a free
edge and in-the-field is discussed
herein. These two conditions cause the
capacity to consistently be one of steel

failure.
This paper provides background in
formation for the steel capacity equa
tions in ACI 3 18-02 Appendix D. The
remaining components of this shear
research program including the front
edge effects, side edge effects, and
combined front and side edge effects
or corner influences will be reported
in future issues of the PCI JOURNAL.
OVERVIEW OF
RESEARCH PROGRAM
In mid-1996, PCI selected Wiss,
Janney, Elstner Associates, Inc. (WJE)
of Northbrook, Illinois, an engineering
consulting firm, to undertake this re
47
headed studs embedded in structural

concrete loaded in shear. Some experi
mental data on headed studs subjected
to tension, shear, and combined ten
sion and shear loadings have been
published. However, the database for
headed stud anchors is limited espe
cially when compared to the database
existing for post-installed anchors.
Additionally, cast-in-place anchorages
having head geometries similar to that
of headed studs were included in this
review, as applicable; for example,
some cast-in-place anchor bolts fall in
this category.
A literature search and analysis of
existing data were used to formulate a
laboratory test program. The shear
testing program, conducted in the
WJE laboratory, had the following
objectives:
Verify that a well-developed stud
connection, that is, studs long
search program under the direction of

PCIs Research and Development
Committee. An advisory panel was
appointed to closely monitor the pro
ject and consult on the testing scope.
The advisory panel includes individu
als from academia, consulting engi
neering, and precast concrete producer
members.
Program Development
6 focused on an
This research work
chorages and geometric conditions
typically used in the precast/pre
stressed concrete industry. The re
search concentrated on diameter, em
bedment depth, and number of welded
headed studs on a connection plate in
configurations commonly used in pre
cast concrete applications; the study
excluded post-installed anchors.
The first task of the research pro
gram was to review existing data on
enough to preclude a concrete pryout failure, located away from edge

influences will develop a capacity
dictated by characteristics of the
stud steel properties.
Evaluate group behavior near the
side (del), front (de3), and back
(de4) free edges considering as vari
ables: distance from an edge, spacing
in the x- and y-directions, number of
anchors in the group, embedment
depth, and anchor diameter.
Review or refine the concrete break
out model with respect to the del,
de3, and de4 edge distances.
Determine the influence of the slab
thickness effect for shear-loaded
connections.
Evaluate anchor group behavior at a
corner, where the del and de3 edge
distances meet.
Evaluate anchor group behavior
when simultaneous del and de2
edge conditions exist, such as in a
column.
Table 1. Summary of the WJE/PCI test program.

Testing Program Description
Tests completed
328 shear tests performed
Test specimens
34 concrete slab specimens cast

(12)5 x 5 ft x 6 in.
(4) 4xlOftxl6in.
(3) 4x4ftxI6in.
(I5)Sx5ftx 16 in.
16 push-off specimens utilizing a steel wide
flange section
Fabrication
Six separate concrete castings made

40.5 cu yd of concrete used for an average of
6.5 cu yd per casting
Supporting tests
Concrete compressive and tensile strength

Concrete modulus of elasticity
Tensile strength of all headed stud sizes used
(23 tests total)
Double shear strength of all headed stud sizes
used (18 tests total)
..
.
.
/2
J,
Anchorages
/2
--
14 anchorage plate layouts

1, 2, 3, 4 and 6 stud combinations evaluated
5000 to 6000 psi (typical of precast concrete)

Front edge (de3)
Side edge (del)
Back edge (de4)
Front corner (del-de3)
In-the-field
Push-offs
Test locations
Note: I in. = 25.4 mm; I ft= 0.3048 m; 1 lb

.
2
cu yd = 0.7646 m
Concrete strength
48
h, 5.38d
=
1
hf= 9.84d
5.93d
1
h,
x 3/s in.
55/ in.
x
s4x4
ioin.
5
/
3
Stud sizes
4.448 N; 1 kip = 4.448 N; I psi
102 tests
94 tests
23 tests
67 tests
26 tests
16 tests
6.895 kPa; 1 ksi
6.895 MPa
The WJE experimental program in

cluded 312 plate configurations in
shear and 16 push-off type specimens,
as summarized in Table I. The tests
were typically conducted in slabs mea
suring4x lOftor5 x5ft(l.2x3.Om
or 1.5 x 1.5 m) with either a 6 or 16 in.
(152 and 406 mm) thickness. The 16
push-off specimen tests were con
ducted to simulate the shear loading
conditions when an embedded anchor
group is adjacent to two longitudinal
edges simultaneously.
In the shear testing program, a total
of 14 different plate designations
were evaluated, which included dif
ferent combinations of plate size, stud
spacing, stud embedment depth, and
stud diameter. A conscience decision
was made that plate thickness and
concrete compressive strength would
not be variables in the test program.
Headed stud diameters of /2 and
/8 in. (12.7 and 15.9 mm) were tested
in this program.
The test program evaluated the ca
pacity of single and group connection
configurations in several geometric
conditions. Referring to Fig. 1, an
chorages were tested toward the free
edge (de3), at a corner, adjacent or
parallel to a free edge (del), and away
PCI JOURNAL
from a free or back edge (de4). Addi

tional testing was directed to evaluate
connection capacity when it was posi
tioned far away from the influence of
an edge, or the so-called in-the-field
tests. The in-the-field tests were antic
ipated to produce a better understand
ing of connection capacity when the
stud steel governs the failure mode.
Several test series were repeated in
both 6 and 16 in. (150 and 400 mm)
thick specimens to evaluate the effects
of member thickness.
The test findings reported in this
paper represent one part of an overall
comprehensive report on the shear be
havior of headed stud anchorages
loaded in shear. Results reported here
are limited to defining the capacity of
the stud steel in shear.
LITERATURE REVIEW
The welded headed stud gained con
siderable research attention in the late
1950s and through the 1960s. The
early research work on welded headed
studs was focused on applications in
the concrete slab-steel beam compos
ite member. The headed stud was
viewed to be an efficient and effective
shear transfer device, replacing chan

nels, angles, or fabricated spirals at
tached to the top flange of a steel
beam. The headed stud arc welding
process represented a labor and mate
rial cost savings over manual arc
welding the aforementioned shapes to
a steel beam.
Push-Off Tests
Testing to evaluate composite beam
behavior typically utilized a push-off
specimen to study shear transfer
through the headed studs. The push-off
test specimen commonly used a wide
flange beam section sandwiched be
tween two concrete slabs. Headed studs
were welded to both flanges in some
prescribed pattern or spacing and em
bedded into a thin concrete slab repre
senting the composite deck slab. The
concrete slab was usually reinforced to
simulate a bridge deck. As shown in
Fig. 2, the steel beam was held above
both the top and bottom elevation of
the slabs. Both the beam and two slabs
were oriented vertically fitting conve
niently into a universal testing machine.
Early composite beam research,
using the push-off specimen, was con-
ducted by Viest at the University of

7 Slutter, Fisher and others at
Illinois,
Lehigh University,
9 Baldwin, Dal
8
lum, and others at the University of
2 Goble at
Missouri-Columbia,
3
Case Western Reserve University,
Chinn at the University of Colorado,
14
and Hawkins at the University of Syd
5 These early test programs pro
ney.
duced a significant amount of shear
data on headed stud behavior with a
particular emphasis on groups. Several
of the push-off test failure loads were
due to stud steel shear, which is rele
vant to this paper.
Review of the push-off test results
provides good comparative data for
headed studs loaded in pure shear. As
stated earlier, previous testing on the
headed stud connections used in pre
cast concrete attachments is limited,
especially when groups are consid
ered. To evaluate group stud connec
tions, with an emphasis on steel fail
ure, there are no known published test
results.
Most of the non-push-off testing
programs were conducted by loading
the connection toward a free edge with
the intent of studying anchorages
loaded in shear and failing in a con-
Fig. 2. Typica push-off test specimen (from Oflgaard, Slutter and Fisher).
9
49
crete breakout mode. Therefore, pub

lished behavioral results on headed
stud groups loaded in pure shear with
out the influence of any edge effects
and failing the steel is entirely con
tained in the aforementioned refer
enced push-off tests.
It should be noted that the design of
the push-off test specimen has charac
teristics that limit its full applicability
to emulating a precast concrete an
chorage. Most of the thin concrete
slabs used in push-off tests contained
nominal reinforcement, more repre
sentative of bridge deck construction.
The reinforcement had no influence on
the first cracking load, but it is
likely that the reinforcement in the
concrete slab held the slab together to
allow for additional displacement and
ductility.
The early researchers were particu
larly concerned with load-slip charac
teristics of the connections. Unrein
forced concrete specimens, reported in
the literature, oftentimes produced a
transverse splitting failure in the con
crete slab, a failure mode unlikely to
occur in actual bridge deck construc
tion because of the presence of trans
verse reinforcement.
Another limitation of the push-off
specimen relates to the mechanism to
apply load to the embedded studs.
Load transfer from the steel column
through the headed studs into the two
concrete slabs results in the best con
ditions to place the studs in pure shear.
However, the external applied load
causes a reaction against the ends of
the two concrete slabs, placing them in
compression. This condition is viewed
to be analogous to a headed stud an
chorage located in-the-field of a mem
ber; that is, a significant amount of
concrete slab is located in front of the
anchorage to preclude any front edge
influences.
The favorable concrete compression
stress developed in front of the studs
does not affect tests having one trans
verse row (or one y-row) of studs. On
the contrary, when stud groups with
multiple longitudinal rows were tested
using the push-off specimen, the test
results become more difficult to inter
pret. Each longitudinal row in the
group is subjected to a different level
of compressive confinement stress.
50
1.4
1.2
D 1.0
c.
0.8
0.6
0.
0.4
0.2
0.0
4
10
Embedment Depth Ratio (li.

1 Id)
(a) Concrete and steel failures.

6.0
5.0
U
4.0
3.0
U
2.0
1.0
0.0
0
Embedment Depth Ratio (h., / d)
(b) Shallow embedment capacity predictions.
3.0
PCI Ply-Out Equation
Llghw.Ight Concrete
Steel Equation (AoFut)

-.
2.5
2.0
::
1.5
1.0
0.5
0.0
0
Embedment Depth Ratio (h, Id)
(c) Concrete and steel failures in push-off tests with lightweight

concrete using PCI pryout capacity equation.
Fig. 3. Test-to-predicted capacity ratio as a function of embedment depth.
PCI JOURNAL
Likewise, multiple longitudinal (or y)

rows which are spaced at large dis
tances reduce the efficiency of the an
chor group due to shear lag effects.
A push-off specimen with multiple
longitudinal rows of studs is similar to
a long bolted connection whose effi
ciency is reduced in proportion to
/ L (where
is the distance be
tween the shear plane of the connected
parts and the centroid of the connected
component) in accordance with steel
tension member design.
Our review of the available push-off
data shows the overall connection
length, L, may be a significant vari
able in determining the capacity that
can be achieved by the stud because of
shear lag. For further discussion of
shear lag, see the book by Kulak,
Fisher, and Struiki
6
Keeping the above limitations of the
push-off test in perspective, some
valuable data were applicable to the
present study. Relevant information
from these early tests is discussed
below.
Embedment Depth
In 1955,
7 performed 12 push-off tests at
Viest
the University of Illinois as part of re
search into composite beam behavior.
Stud diameters ranged from 1/2 to 11/4
in. (12.7 to 31.8 mm), with a reason
ably constant effective embedment
depth (hej) between 3 and 3/2 in. (76.2
to 88.9 mm). All studs were placed in
one row with an approximate 4 in.
(101.6 mm) x-spacing for 10 of the 12
tests. Two tests had four, 3/4 in. (19.1
mm) diameter studs in one y-row with
an approximate 2 in. (50.8 mm) cen
ter-to-center spacing. To ensure the
studs were the only shear transfer
mechanism, the steel I-beams were
coated with grease to minimize any
frictional transfer of the shear load
along the flange width.
As reported in the work by Viest,
the two-stud tests having ratios of ef
fective depth to stud diameter (held)
of 4.53, 5.5, and 7.0 failed in stud
shear (steel failure). The four tests
with two studs having diameters of 1
and 1/4in. (25.5 and 31.5 mm)experi
enced concrete failure. These four
tests had average held ratios of 3.22
and 2.51, respectively.
In the original Viest work, because
the stud height was relatively con.
stant, two prediction equations were

presented for stud diameters less than
1 in. (25.4 mm) and greater than or
equal to I in. (25.4 mm). In a subse
quent research summary paper,
7 described testing ten addi
Viest
tional push-off specimens and modi
fying the equations for the 1957
AASHO Specifications. Instead of
making the design equation a function
of diameter only, the critical parame
ter became held.
Another observation from the Viest
test data is the apparent good correla
tion of steel shear capacity using a
prediction of 1.0 A
F, when held
5
4.53. In this equation, A
5 is the crosssectional area of the stud shank and
F, is the ultimate tensile strength of
the stud steel.
This predicted steel shear failure
load corresponds quite well to test
data when ultimate tensile strength of
the steel is used, instead of a value re
duced for tensile yield (F = 0.9 F,),
where F is the offset tensile yield
stress for the stud steel. Likewise, it is

a better predictor than the shear yield
(F = O.
F, that is, 1/ J), where
58
the
is
shear yield stress according
5
F
to the Huber-von Mises-Hencky yield
8
criterion.
WJE compiled push-off test data
from a number of the referenced re
search studies to evaluate concrete and
steel failures. To eliminate having the
data influenced by shear lag, only
push-off specimens with one longitu
dinal (y) row of studs were evaluated.
Fig. 3(a) shows a graph of the test-topredicted ratio for steel failure plotted
against embedment depth ratio (held),
where the predicted capacity is based
on 1.0A
F,.
5
The trend of the data indicates that
1.0 A
F, is a good predictor for a steel
5
failure when the embedment depth
(held) exceeds about 4.5. This is just
slightly greater than the value of 4.2
identified by Driscoll and Slutter
5 and
incorporated into the 1961 AASHO
Specifications.
Table 2. Review of PCI Design Handbook requirements for stud strength

governed by steel.
PCI Handbook
edition
Parameters
phi (ci)
Steel
Steel strength
equation
none
f=60ksi
V=0.75Aj
0.85
f. = 60 ksi
1 (1971)
;
--2(1978)
(shear-friction concept)
45.0 Ab
(where u= 1.0)
=
---+----3 (1985)
f. = 60 ksi
1.0
=
=
4(1992)
1.0
5 (1999)
0.90
I
J=60ksi
50 ksi
4Vci(O.75)fA
n
5
= 45 0 A, n
=
ci(0.9)fA,,n
=40.5A,,,z
Note:
= A,, = cross-sectional area of the stud shank (sq in.);
= ultimate tensile strength (ksi);f = yield strength (ksi).
(0.75) A,.L
45.0 A
5
number of studs in the connection;
f,
Table 3. Minimum mechanical property requirements for headed studs

adapted from AWS Dl .1 2000.31
Property
Type A
Type B
61.000 psi (420 MPa)
65.000 psi (450 MPa)
Yield strength (0.2 percent offset)
49.000 psi (340 MPa)
51,000 psi (350 MPa)
Elongation (mm. percent 2 in.)
17 percent
20 percent
Reduction of area (mm.)
50 percent
50 percent
Tensile strength
(mm.)
51
For some tests conducted in normal

weight concrete, steel stud shear fail
ure occurred at embedment depth ra
tios (held) less than 4.5. The tests with
shorter stud embedment depths gener
ally have predicted steel shear capaci
ties (using 1.0 ASFUF) greater than the
actual test results, even though the re
ported failure mode was that of stud
failure. The steel failure mode may
have, in fact, been a secondary failure
after considerable concrete crushing or
stud deformation had occurred.
Work performed by Ollgaard, Slut
ter, and Fisher
9 at Lehigh University
was an extensive study using studs
with an effective embedment depth
(held) of 3.26 and different types of
lightweight and normal weight con
crete. Failures were noted in both stud
steel shear or by a concrete mecha
nism. Results of this work produced a
prediction equation, independent of
failure mode, basing individual stud
strength on stud area, concrete com
pressive strength, and elastic modulus
of the concrete. Their final prediction
equation was:
=
(a) Overall view of test setup.
(1)
where
nominal strength of a shear
stud connector embedded in a
solid concrete slab (kips)
5 = effective cross-sectional area
A
of a stud anchor (sq in.)
=
specified
compressive strength
f
of concrete (ksi)
= modulus of elasticity of con
crete (ksi)
This equation is applicable to both
normal and lightweight aggregate con
crete. Unlike earlier prediction equa
tions from the push-off test, this equa
tion did not have a limitation on
effective stud embedment depth, held.
This equation has a long history of
being a good predictor of shear capac
ity, as it has been referenced in the
AISC Specification since 1978.20 In
AISC, the upper bound on the stud
strength is A
F, where A
5
5 is the
cross-sectional area of a stud shear
connector and F is the minimum
specified tensile strength of the stud
shear connector.
A simplified lower bound form of
the Ollgaard, Slutter, and Fisher equa
52
(b) Close-up view of test fixture.

Fig. 4. Test setup for tension testing headed studs welded to a steel plate.
21
tion was proposed by Shaikh and Yi
in 1985:
800AA,/j
(2)
where
V, = nominal shear strength (ib)
A
specified compressive strength
of concrete (psi)
= concrete unit weight factor
The conversion of Eq. (1) to Eq. (2)
with its assumptions and use of A, re
sulted in an average prediction equa
tion. Consequently, Shaikh and Yi Se

lected a lower bound line through the
data, resulting in a constant of 800.
This equation appeared in both the
Third and Fourth Editions of the PCI
Design Handbook.
When the concrete failure test loads
for short studs (held < 4.5) in normal
weight concrete are predicted in ac
cordance with the PCI Design Hand
book Fourth Edition
22 Eq. (6.5.8) [Eq.
(2) above], there is reasonably good
correlation with the data. The data
plot in Fig. 3(b) shows the data trend
with test-to-predicted capacities of
about 1.0.
PCI JOURNAL
In summary, the work by Ollgaard,

Slutter, and Fisher, as well as the WJE
review of other data showed that 1.0
is a good predictor of steel stud
shear capacity when he/d 4.5. When
headed studs are shorter than this ef
fective embedment depth, a concrete
pryout failure mechanism occurs.
Concrete pryout failure is a concrete
breakout failure mode that is not a
function of edge distance but a func
tion of the stiffness of the headed
stud. Thus for short, stocky anchors
with he/d less than 4.5 it appears that
Eq. (6.5.8) in the Fourth Edition of the
PCI Design Handbook, Eq. (2), appro
priately predicts the concrete failure
mode.
Lightweight Aggregate Concrete
Our analysis of reported steel shear
failures in studs embedded in
lightweight concrete indicates test
strengths less than that predicted by
1.0 AF
. The data were found in the
1
work by Ollgaard, Slutter, and Fisher;
9
4 and studies at the University
Chinn;
of Missouri-Columbia.
1 1,12 Lightweight
aggregate concrete appears to provide
an embedment environment whereby
the stud causes greater crushing of the
concrete, producing more bending de
formation, resulting in larger overall
slip of the embedment.
On a macro level, lightweight con
crete is a matrix of cement paste and
lightweight aggregates. The cement
paste, taken alone, has a higher com
pressive strength than the individual
lightweight aggregate particles. In
combination, this mixture usually does
not exhibit a reduction in compressive
strength over comparable normal
weight concrete.
When a steel shear failure occurs,
the welded end of the stud deforms by
bending as it bears on the concrete.
Extremely high, concentrated bearing
stresses develop in the concrete, which
eventually cause localized concrete
crushing. Lightweight concrete has a
lower concentrated point bearing
strength than the point bearing
strength in normal weight concrete be
cause of the lightweight aggregate par
ticles in the matrix. The weld connect
ing the stud to the plate concentrates
the transfer of the shear load over a
small bearing area in the concrete. As
such, the bearing and localized crush
Fig. 5.
Test setup for shear
testing of headed
studs.
ing zone in the lightweight concrete

needed for the concrete to mobilize
bearing resistance is deeper.
Lightweight concrete appears to
allow the headed stud a greater oppor
tunity to deform, and the increased
concrete deformation induces more
bending into the stud and attachment
weld. The failure mode appears to be
the result of combined shear and ten
sion from the stud bending on the crit
ically stressed portion of the weld.
In our analysis, this apparently
higher bending deformation combined
with shear deformation reduces the ca
pacity of the headed stud to a lower
value than 1 .0
The stud shear
capacity well away from a free edge in
lightweight concrete is better pre
dicted by Eqs. (1) or (2) for either con
crete or steel failures, regardless of
embedment depth, as shown in
Fig. 3(c).
Connection Plate Thickness
To
ensure adequate distribution of an ap
plied shear force to the individual
headed studs on a connection plate,
the plate must have a certain minimum
thickness. The PCI Handbook Fourth
22 required the minimum plate
Edition
thickness (t,,
) to be two-thirds the stud
1
diameter (t,, 0.67d). The minimum
plate thickness requirement was re
duced in the Fifth Edition of the PCI
1 to one-half the stud diame
Handbook
ter (t,,
1 0.5d).
Research on the minimum plate
thickness appears to be limited to
work by Goble at Case Western Re
serve University.
3 This research fo
cused on the minimum flange thick

ness required in light-gage steel in
order to develop the full capacity of a
welded stud connection. Goble deter
mined the stud diameter-to-flange
thickness (dItf ratio was required to be
less than or equal to 2.7 to develop the
stud weld. A total of 41 push-off spec
imens were tested to develop this find
ing.
Alternatively written, the minimum
flange thickness required must be
greater than 0.37d. Flange thickness
meeting this minimum criterion exhib
ited steel stud shear failure. Thinner
flanges, exceeding a d/t of 2.7,
showed partial or full pullout of the
stud from the flange, producing a
divot or crater.
Based on this work, it appears that
the current PCI minimum plate thick
ness requirements are slightly conser
vative when a connection is loaded in
pure shear. An increased plate thick
ness may be justified when tension or
combined shear and tension are con
sidered to ensure more uniform load
distribution to the stud group.
Minimum Slab Thickness
Steel
stud failures, loaded in shear in the
push-off specimens, were achieved in
some relatively thin slabs. Review
of the data shows steel failures oc
curred in slabs ranging in thickness
from 4 to 7 in. (102 to 178 mm). For
these tests, the clear cover over the
head of the stud on the free surface
side of the slab ranged from 1 to 3.1
in. (25.4 to 78.7 mm). A more conve
nient description is the slab thickness
53
>
0
C
-U
(21
*Column 7: 0.2 percent offset yield strength.

Column 13: NR denotes Not Reported.
Note: 1 in. = 25.4 mm; 1 lb = 4.448 N; 1 psi
5.335
5.343
5.346
5.347
5.346
5.346
5.344
5.345
5.343
4.237
4.240
4.237
4.236
4.230
4.238
0.1917
0.1917
0.1917
0.1909
0.1913
0.1909
0.1909
0.1909
0.1909
0.1909
0.1971
0.3058
0.3048
0.3058
0.3048
0.3048
0.3058
0.3058
6.895 kPa.
3.145
3.152
3.154
0.1932
0.1932
0.1932
0.1917
0.1932
..
3.155
3.158
3.153
4.01
3.989
4.030
4.013
4.010
4.018
4.015
5.18
5.188
5.177
5.188
5.20
5.205
5.212
5.201
5.199
5.188
5.187
2.96
2.963
2.926
2.991
2.95
2.953
2.956
2.930
(4)
(5)
Length
Length
before
after
welding (BW) welding (AW)
(in.)
(in.)
0.1917
0.1917
0.1917
0.1909
0.1917
0.1917
(3)
Crosssectional
area
(sq in.)
eiisiie ess on I4eison neaaea stuas.
(2)
Measured
Stud
actual
description
diameter
(in.)
/2 in. x 3Is in. (Heat F48266)
WJEI/12/99(l)
0.494
WJE 1/12/99 (2)
0.494
WJE 1/12/99 (3)
0.494
Nelson 9/23/98
0.493
Nelson 11/98
0.494
WJE 1/15/99
0.494
Average
0.494
Std.dev.
112 in. x 3Is in. (Heat 792680)
WJE /12/99 (4)
0.496
WJE 1/12/99 (5)
0.496
WJE 1/12/99 (6)
0.496
Nelson 4/14/98
0.494
WJE 1/5/99
0.496
Average
0.496
Std. dev.
/2 in. x 5/16 in. (Heat F48262)
WJE 1/13/99 (7)
0.494
WJE 1/13/99 (8)
0.494
WJE 1/13/99 (9)
0.494
WJEI/18/99(D)
0.493
WJEI/l8/99(E)
0.494
WJE 1/18/99 (F)
0.493
Nelson 9/23/98
0.493
Average
0.494
Std. dev.
1/2 in. x 5/16 in. (Heat F47841)
WJEI/18/99(A)
0.493
WJE 1/18/99 (B)
0.493
WJE 1/18/99 (C)
0.493
Nelson 9/23/98
0.501
Average
0.495
Std. dev.
/s in. x 4/16 in. (Heat A43765)
WJE 1/13/99 (A)
0.624
WJE 1/13/99 (B)
0.623
WJE I/I 3/99 (C)
0.624
WJE 1/18/99 (D)
0.623
WJE 1/18/99 (E)
0.623
WJE 1/18/99 (F)
0.624
Nelson 4/14/98
0.624
Average
0.624
Stddev
(1)
I dUI 5. I\tSI1S UI
0.248
0.210
0.224
0.226
0.212
0.223
0.156
0.168
0.155
0.130
0.131
0.145
0.148
0.158
0.160
0.182
0.226
0.163
0.202
0.202
0.223
(6)
Amount
of
burnoff
(in.)
58205
46582
61475
67577
61673
56897
61475
59126
61554
59982
62863
60871*
61466
56098
56348
53218
55305
58149
54894
58673
62863*
59518
64952
66763
64175
69351
62609*
62609
66653
64174
70435
69392
64958*
(7)
Estimated
stress at
proport. limit
(psi)
76569
72373
76383
83104
81972
81468
74391
78.04 ksi
4l5ksi
81199
79627
80151
75329
79.08 ksi
2.58 ksi
81392
82957
80870
80675
81034
80151
81198
81.18 ksi
0.88 ksi
78667
79184
79184
74348
76907
77.66 ksi
2.07 ksi
82770
77320
81705
81.60 ksi
2.17 ksi
82957
81914
82957
Failure
stress
(psi)
(8)
0.368
0.375
0.369
0.385
0.374
0.376
0.401
0.321
0.306
0.311
0.321
0.312
0.303
0.315
0.312
0.305
0.320
0.293
0.291
0.286
0.316
0.302
0.306
0.305
0.308
0.302
(9)
Diameter
at fracture
section
(in.)
0.1064
0.1104
0.1069
0.1164
0.1099
0.1110
0.1263
0.0809
0.0733
0.0760
0.0809
0.0765
0.0721
0.0779
0.0762
0.0731
0.0804
65.2
63.8
65.0
61.8
64.0
63.7
58.7
57.6
61.6
60.2
58.9
60.1
62.4
59.3
601
61.8
57.9
61.4
65.1
65.6
66.8
59.1
62.9
61.6
0.0735
0.0674
0.0665
0.0642
0.0784
0.0716
61.9
61.1
62.6
60.9
(11)
Percent
reduction
in area
(percent)
0.0731
0.0745
0.0716
(10)
Reduced
crosssectional area
(sq in.)
.?
28.0
21.0
21.0
22.0
21.3
Percent
elongation
(percent)
(12)
X
X
X
X
X
X
X
X
X
X
NR
X
X
X
X
X
X
NR
X
X
X
NR
X
X
X
X
NR
NR
X
Failure
in shank
(13)
to-effective embedment depth ratio

(h/hef), which ranged from 1.4 to 2.2
for the published push-off test data.
No definite conclusions can be ex
tracted from the push-off data regard
ing minimum slab thickness within the
range tested. However, we conclude
that slab thickness is not a variable in
fluencing a stud steel shear failure.
X-Spacing Effect
In early work
by Viest,
7 a comparative set of tests
was conducted with 3/4 in. (19 mm) di
ameter studs at differing x-spacings;
that is, the distance between adjacent
studs in one row. Tests 6A2 and 6B2
in that series had two studs in a trans
verse row with a 3.0 in. (99 mm) x
spacing. Companion specimens 6A4
and 6B4 had four studs across with a
center-to-center spacing (x-spacing) of
1.9 in. (48.3 mm). All studs in these
tests had embedment depth to stud di
ameter ratios (held) greater than 4.5,
such that the concrete pryout failure
mode should not occur.
The two specimens with four studs
in a row (Tests 6A4 and 6B4) failed in
a concrete failure mode, even though
their held values were greater than
4.5. This implies that the anchor spac
ing in the x-direction can be a factor in
the ultimate load capacity of the con
nection. In the ACT 318 Code Ap
pendix D design provisions,
4 the mini
mum anchor spacing for multi-anchor
connections is 4d. For the Viest tests
with four studs in one row, the x-spac
ing ratio (xld) was 2.5 while tests with
two studs per row had a ratio of 4.0.
Thus, it appears that if a minimum
spacing of 4d is imposed on anchor
spacing in a group with large edge dis
tances, a steel failure should occur
when held is greater than 4.5.
Other than these tests by Viest,
push-off tests by Goble
3 and
5 studied conditions with
Hawkins
held greater than 4.5 and reported stud
steel failure with a x-spacing as low as
4.8d. Other shear tests, exceeding a x
spacing of 4.8d, caused the stud to fail
in shear.
PCI Design Handbook Review

Throughout the years, the PCI De
sign Handbooks
25 have contained
22
upper limits on the strength of a
headed stud connection loaded in
shear. The maximum permissible
shear load has always been limited by

the studs steel strength in shear.
Table 2 summarizes the history of the
PCI Design Handbook requirements
for stud strength.
As shown in Table 2, the equations,
strength reduction phi () factors, and
steel strengths (yield or ultimate ten
sile) used have varied. However, the
basic resultant allowable stress on
the stud has remained relatively con
stant at 45 ksi (310 MPa), assuming a
minimum ultimate stud tensile strength
(F, or F) of 60 ksi (414 MPa).
Aside from the information listed in
Table 2, the Handbook has contained
some unique provisions. The Second
Edition incorporated the concept of
shear-friction into the limiting equa
tion. But the shear-friction coefficient,
i, was conservatively set equal to 1.0
based on the lack of test data.
The Third and Fourth Editions of
the Handbook also contained a limita
tion on the concrete pryout strength
equal to Eq. (2). As addressed earlier,
this equation was a simplification to
the Ollgaard, Slutter, and Fisher equa
tion proposed by Shaikh and Yi, Eq.
(2) was actually based on both con
crete and steel failures. Note that Eq.
(2) was dropped from the Fifth Edition
of the Handbook.
Steel Code Review
The nominal shear strength of one
stud connector embedded in a solid
concrete slab is given in the AISC
26 as Eq. (1). This pryspecifications
out equation in AISC carries an upper
limit of
AISC does not specifi
cally stipulate a separate phi reduction
factor () for studs. The reduction fac
tor is apparently grouped with the
bending phi factor, b which is 0.85
for plastic redistribution of stress or
0.90 for an elastic stress distribution
on the section. These reduction fac
tors are also calibrated to the
AISC/ASCE load factors, which dif
fer slightly from the traditional ACT
load factors.
AISC does not place a limit on the
effective stud embedment depth, held.
They do, however, place limitations
on the stud spacing. The minimum
center-to-center spacing of studs is to
be 4d in the transverse direction, and
6d along the longitudinal axis of the
Fig. 6. Ductile failure of welded headed

stud away from the weld.
composite beam. Finally, AISC limits

the stud diameter to less than 2.5 times
the flange thickness.
The code requirements in the Hand
27 from the
book of Steel Construction
Canadian Institute of Steel Construc
tion (CISC)
27 are similar to AISC. The
Canadian Code does place a capacity
reduction factor (b) on the pryout
equation, Eq. (1), while keeping the
The phi resis
upper limit to
factor
is
0.80.
Moreover,
tance
(&c)
are
to
have
an
effective
welded studs
height to stud diameter ratio (held)
greater than 4.
Summary of
Code Material Requirements
For the past 35 years, the material
strength requirements for welded
headed studs have remained unchanged.
The 1965, Ninth Edition of the AASHO
Standard Specifications for Highway
28 representing the earliest codi
Bridges,
fied properties, specified that shear con
nector studs should conform to ASTM
A108 in Grades 1015, 1017, or 1020.
The physical material properties listed
in the 1965 AASHO were:
Tensile strength (mm.): 60,000 psi
(415 MPa)
Yield strength (mm.): 50,000 psi
(345 MPa) (0.2 percent offset)
Elongation in 2 in. (52 mm) (mm.):
20 percent
Reduction of area (mm.): 50 percent
These material properties are identi
cal to the first requirements published
by the American Welding Society
(AWS) for shear connectors in the
1968 Supplement to AWS D1.0-66
and D2.0-66.
29 AWS also designated a
second class of studs as studs other
than shear connectors having the fol
55
lowing material properties:

Tensile strength (mm.): 55,000 psi
(380 MPa)
Yield strength (mm.): No requirement
Elongation in 2 in. (52 mm) (mm.):
20 percent
In the 1982 edition of AWS Dl.l,
3
these lower strength studs were
classed as Type A studs, whereas the
60 ksi (415 MPa) studs were classed
as Type B studs.
The current Structural Welding
Code, AWS Dl.l 2000, has recog
nized that mild steels conforming to
ASTM A108
32 (Grades 1010 through
1020) and used for headed studs have
increased material properties. Table 3,
adapted from AWS D1.l-2000 Table
7.1, shows the current minimum ten

sile strength (F) to be 65 ksi (450
MPa) and yield strength (F) to be 51
ksi (350 MPa) for Type B studs. Cur
rently, AWS classifies Type B studs as
studs that are headed, bent, or of other
configuration in 1/2 /8, 3/4,
and I
in. (12, 16, 20, 22, and 25 mm) diame
ters used as an essential component in
composite beam design and other con
struction. The stud diameters listed ap
pear to be the majority of those also
used in precast concrete construction.
Type A studs cover the 1/4 and /8 in.
(6 and 10 mm) diameter stud sizes,
used occasionally in precast construc
tion. As shown in Table 3, Type A
studs currently have a 61 ksi (420
MPa) minimum tensile strength (F)

and a 49 ksi (340 MPa) minimum
yield strength (F). AWS defines Type
A studs as general purpose of any
type and size used for purposes other
than shear transfer in composite beam
design and construction.
STUD TESTS
Design rules for steel anchorages
are generally based on the tensile
properties of steel. For most design
cases, it is convenient to base the ca
pacity of headed studs on the tensile
yield or strength values and relate the
shear capacity to a factored reduction
of either value. Materials used for
Table 5. Results of double shear tests on Nelson headed studs.

(4)
(5)
Crosssectional
area
(sq in.)
Length
before
welding
(8W)
(in.)
(6)
Estimated
stress at
proportional
limit
(psi)
Unrestrained
0.494
0.493
0.493
0.493
0.19 17
0.1909
0. 1909
0. 1911
3.160
3.161
3.160
3.160
46624
37718
39505
Restrained
0.494
0.494
1 0.494
0.494
0.1917
0.1917
0. 1917
0.1917
3.159
3.160
3.164
3.161
34174
38087
45913
39392
x 3/s in. (Heat 792680) Unrestrained

WJE 4/21/99
1
0.496
WJE 4/21/99
2
0.497
WJE 4/21/99
0.497
3
Average
0.497
0.1932
0. 1940
0.1940
0. 1937
3.151
3.148
3.148
3.149
32088
36082
35825
34665
48132
48196
46392
47573
x
x
x
Unrestrained
0.494
0.494
0.493
0.494
0. 19 17
0. 19 17
0. 1909
0. 19 14
5.347
5.345
5.345
50609
47479
51338
49809
x
x
x
5.346
34696
34174
35623
34831
Restrained
0.494
0.494
0.494
0.494
0.1917
0.1917
0.1917
0.1917
5.344
5.345
5.344
5.344
45653
46957
35218
42609
54783
55826
55044
55218
x
x
Unrestrained
0.624
0.625
0.624
0.624
0.3058
0.3068
0.3058
0.3061
4.237
4.245
4.229
4.237
34171
36669
30411
33750
51829
51337
50194
51120
x
x
x
(1)
Stud
description
(3)
(2)
Measured
actual
diameter
(in.)
Test
number
/ain(IeatF48266)
t
3
WJE 4/21/99
1
WJE 4/21/99
2
WJE 4/21/99
3
/zx
x 3/ in. (Heat F48266)

WJE 4/23/99
IA
WJE 4/23/99
2A
WJE 4/23/99
3A
Average
Is x 5/16 in. (Heat F47841)

WJE 4/21/99
WJE 4/21/99
2
WJE 4/21/99
3
Average
in. (He F47841)
WJE 4/23/99
1A
WJE 4/21/99
2A
WJE 4/21/99
3A
Average
1/2 x 5/16
/s x
in. (Heat A43765)
WJE 4/21/99
1
WJE 4/21/99
2
WJE 4/21/99
3
Average
34174
(7)
(8)
Maximum
shear
stress
(psi)
Failure
in shank
49566
56053
51600
52406
x
x
x
50870
x
x
x
55044
55305
53740
(9)
(10)
Shear-Tensile data
Avg. shank
Ratio of
tensile
shear to
strength
tensile
(psi)
strength
81600
0.642
81600
0.659
77660
0.613
79080
0.630
79080
0.698
78040
0.655
Note 1: Tests identified with an A were run with side plates and the center plate, in the double shear test, having the same size drill holes. All other tests allowed the
stud bearing on the side plates to rotate unrestrained.
Note 2: 1 in. = 25.4 mm; 1 lb = 4.448 N; I psi = 6.895 kPa.
56
PCI JOURNAL
manufacturing headed studs do not

generally exhibit well defined yield
point values. Therefore, the capacity
of the stud is much easier to calculate
when based on the easier to measure
tensile strength. The headed stud prop
erties used in these shear tests will
also be related to the measured tensile
strength properties of the stud.
Testing for Material Properties
Upon receipt of the headed stud ma
terials from Nelson Stud Welding,
WJE independently tested the geome
try and physical characteristics for the
various steel heats in the project stock.
Four different stud length and diame
ter configurations were received.
These headed studs were manufac
tured from six different heats of steel
wire.
Headed studs were tested for their
tensile and shear strength properties,
in air. The test fixture was similar to
that suggested in AWS Dl.l-2000.
Double shear, guillotine tests were
also conducted on the middle third of
the stud shank to determine the steel
shear strength.
The steel plate for the test anchor
ages was purchased as ASTM A36
steel. Strength properties and chemical
compositions of the plate steel were
taken from mill test reports.
Test Setups
A universal test
Direct Tension
ing machine was adapted to test
headed studs that had been welded to a
square plate. The plate thickness was
1/2 in. (12.7 mm) in all cases. Both 1/2
and /8 in. (12.7 and 15.9 mm) diame
ter studs were evaluated. The pho
tographs in Figs. 4(a) and 4(b) show
the overall test setup and a close-up
view for the welded, headed stud in
the test fixture.
A load-deformation plot was made
for each test. Deformation was mea
sured between the cross-heads of the
testing machine and, therefore, repre
sents an overall behavior and not the
actual stud shank strain.
Double Shear
Two different
types of double shear, guillotine tests
were conducted. The test setup, shown
in Fig. 5, was essentially the same for
both test types. A three-plate fixture
30
Fig. 7.
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
Dofloction (in)
was used similar to the typical pushoff test specimen discussed earlier.
The side plates contained either a slot
or oversized hole. The interior plate
reacted against the test machine head
and was guided between the exterior
plates, bearing on the stud. The fixture
placed the studs in double shear.
The majority of tests allowed unre
strained rotation to occur in the stud
shank as it was bearing against the two
side plates. Some additional tests were
conducted by drilling a hole slightly
larger than the stud diameter through
the side plate and thereby restraining
the stud rotation.
Test Results
Tension test results and
Tension
geometric properties for the various
steel heats are listed in Table 4. Tests
conducted in-house by Nelson Stud
Welding are also incorporated into the
table. The tensile test results typically
accompanied the mill certificates for
each steel heat.
Each tested stud exhibited a round
house load-deformation curve, requir
ing the 0.2 percent offset method to
determine its yield strength. The stud
yield strength was approximately 80
percent of the tensile strength.
All studs failed in a ductile manner
forming a cup and cone fracture sur
face in the neck-down region, as
shown in Fig. 6. The fracture was also
0.40
0.45
Shear failure of headed stud

loaded in double shear.
consistently away from the weld. The

reduction of area was approximately
60 percent for all tests.
Double shear test results
Shear
are listed in Table 5. Two of the series
were tested restrained and the remain
der were unrestrained. The restrained
tests, similar to the condition that may
exist when the stud is embedded in
concrete, had a slightly higher shear
strength. The shear stress at the pro
portional limit is approximately 70
percent of the measured shear
strength. A typical failure of the stud
in double shear is shown in Fig. 7.
Mill
Plate Material Properties
test reports showed that the flat bar
stock used for the anchorage plates
had an average yield stress of 47.6 ksi
(328 MPa). an ultimate tensile
strength of 70.1 ksi (483 MPa), and an
elongation in 2 in. (51 mm) of 34.0
percent. Chemical analyses of the
three steel heats used for the plates
show an average carbon equivalent
(CE) of 0.35 with a range of 0.33 to
0.37. The mechanical properties of the
plate are on average less than those of
the stud steel. The plate steel can be
considered as very weldable for this
test program.
Summary
Tables 4 and 5 summarize the mea
sured tension and shear strength basic
material properties. The tensile and
57
yield strengths of each of these steel

heats exceed the AWS Dl.1-2000 re
quirements of 65 and 51 ksi (450 and
350 MPa), respectively. From these
tension and shear tests, it would be ex
pected that the shear strength of the
headed stud embedded in concrete
would be about 65 percent of the ten
sile strength, as shown in Table 5.
The average shear strength to tensile

strength ratio for rivets, a similar fas
tening material to studs, has been re
ported to be about 0.75. This rivet
shear to tensile strength ratio was seen
to vary from 0.67 to 0.83 for the rivet
tests. However, the data were indepen
dent of whether the rivet was driven or
undriven, or on the grade of the rivet
material. 16 The PCI Design Handbook

reduces the design shear strength of
headed studs by 10 percent presum
ably to account for the in-air shear
strength to tension strength ratio being
less than 1.0.
Earlier reported push-off test results
indicate a shear strength behavior that
Fig. 8. Plan view of shear test setup.

58
PCI JOURNAL
is better than what these material test

results imply.
SLAB TESTS
The majority of the shear tests in
this test program were conducted with
the anchors embedded in a concrete
slab. A slab specimen is more repre
sentative of the conditions used in pre
cast construction than the push-off
specimen.
Test Specimens
Two concrete slab sizes were se
lected for the investigation. The
larger slab was a 4 x lOft (1.22 x
3.04 m) slab with a 16 in. (406 mm)
thickness. Four of these slabs were
fabricated because of their efficiency
in locating edge shear tests. The large
interior areas of these slabs were used
for in-the-field tests, while the
perimeter edges accommodated nu
merous test specimens with small
edge distances. These large test slabs
permitted many tests to be conducted
without physically moving the speci
men; only the loading apparatus
HOLD
DOWN
SPACE FOR
HORIZONTAL
REACTION
STRUT
needed to be repositioned.
Twenty-seven, 5 ft (1.52 m) square
slabs were also fabricated in the test
program. The slab thickness was ei
ther 6 or 16 in. (152 or 406 mm), con
stituting the variable range for the slab
thickness effects. Because most in-thefield tests were contained in the larger
slabs, this smaller square specimen
proved efficient for laying out edge ef
fect tests where spacing between the
test samples needed to be large to
avoid overlapping concrete breakout
zones. The test anchorage locations
were established to minimize the slab
setup and handling during the tests.
Three additional 4 ft (1.22 m)
square slabs, 16 in. (406 mm) thick,
were cast during the course of the ex
perimental work. These slabs were
used to test conditions that were dam
aged by adjacent tests in earlier slabs
or where repeated tests were needed.
All anchorage plates were 1/2 in.
(12.7 mm) thick, Grade A36 (248
MPa) steel plate, having good weldability characteristics. Eight separate
plate sizes, in plan, were used in the
test program to accommodate the 14
different headed stud configurations.
HOLD
DOWN
L48ORA TORY
STRONG FLODR
Concrete selected for the test speci

mens was a commercially available
5000 psi (34.5 MPa), normal weight
concrete mix containing 3/4 in. (19.1
mm) limestone coarse aggregate. All
slabs were cast with the stud anchor
age specimens in the bottom of the
form to avoid conditions related to the
so-called top bar effect. This proce
dure was also used to ensure good
concrete consolidation around the in
dividual headed studs.
Reinforcement was only used in the
6 in. (152 mm) thick specimens for
handling purposes. In these slabs, the
reinforcement was placed so as not to
interfere with the stud anchorage
plates or provide confinement to the
anchorage.
Testing
All slabs were tested flat (horizon
tal) on the laboratory test floor. Fig. 8
shows an overall plan view of the
shear test specimen setup for a 5 ft
(1.52 m) square slab test. Fig. 9 shows
a cross-sectional view through the
load application framework. Two tri
angular frames, of the shape shown in
HOLD
DOWN
SHEAR TESTING RIG SECTION

Fig. 9. Cross-sectional view of shear test setup.
59
Fig. 9, were used to produce the shear

force reaction. The frames were located
8 ft (2.45 m) apart and anchored to the
laboratory strong floor.
A rectangular tube section, located be
tween the test specimen and the triangu
lar reaction frame, was used as a horizon
tal reaction beam for struts bearing against
the bottom of the slab. A double channel
spreader beam was elevated to the proper
height to react against the triangular test
frame columns. A gap between the double
channels accommodated a high strength,
threaded rod, which was positioned
through a center-hole ram and load cell.
At the concrete slab, the threaded
loadmg rod was attached to a pinned cle
vis, which in turn was bolted to a chan
nel shaped pulling device. As shown in
Fig. 10, the channel shape had a shoe
plate welded to the web (bottom), which
reacted on the backside of the stud an

chorage plate. This loading scheme was
used in order to eliminate almost all ec
centricity from the shear tests. Theoreti
cally, the actual eccentricity was one-half
the plate thickness or 1/4 in. (6.4 mm).
A threaded stud was used to tie
down the channel and prevent it from
kicking up while loading in the hori
zontal direction, as shown on Fig. 10.
This threaded rod was stud welded to the
exposed face of the plate prior to testing.
During the test, the threaded stud was
positioned in an oversized, slotted hole in
the channel. To further ensure no shear
resistance was provided at the threaded
stud tie down, a teflon-coated plate
washer was used below a finger-tight
ened nut.
All tests were instrumented with a
load cell and two linear variable dis
placement transformers (LVDTs). The

LVDTs measured the total movement of
the plate parallel to the applied load. Dig
ital data were collected through a data
acquisition system linked to a personal
computer. A peak-reading strain indica
tor, a voltmeter, and an analog x-y plotter
were used to obtain real-time test data.
Load was applied continuously with a
center-hole hydraulic ram loading a 150
ksi (1034 MPa) threaded rod.
Following each test, the failure surface
was documented with photographs and
pertinent geometry was measured in the
case of concrete breakout failures.
TESTS LOADING AWAY

FROM A FREE EDGE (de4)
Shear load on anchorages directed
away from a free edge is not commonly
I 150 ksi THREADED

ROD To RAM / TEST FRAME
PLAN
5/80 THREADED STUD
WELDED To ANCHOR PLATE
SECTION A
Fig. 10. Details of shoe for loading the embedded plate assembly.
60
PCI JOURNAL
encountered in precast construction.

However, special situations or framing
conditions may dictate use of this type of
connection. Pilot tests
33 conducted at the
University of Wisconsin-Milwaukee
(UWM), serving as a precursor for the
present study, showed a potential con
crete breakoutlpryout failure mode asso
ciated with this edge condition connec
tion. However, an eccentric shear
loading, brought on by the testing setup,
may have mduced the concrete breakout
observed in the pilot study. Conse
quently, a few test series were planned in
this program to evaluate the back edge
distance, de4.
In this study, 23 tests were conducted
with the shear force applied away from
the back free edge. This can be seen in
Fig. 1, which shows the shear force di
rection and the definition of the de4 back
edge distance. Three separate test series
were evaluated; two series had single
studs and the third series had two headed
studs oriented in one y-row.
The two series of single stud anchor
ages examined both /2 and /8 in. (12.7
and 15.9 mm) diameter studs. The two
stud anchorage groups used 1/2 in. (12.7
mm) diameter studs, spaced 21/4 in. (57.2
mm) or 4.5d apart. All three series were
tested in 16 in. (406 mm) thick speci
mens. The test series identifications for
the de4 tests are shown in Table 6. As
shown in Table 6, he/d ratios for these
tests were 5.3 and 5.93.
-D
0000.
00
0000000
C 0 C C C C
CC
n n
Sn
Sn
sn
Sn
-t
N N N
N N
N N
CCCC
00
00
00
00
000000
00
Sn
00
SC
Sn
Sn
Sn
00000000
00
N
N
000000000000
N C r
SC Sn
N N
N
N N
N N
-so
Do
;
:0 00 00
00
00
-sI-
Sr
C C C C C C
C-N
C--CN N N N N N
N N N N N N
C C N
N C
N N C-N rN
C-N
Os
N
00
-
-t
II
CCCCC
C
l- C
00
0
Sn Sn Sn Sn Sn
N N N N N N
N N N N N N
SI-s
Sn
C C C C C C C
--
C C C C C C C
d
0
00
SIN
C C
qQNNNNNJNNN
00 C Sn C C C SI
C N C N _00 C C
C C C C C Sn Sn
C.CCCNN
C C C Sn
N c C-N 00 Os
C Sn Sn C Sn C
CNNCNC
Sn N 00 N
00 N C 00
SnS(NSnSn
Sn
CC
C.QN.N.N.N.N.N.NJN.
N C 00 C C Cs 00
C t
NNNSINSn
C C C Sn IN Sn Sn
CCCNNNN
N N Cs
C N
N N C-N C-N N
IN C C Sn C Sn
NCCNSnN
N Sn Cs Sn
C-N C-N
t C-N
C 00
SO
0
a-so
555
II
C.)
.CCCCCCCCCC
N N C-N C-N -2- -2- Sn IN C sO
INSIN00CCVNSn
N N N Sn IN N N
CCCCC
N N sf s1- C C
C.
El
C
S
-q)
o
N
C C C C C C
N N N N N N N N N
Sn
4-.
CO
00
0-
Sn
.
t0
S
0
CCCCCCCCCC
C 00 00 00 00 00 C C C C
N IN 00 C C C 00 00 N N
0000S0000000NSnSn
CCCCCCC
00 C C-N N N C C
N N
Sn N
CCCC\0SnSn
C C C C C C
CC C-N C N C C
N
N N
C C 00 C Sn Sn
C C C C C C C
NNNrt
CC
Test Behavior
For the 1/2 in. diameter single stud
connection (identified as Series
VI 40_), five de4 edge distances were
evaluated with two tests performed per
edge distance. The five edge distances
were nominally 4d, 6d, 8d, lOd, and
12d. Eight of these tests failed due to
steel stud failure, and two failed in the
weld to the plate. After failure in all
cases, only minor concrete damage
was observed. Crushing of the concrete
in front of the stud was accompanied
by hairline, transverse cracks (cracks
perpendicular to the applied shear
load) propagating 2 to 4 in. (51 to 102
mm) either side from the stud center.
Seven tests were conducted (identi
fied as Series V141_) using /8 in.
(15.9 mm) diameter studs. Edge dis
tances evaluated were 4d, 8d, and 1 2d.
Three tests were conducted at the cbs-
NJ
N
t
ON
CCC
-so0
C-N
-S
CC
SC
(C
(C
N N
-D
2-
SC
CC
(C
(C
CC
C-N
N N N N N N
CC
CC
C-N
CC
CC
SC
CC
C-N
CC
N N N
(C
CC
C-N
N N N
II
C0
Co
S.C.)
I
2-
C C C C Cs Cs C Cs Cs
)
.
C C C C C
Cs
C C C C
t
-2- -2-2N N N N N N N
C 00 COO C C C
CCCCCCC
C C C C sO C
Cs C C C C C
Do
Co
5<0
CC
2
2
SI)
SI)
C)
II
.1
C C C C C C C C C C
C)
N N N
N N N
C
CC
I)
a)
Cr1
IL)
5-
C-INN
0
U
.
C-
CC
-.U<O0CC
<<-2
IN Sn
N N CC C-N
C
CCCCCCCCCC
NNC-NC-N
NNCCCC
>>>>>>>.>>>
>>>>>>>
>>->>>
ICe
EZ
C)
a)
NNN
zo
CC
C
Co
.0
C)
C.)
C
C
CO
CE
C
II.
Csi
L- o
II
-2 C
II
-
II
I!
0 C 0
ii
ii
Z N
0<
-(0
61
viqlt B.
4
Fig. 11. Steel shear failure of in. (16 mm) diameter stud with
a back edge distance (de4) of 4d.
est edge distance of 4d or 212 in. (63.5

mm). All tests at the three edge dis
tances failed in a steel shear mode,
with no weld failures occurring in this
series. Fig. 11 shows a representative
test specimen, Test VI411B, follow
ing stud failure. Based on these test re
sults, it was apparent that the back
edge distance, de4, variable is not a
factor that causes concrete breakout of
single stud anchorages.
The two stud anchorage series
(identified as Series V240_) used 1/2
in. (12.7 mm) diameter studs at nomi
nal edge distances of 4d, 8d, and 12d.
Six total tests were conducted in this
series. Two of the tests exhibited weld
failures in one or both of the studs,
while the other four tests failed by the
stud shearing through both of the stud
shanks.
Following failure, it was found that
at these anchorage locations, only lo
calized damage occurred to the con
crete in the form of concrete crushing.
Transverse splitting cracks were found
to propagate from the studs but were
arrested in a short distance. No con
crete cracking propagated longitudi
nally to the back free edge, even after
isopropyl alcohol was used to visually
aid in tracking cracks.
Fig. 12 shows a typical steel stud
failure for the two-stud anchorage
loaded away from the back free edge.
It is important to note that there was no
supplementary reinforcement around
or near any of these anchorages.
Table 6, Column 17, presents the
failure loads for all of these tests. Be
cause steel failure dominated the fail62
Fig. 12. Local concrete crushing and steel shear failure of

in. (12.7 mm) diameter studs with back edge distance (de4)
of 8d.
1/2
40.0
35.0
30.0
2541
20.0
15.0
10.0
5.0
0.0
0.0
0.1
0.2
0.3
Average Lateral Deflection -A (In.)
(a) de4 tests.

70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
0.0
0.1
Average Lateral Deflection -A
0.2
0.3
(in.)
(b) In-the-field tests.

Fig. 13. Load-deformation behavior for steel failure.
PCI JOURNAL
0)
0)
CD
(I)
CD
0
CD
2
2
2
2
2
2
2
2
V2F21A
V2F2IB
V2F22A
V2F22B
V2F3IA
V2F3IB
V2F32A
V2F32B
V4FOIA
V4FOIB
V4FO2C
V4FO2D
V4FI IA
V4EI lB
V4FI2A
V4FI2B
V2F2
Series
2 Y-rows
4.5d
7
V2F3
Series
2 Y-rows
v=7.Od
V4FO_
Series
2 X and Y
rows
V4FI_
Series
2 X andY
rows
x=4.5d,
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
(5)
Back
row
(BR)
2
2
2
2
2
2
2
2
2
2
2
2
0.496
0.496
0.495
0.495 I
2.63
2.63
4.87
4.87
2.63
2.63
4.87
4.87
0.496
0.496
0.495
0.495
,
2.63
2.63
4.87
4.87
2.63
2.63
4.87
4.87
6570
6570
6570
6570
5160
6570
5160
5160
6580
6700
6700
6490
6580
6700
6700
6700
6700
6700
6700
6580
2.63
2.63
4.87
4.87
-
6450
6450
6580
6700
2.63
2.63
4.87
4.87
(9)
(8)
Embed Concrete
depth strength
h(in.) f (psi)
0.496
0.496
0.495
0.495
0,496
0.496
0.495
0.495
0.496
0.496
0.495
0.495
1
1
1
1
2
2
2
2
0.496
0.496
0.495
0.495
(7)
Stud
dia.
d (in.)
1
1
1
(6)
Side
row
(SR)
5.30
5.30
9.84
9.84
5.30
5.30
9.84
9.84
5.30
5.30
9.84
9.84
5.30
5.30
9.84
9.84
5.30
5.30
9.84
9.84
5.30
5.30
9.84
9.84
hdd
Ratio
(10)
Note: Number of tests = 24; Columns 11 to 16 refer to Fig. 1 for geometric notation; Columns 17 and 18 ED
failure; note: 1 in, = 25.4 mm; 1 kip = 4.448 kN; 1 psi = 6.895 kPa; 1 ksi = 6.895 MPa.
4
4
4
4
4
4
4
4
2
2
2
2
V2FI IA
V2FIIB
V2F12A
V2FI2B
V2FL
Series
2Xrows
x=7.Od
x=4.5d
2
2
2
2
V2FO1A
V2FO1B
V2FO2A
V2FO2B
Test
no.
Group
information
(3)
(4)
No. of Front
row
studs
(FR)
(n)
V2FO
Series
2 X-rows
(2)
(1)
Table 7. In-the-field test summary.
___
________________________
18.00
30.00
30.00
30.00
14.50
14.50
30.00
30.00
45.75
30.50
38.75
22.50
18.00
55.75
21.75
21.75
46.25
34.25
22.25
19.50
28.25
28.25
28.25
22.25
19.50
15.75
19.50
19.50
67.75
79.75
91.75
103.75
20.25
27.25
42.75
58.25
I--
-,.
(15)
3.50
3.50
3.50
3.50
2.25
2.25
2.25
2.25
30.6
27.6
30.5
30.0
54.7
58.7
58.3
54.6
55.0
60.4
60.1
57.8
39.8
39.8
39.9
52.0
81.7
65.0
81.8
81.8
67.5
67.5
36.4
36.4
32.8
32.8
32.8
44.9
77.1
60.5
77.3
77.3
60.5
60.5
29.3
29.3
30.9
29.0
32.0
30.6
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
i
60.1
601
60.9
60.9
60.1
60.1
60.9
60.9
30.0
30.0
30.4
30.4
30.0
30.0
30.4
30.4
30.0
30.0
30.4
30.4
30.0
30.0
30.4
30.4
1.03
0.97
1.05
1.01
0.92
11,01
0.9
0.95
0.91
0.98
0.96
0.90
1.02
10.92
1.00
0.99
0.97
0.97
0.96
0.89
0.98
1.05
0.99
0.87
(21)
(221
(20)
Steel
Steel
Stud
strength capacity testl
Pred
F (ksi) (kips)
(predicted steel capacity); Column 23 stud = ductile failure of the steel stud and W-Stud
72.00
87.25
79.00
95.25
2.25
2.25
2.25
2.25
2.25
2.25
2.25
2.25
3.50
3.50
3.50
3.50
39.75
62.00
36.00
36.00
0.00
0.00
0.00
0.00
73.75
85.75 I
97.75
100.50
,.
29.1
29.1
29.1
27.2
29.4
31.4
30.0
26.5
62.8
62.8
62.9
62.9
60.5
36.3
36.4
36.4
64.5
64.5
60.6
60.6
58.2
58.2
58.3
58.3
60.5
36.3
36.4
36.4
64.5
64.5
60.6
60.6
(18)
(19)
(17)
V
ED Factors
1
Primary Secondary (kips)
de3/d [de3+yJId
___
2.25
2.25
2.25
2.25
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
y
(in.)
(16)
0.00
0.00
0.00
0.00
3.50
3.50
3.50
3.50
2.25
2.25
2.25
2.25
x
(in.)
--
____
52.25
40.25
28.25
16.25
96.25
89.25
77.25
61.75
30.75
15.00
82.00.
67.75
15.00
30.75
35.75
50.00
16.88
16.88
16.88
16.88
I..
16.00
16.00
18.00
18.00
de2
(in.)
(14)
del
(in.)
(13)
edge distance; Column 21 V. =
30.00
30.00
14.50
14.50
38.25
30.00
38.25
38.25
16.25
16.25
6.25
22.25
28.88
28.88
28.88
I 28.88
30.00
18.00
18.00
18.00
32.00
32.00
30.00
30.00
(12)
(11)
Test geometry
de4
de3
(in.)
(in.)
________
________
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
stud weld
Stud
Stud
Stud
W-Stud
Stud
Stud
W-Stud
W-Stud
(23)
Actual
failure
mode
Table 8. Steel failure test results.
studs
(a)
(4)
Stud
diameter
d (in.)
(5)
Tensile
stress
F, (ksi)
(6)
Steel
capacih
(kips)
(7)
Failure
load
(kips)
(8)
Test!
predict
(actual)
(9)
Test!
predict
(design)
1
I
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
0.494
0.494
0.494
0.624
0.624
0.624
0.624
0.624
0.495
0.495
0.494
0.494
0.494
0.624
0.496
0.496
0.496
0.496
0.496
0.496
0.496
0.496
0.496
0.496
0.496
81.6
81.6
81.6
78.0
78.0
78.0
78.0
78.0
79.1
79.1
81.6
81.6
81.6
78.0
77.7
77.7
77.7
77.7
81.6
81.6
77.7
77.7
77.7
77.7
77.7
15.6
15.6
15.6
23.9
23.9
23.9
23.9
23.9
15.2
15.2
15.6
15.6
15.6
23.9
30.0
30.0
30.0
30.0
31.5
31.5
30.0
30.0
30.0
30.0
45.0
16.1
16.1
14.3
22.6
25.6
1.03
1.03
0.91
22.8
25.1
14.3
14.5
15.0
15.8
18.2
25.0
28.1
30.4
30.1
27.7
27.5
27.6
30.4
28.5
29.9
28.4
38.1
1.07
1.03
0.96
1.05
0.94
0.95
0.96
1.01
1.16
1.05
0.94
1.01
1.00
0.92
0.87
0.88
1.01
0.95
1.00
0.95
0.85
1.29
1.29
1.15
1.14
1.29
1.23
1.15
1.26
1.14
1.16
1.20
1.27
1.46
1.26
1.12
1.21
1.20
1.10
1.09
1.10
1.21
1.13
1.19
1.13
1.01
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
1.09
1.12
1.06
1.04
1.07
1.10
0.94
0.97
0.90
0.97
0.91
1.36
1.40
1.27
1.25
1.30
1.34
1.13
1.15
1.13
1.22
1.14
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
1.01
1.05
1.01
1.13
1.09
1.12
1.11
1.16
0.96
0.91
1.01
1.12
1.07
1.00
1.01
0.90
1.03
1.04
1.05
1.27
1.32
1.27
1.41
1.37
1.40
1.40
1.45
1.15
1.10
1.22
1.35
1.28
1.20
1.21
1.07
1.23
1.24
1.26
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
(2)
(3)
Test
Number of
Test
number
type
(edge)
VI 102B
VI1O3A
VIIO3B
Viii lB
V11I2A
VII12B
VI1I3A
VI1I3B
V1122A
V1122B
V1152A
V1I52B
VII53B
V1163A
V21O2A
V2103A
V2lllA
V21IIB
V2124A
V2124B
V2161A
V2161B
V2174A
V2174B
V3174A
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
del
0.95
24.5
(10)
Failure
type
Stud
Stud
Stud
Stud
V1303A
V1303A
V1313A
VI3I4A
V1323B
\11323C
V2324A
V2324B
V2334A
V2334B
V2335A
de3
de3
de3
de3
de3
de3
de3
de3
de3
de3
de3
1
1
1
1
1
1
2
2
2
2
2.
0.494
0.494
0.624
0.624
0.495
0.495
0.496
0.496
0.496
0.496
0.496
81.6
81.6
78.0
78.0
79.1
79.1
77.7
77.7
81.6
81.6
81.6
15.6
15.6
23.9
23.9
15.2
15.2
30.0
30.0
31.5
31.5
31.5
17.0
17.5
25.3
24.9
16.3
16.7
28.3
29.0
28.3
30.7
28.7
V14OIA
V1402A
VI4O2B
V1403A
V1404A
V1404B
V1405A
V1405B
VI41IA
Vl4llB
VI4I1C
V1412A
V1412B
V1413A
V1413B
V2401B
V2402A
2402B
1
\
V2403A
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
de4
1
1
1
1
1
1
0.494
0.494
0.494
0.494
0.494
0.494
0.494
0.494
0.624
0.624
0.624
0.624
0.624
0.624
0.624
0.496
0.496
0.496
0.496
81.6
81.6
81.6
81.6
81.6
81.6
81.6
81.6
78.0
78.0
78.0
78.0
78.0
78.0
78.0
77.7
77.7
77.7
77.7
15.6
15.6
15.6
15.6
15.6
15.6
15.6
15.6
23.9
23.9
23.9
23.9
23.9
23.9
23.9
30.0
30.0
30.0
30.0
15.8
16.5
15.8
17.6
17.1
17.5
17.4
18.1
22.8
21.8
24.2
26.8
25.5
23.9
24.1
27.0
30.9
31.2
31.6
1
1
1
1
1
1
1
1
2
2
2
2
..
Note: Column 6: V = AF,, (steel capacity); Column 8: F, = actual from Column 5; Column 9: F, = 65 ksi (minimum design from Table 3); Column 10: stud = ductile failure of
the steel stud; W-stud = stud weld failure; note: 1 in. = 25.4 mm; 1 kip = 4.448 kN; 1 ksi = 6.895 MPa.
64
PCI JOURNAL
Table 8. (cont.). Steel failure test results.

(1)
Test
number
(2)
Test
type
(edge)
(3)
Number of
studs
(n)
V2FOIA
V2FO1B
V2FI 1A
V2FIIB
V2FI2A
V2F2IA
V2F2IB
V2F22A
V2F22B
V2F31A
V2F3IB
V2F32A
V2F32B
V4FOIA
V4FO1B
V4FO2C
V4FO2D
V4F1IA
V4FIIB
V4FI2A
V4FI2B
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
2
2
2
2
2
2
2
2
2
2
2
2
2
4
4
4
4
4
4
4
4
P012-10
P012-7
P012-8
P012-9
Push-Off
Push-Off
Push-Off
V1IOIA
VI lO2A
VII2IA
VI l53A
V2102B
V3173A
V2325A
V2325B
V2335B
V2375B
VI4OIB
V1403B
V2401A
V2403B
V2FO2A
V2FO2B
V2FI2B
del
del
del
del
del
del
de3
de3
de3
de3
de4
de4
de4
de4
Field
Field
Field
6
6
6
6
1
1
1
1
2
3
2
2
2
2
1
I
2
2
2
2
2
(4)
Stud
diameter
d (in.)
(5)
Tensile
stress
(ksi)
(6)
Steel
capacity
(kips)
(7)
Failure
load
(kips)
(8)
Testf
predict
(actual)
(9)
Test!
predict
(design)
Failure
type
0.496
0.496
0.496
0.496
0.495
0.496
0.496
0.495
0.495
0.496
0.496
0.495
0.495
0.496
0.496
0.495
0.495
0.496
0.496
0.495
0.495
77.7
77.7
77.7
77.7
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
77.7
77.7
79.1
79.1
30.0
30.0
30.0
30.0
30.4
30.0
30.0
30.4
30.4
30.0
30.0
30.4
30.4
60.1
60.1
60.9
60.9
60.1
60.1
60.9
60.9
29.4
31.4
29.1
29.1
29.1
30.9
29.0
32.0
30.6
30.6
27.6
30.5
30.0
54.7
58.7
58.3
54.6
55.0
60.4
60.1
57.8
0.98
1.05
0.97
0.97
0.96
1.03
0.97
1.05
1.01
1.02
0.92
1.00
0.99
0.91
0.98
0.96
0.90
0.92
1.01
0.99
0.95
1.17
1.25
1.16
1.16
1.16
1.23
1.15
1.28
1.22
1.22
1.10
1.22
1.20
1.09
1.17
1.17
1.09
1.09
1.20
1.20
1.16
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
Stud
0.494
0.494
0.494
0.494
82.4
82.4
82.4
82.4
94.7
94.7
94.7
94.7
84.7
84.1
86.5
96.6
0.89
0.89
0.91
1.02
1.13
1.13
1.16
1.29
Stud
Stud
Stud
Stud
0.494
0.494
0.495
0.494
0.496
0.496
0.496
0.496
0.496
0.496
0.494
0.494
0.496
0.496
0.495
0.495
(1.495
81.6
81.6
79.1
81.6
77.7
77.7
77.7
77.7
81.6
81.6
81.6
81.6
77.7
77.7
79.1
79.1
79.1
15.6
15.6
15.2
15.6
30.0
45.0
30.0
30.0
31.5
31.5
15.6
15.6
30.0
30.0
30.4
30.4
30.4
10.8
12.2
10.1
14.0
28.3
27.4
22.8
24.0
27.4
18.6
16.2
14.5
26.2
29.7
30.0
26.5
27.2
0.69
0.78
0.66
0.90
0.94
0.61
0.76
0.80
0.87
0.59
1.04
0.93
0.87
0.99
0.99
0.87
0.89
0.87
0.98
0.81
1.12
1.13
0.73
0.91
0.96
1.09
0.74
1.30
1.16
1.04
1.18
1.20
1.06
1.09
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
W-Stud
,.
(10)
Note: Column 6: V, = A,F, (steel capacity); Column 8: F, = actual from Column 5; Column 9: F,, = 65 ksi (minimum design from Table 3); Column 10: stud = ductile failure of
the steel stud; W-stud = stud weld failure; note: 1 in. = 25.4 mm; I Icip = 4.448 kN; I ksi = 6.895 MP&
ure mode, the test loads are evaluated

based on a steel capacity predicted
using 1.0 AF
. Column 19 lists the
51
predicted steel capacity and Column
20 shows the test-to-predicted ratio for
the tests.
The test-to-predicted capacities for a
majority of the stud failures are generally above 1.0. Tests V1411A,
Vi 411 B, and V240 1 B, which were
categorized as stud shear failures, exhibited lower test-to-predicted capacities of 0.90 to 0.96. A review of the
photograph for Test V2401B showed a

potential weldment problem in one of
the two studs. In assessing the weldment after failure, it did not exhibit
significant enough porosity (air voids)
to classify it as a weld failure,
The four weld failures encountered
in this test series also exhibited a
unique but variable behavior. Two of
the weld failures had test-to-predicted
capacities of 0.87 and 0.93; however,
the two remaining weld failures had
test-to-predicted capacities approxi-
mately equal to 1.0. Evidence of

porosity was observed within the cir
cumference of the weld; however, this
porosity appears to have had minimal
influence on the connection capacity
when compared to 1.0 AF
.
1
Fig. 13(a) shows two representative
load-deformation curves for the two
anchor tests. Test V2401A was found
to exhibit a partial weld failure as dis
cussed above and failed in a more brit
tie manner, but only after achieving
about 0.1 in. (2.5 mm) of deformation.
65
Companion Test V2401B was a stud

shearing failure. As shown in Fig.
13(a), this test showed good ductile
behavior.
studs used in this study. The test series

identifications relevant to the in-thefield tests are shown in Table 7.
Test Behavior
TESTS-IN-THE-Fl ELD
Some anchorages used in precast
concrete members are located a suffi
ciently large distance away from all
edges that all concrete breakout capac
ities are in excess of the capacity that
can develop by the individual studs
failing in steel shearing. This series of
testing was classified as in-the-field
tests. Six test series were planned and
tested in this program to test two- and
four-anchor connections. For these six
series, the emphasis was on evaluating
if the x- and y-row spacing had an ef
fect on capacity and if stud embed
ment depth had an influence.
This portion of the overall test pro
gram had 24 total tests. Each test se
ries used /2 in. (12.7 mm) diameter
studs. Two tests in the series used
studs with an effective embedment
depth (hef) of 2.69 in. (67.7 mm).
Longer studs having an hef = 4.87 in.
(124 mm) were used for the second
two tests in the series.
All 24 tests were conducted on 16
in. (406 mm) thick test slab speci
mens. Based on previous push-off
testing, discussed earlier, steel stud
failure can be achieved in relatively
thin slabs. As such, the influence of
slab thickness on the ability of an an
chorage to develop steel failure was
viewed to have little effect, especially
with the /2 in. (12.7 mm) diameter
Fig. 14.
Shear failure of
stud shank with
local crushing of
concrete.
66
Of the 24 tests conducted in the six

in-the-field test series, three of the an
chorages failed by fracture in the stud
weld. In a multi-stud anchorage, typi
cally one of the stud welds failed.
However, the entire anchorage was
classified as a weld failure.
Weld failures were found to occur
in the two stud anchorages evaluating
x-spacing effects. Review of these
weld failures showed two of the an
chorages failing at a test-to-predicted
ratio less than 0.90. The remaining
weld failure, Test V2FO2A, showed a
test-to-predicted ratio of 0.99.
This test also showed good defor
mation behavior prior to failure.
Moreover, the stud weld failure sur
face was marked by only minor poros
ity, estimated to be about 10 percent
of the stud cross-sectional area. There
fore, this test can be considered as a
steel failure, even though some weld
failure tendencies were observed. Fig.
13(b) shows representative load-defor
mation curves of two tests that sheared
the stud shank.
The test-to-predicted steel stud
shear capacities, in Table 7, ranged
from 0.90 to 1.05 for the six series of
tests examined in the in-the-field test
ing. When the long stud results are
compared with the short stud results
for all series, there is no discernable
increase or decrease in the ultimate
steel shear capacity due to stud

length.
With respect to stud spacing in the
two-stud connection, the test-to-pre
dicted capacity ratios are within 4 to 5
percent of each other, within the ex
pected scatter for the stud strength
data (discussed in the following sec
tion). The data in Table 7 show the
variation of the capacity with x- and y
spacing as the variable. Because of the
limited amount of data, the authors be
lieve that these results reflect the nor
mal scatter in steel strength rather than
an influence of spacing.
For the four-stud connection, the av
erage test-to-predicted capacity of the
connection having a larger y-spacing
(y 7.Od) is greater than those with a
smaller y-spacing of 4.5d. The capac
ity ratio for the two series are within 3
percent, which is again considered to
be within the scatter of the strength of
the stud material.
In summary, it appears that x- or y
spacings of 4.5d and 7.Od in different
combinations and loading orientations
did not have a significant effect on the
ultimate steel strength of the given
connection. Mean test-to-predicted
steel capacities ranged from 0.93 to
1.03 for the six series.
STEEL FAILURE ANALYSIS

As discussed in the previous sec
tions, a failure mode governed by steel
stud failure typically occurred for back
edge (de4) and in-the-field tests per
formed in 16 in. (406 mm) thick slabs.
Approximately one-half of the
database from this program on steel
shear failures is based on the de4 and
in-the-field testing. Additionally, steel
failures were obtained in the side edge
(del) and front edge distance (de3)
testing.
Steel failures, in these other two
testing regimes, occurred when the
edge distance was large enough to
transition from concrete breakout to
steel stud shear failure. Those steel
shear failures are compiled with the
other de4 and in-the-field tests result
ing in a shear strength steel capacity
database of 97 total tests, shown in
Table 8.
In all cases, steel failures were typi
cally marked by two failure modes: a
ductile, yielding-type failure of the
PCI JOURNAL
25
20
15
I:
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0......
..00
1..5
1.10
1.15
1.20
AIFa#ures
16
20
21
18
OrnitWejdFaikrres
14
19
19
17
Test/Predicted Steel Strength
Test/Predicted Steel Strength
stud or a stud weld failure at the plate

interface. Ductile failures were accom
panied by appreciable lateral deforma
tion. Figs. 13(a) and 13(b) showed
representative load-deflection curves
of a ductile steel shear failure of the
stud.
These figures illustrate that postelastic yielding in shear exists, as evi

denced by the flat topped portion of
the plot. After fracture, the stud failure
surface exhibited a shear yielding-type
profile indicative of good ductile be
havior. This is opposed to a failure
Fig. 15(a).
Histogram
showing the
distribution of
steel failure data
with actual
ultimate tensile
strength.
Fig. 15(b).
Histogram
showing the
distribution of
steel failure data
with minimum
design tensile
strength.
surface marked by porosity from inad

equate weld fusion. The ductile failure
surface was generally located in the
stud shank directly adjacent to the
weld flash.
When the corresponding failure area
on the concrete slab specimen was ob
67
served, the still embedded studs had

fracture surfaces that were ellipticalshaped in plan with the major axis par
allel to the load direction. The con
crete bearing surface in front of the
stud was locally crushed, with a pow
dery appearance. Also, because of the
elliptical nature of the stud failure sur
face and associated lateral deforma
tion, a void (pocket) behind the stud
was created; this void represented the
former location of the stud when the
concrete was cast.
An example of this classical failure
mode is illustrated in Fig. 14. As
shown in this figure, transverse cracks
perpendicular to the applied shear load
propagated from the concrete crushing
zone. These cracks were typically
quite shallow and were representative
of the stress state in the concrete just
prior to failure.
The second steel failure type experi
enced was failure of the stud weld to
the plate. Varying degrees of porosity
in the weld region confined within the
shank diameter typically marked the
weld failure surface. Porosity amounts
often ranged from 25 to 75 percent of
the shank area.
For this test program, the stud fail
ures due to welding appeared to be a
random occurrence. As an example,
WJE tested numerous multi-stud an
chorage plate configurations. In most
instances, only one of the studs in a
multi-stud anchorage would experi

ence a weld failure, whereas the re
maining studs appeared to have ade
quate welds because they failed by
shearing the stud shank. On any
given multi-stud connection plate, the
studs used were exactly the same, and
the settings on the stud-welding ma
chine remained unchanged, confirm
ing the apparent randomness of the
weld failure.
Failure load achieved for an anchor
age governed by steel stud shank fail
ure can typically be predicted using
the ultimate tensile strength of the
stud. In other words, the connection
shear strength can be calculated by the
number of studs times the stud area
multiplied by the ultimate tensile
strength of the material. Weld failures
typically occurred at steel stresses less
than the ultimate tensile strength.
All Steel Failures

WJE reviewed the steel shear failure
data by first evaluating the actual ten
sile strength of the stud material. Ulti
mate tensile stresses were obtained for
each stud heat and stud size, as re
ported earlier. WJE then compared the
calculated A
F strength to the actual
5
load in shear, as shown in
failure
test
Column 8 of Table 8. Fig. 15(a) pre
sents a histogram of all the steel fail
ures, with all steel failures shown as
Table 9. Steel test-to-predicted statistics using actual ultimate and design

tensile strengths of the studs.
All failures
Design
Test
F, = actual
(mm)
= 65 ksi
value
Statistic
Stud failures only

Design
Test
(mm)
F,, = actual
= 65 ksi
value
Weld fatlures
17
17
Total tests
97
97
80
80
Average
0.97
1.18
1.00
1.21
Median
0.98
1.18
1.00
1.20
Standarddeviation
0.105
0.132
0.071
0.095
Variance
0.011
0.017
0.005
0.009
11.2 percent
7.1 percent
7.8 percent
COV
10.8 percent
,4,
Kfactor
1.95
1.95
1.96
1.96
5 percent fractile value
0.76
0.92
0.86
1.03
Note: 1 ksj = 6.895 MPa.

68
solid bars.
For the 97 tests representing all steel
failures (steel shank and weldment),
WJE obtained an average test-to-pre
dicted ratio of 0.967 and a standard
deviation of 0.105 as presented in
Table 9. The variance for this sample
is 0.011 and the coefficient of varia
tion (COV) is 10.8 percent. These
statistics include weld failures in the
database. When the data are analyzed
for the 5 percent fractile value using a
90 percent confidence level, one ob
tains a ic factor of 1.95 for the popula
tion of 97 samples.
Using the probability distribu
tion information, summarized by
34 the 5 percent frac
Wollmershauser,
tile value is 0.76. Thus, for all of the
non-concrete break-out failures, WJE
data indicates the characteristic steel
capacity, using the actual steel tensile
strength, can be predicted by the fol
lowing formula:
Vsteei
0.76 AsFug(actual)
(3a)
where
nominal shear strength of
a single headed stud or
group of headed studs
steel
governed by
strength (lb)
= effective cross-sectional
5
A
area of a stud anchor (sq
in.)
Fut(actual) = actual ultimate tensile
strength of headed stud
steel in tension (psi)
The above value was obtained using
the actual ultimate tensile strength
(F,) of the headed stud steel. When
the connection capacity is predicted
using the design steel ultimate strength
value of 65 ksi (450 MPa), the test-topredicted ratio averages 1.18 with a
standard deviation of 0.132. The vari
ance is 0.017 and the COV is 11.2 per
cent. Therefore, the design equation
for all steel failures using a design
= 65 ksi (450 MPa) re
minimum
duced for the 5 percent fractile value
becomes:
Vsteei
Vsteel(design)
0.92 AsFut(design)
(3b)
where
Vstee1(esign)
nominal shear strength

of a single headed stud
or group of headed
studs governed by steel
strength (lb)
PCI JOURNAL
effective cross-sectional
area of a stud anchor
(sq in.)
(design) = design minimum ten
sile strength of headed
stud steel in tension,
per Table 3 (psi)
A histogram of test-to-predicted ra
tios using the design ultimate steel
strength is presented in Fig. 15(b).
Given that the minimum design value
of 65 ksi (450 MPa) ranges from 16 to
20 percent lower than the actual mea
sured ultimate strength values of studs
used in this program, a right shift in
the Fig. 15(b) histogram is observed.
Moreover, only six of the tests showed
test-to-predicted ratios less than one
by this analysis.
A
STUD HEAT
AFFECTED ZONE
Steel Shank Failures Only

When the weld failure data are
omitted from the population, the WJE
database is representative of stud steel
shear failures only. When this
database is independently reviewed,
the total test population becomes 80
with an average test-to-predicted ratio
of 0.996 as shown in Table 9. The
standard deviation on this sample is
0.071, thus indicating the relative
tightness of the data.
Furthermore, the variance is 0.005
and the COV is 7.1 percent. This data
set is illustrated in Fig. 15(a) as crosshatched bars; a right shift in these data
can be observed relative to the data
base with all steel failures. In this dis
tribution, the data are also grouped
tighter about the mean.
Given the data base of steel stud
shank shear failure, the K factor for
the 5 percent fractile analysis in
creases slightly to 1.957. Given this
value, the characteristic strength pre
diction equation when the ultimate
tensile strength is known becomes:
Vsteei = 0.86 AsFut(actual)
(4a)
By examining only the stud steel

failures, the 5 percent fractile value
increases by 13 percent over Eq. (3a).
A further increase is observed when
the minimum design ultimate strength
of 65 ksi (450 MPa) is used to ana
lyze the data. By using the design ul
timate strength, the average test-topredicted ratio is 1.21 with a standard
INDICATES LOCATION OF
ROCKWELL B HARDNESS TEST
HARDNESS READING AT STUD WELD

Fig. 1 6. Cross section of a stud shank, weld zone and anchorage plate showing
hardness results.
deviation of 0.095. The variance for

this analysis is 0.009 and the COV is
7.8 percent. The 5 percent fractile
characteristic prediction equation
thus becomes:
Vsteel(design)
1.0 AsFut(design)
(4b)
In this case, the 5 percent fractile

coefficient actually exceeds 1.0, so 1.0
becomes the default maximum value.
From a probability standpoint, this in
dicates with 90 percent confidence
that over 95 percent of the failure
loads occur at a value represented by
Eq. (4b) above, using the minimum
design ultimate steel strength of 65 ksi
(450 MPa). From the histogram shown
in Fig. 15(b), no tests had test-to-pre
dicted ratios less than 1.0, using the
WJE data.
Steel Failure Behavior

An explanation of the apparent in
crease in the steel shear strength when
embedded in normal weight concrete,
as compared to the shear strength re
suits in air, can be founded in the met

allurgy of the stud weld metal. In the
stud welding process, the shielded arc
weld melts the stud end and a shallow
weld pooi is created beneath the stud.
The stud gun plunges the stud into the
molten weld metal and holds the stud
in position while the liquid metal so
lidifies. Although this process occurs
over a very short period of time, a
heat-affected zone (HAZ) is created in
the weldment.
The American Welding Society de
fines the HAZ as that portion of the
base metal where the mechanical
properties or microstructure have been
influenced by the heat of welding. The
heat developed tends to heat-treat or
temper the steel such that locally the
steels strength and hardness will in
crease. This transformation hardening
process is dependent on the initial ma
terial temperature after arcing, the rate
of cooling, and the final (ambient)
35
temperature.
Fig. 16 shows a cross section of a
69
stud weld from this PCI research pro

gram. The anchorage, which in this
case failed by a concrete breakout
mode, was submitted to Nelson Stud
36
Welding for metallurgical work.
The numbers on the cross section are
locations where Rockwell B Hard
ness tests were performed on the
weld and base metal cross section;
the locations are shown in scale. The
Rockwell B Hardness values were
then approximately related back to
ultimate tensile strength of the re
spective material.
37
As shown in Table 10, the approxi
mate tensile strengths are greater in
the HAZ and immediate surrounding
vicinity. Although a conversion of
hardness to tensile strength is an ap
proximation, the purpose herein is to
show the relative steel strength inTable 10. Rockwell B Hardness
and equivalent tensile strengths of
the stud weld area (see Fig. 16).
Tensile
strength
(ksi)
Hardness
test
location
Rockwell B
hardness
90.1
93.5
96.0
91.8
92.0
95.1
102.0
101.5
122.0
99.2
89.9
DISCUSSION OF
PROPOSED ACI CODE
REQUIREMENTS
88.0
Discussed below are steel failure ca

pacity, pryout capacity, and lightweight
concrete effect.
Steel Failure Capacity
115.0
87.5
79.0
87.5
84.0
10
102.5
125.0
11
106.5
143.0
12
92.3
91.4
13
82.5
75.0
14
93.8
97.6
15
102.1
123.5
16
99.8
115.6
17
879
866
18
81.6
74.4
Note: 1 ksi
70
6.895 MPa.
crease in the weld area.

In relating Fig. 16 to the steel shear
failures encountered in this study, the
stud typically sheared off above the
weld flash region in the parent stud
material. Hardness locations 2, 9, and
14 in Fig. 16 exhibit converted tensile
strengths of 96, 84, and 97.6 ksi (662,
579, and 673 MPa), respectively. Ear
lier tensile testing of this particular
stud heat presented in Table 4, re
vealed an average ultimate tensile
strength (F) of 78 ksi (538 MPa).
Therefore, on a relative basis, the indi
cated tensile strength in the area of the
weld is between 10 to 20 ksi (60 to
138 MPa) higher.
The increase in steel strength at the
weld seems to be the most plausible
cause for the greater than expected
steel shear strength results compared
to those measured in air. Double shear
tests of the studs conducted in the
main shaft body showed good correla
tion with a typical reduced shear yield
value.
The purpose of presenting both the

combined steel and weld shear failures
and the stud-only shear failures intro
duces an important issue that must be
considered in the design equations. Of
the equations presented above, only
the second set of equations [Eqs. (4a)
and (4b)] represent the characteristic
value of the connection in normal
weight concrete if it is to fail due to
steel shear. Eq. (4a) represents the ul
timate load for a ductile steel shank
failure with actual tensile strength
being used, whereas Eq. (4b) uses the
minimum design ultimate strength of
65 ksi (450 MPa).
For design, the equations presented
above would typically be reduced by a
strength reduction or phi (4)) factor.
Arguably, the phi, or material reduc
tion factor, should incorporate the
probability that failure could occur
due to a weld failure. If a value of 4) is
assumed to be 0.75, as proposed in

4 all tests
Appendix D of the Code,
would have failure stresses above the
ultimate design stress. Additional
safety factors are then built in by using
a strength design criterion.
Another consideration involves the
ultimate tensile strength. In this study,
the value of F, ranged from 78 to 83
ksi (538 to 572 MPa) depending on
stud heat and size. The current AWS
call for a minimum design
3
standards
tensile strength of 65 ksi (450 MPa).
Ultimate tensile strengths of studs
used in this study were about 20 to 25
percent greater than that used for de
sign. Eq. (4b) is reflective of the data
analysis using the minimum design ul
timate strength.
Both Eqs. (4a) and (4b) above repre
sent the calculation of a connection
steel shear capacity, which is greater
than the initial proposals for the new
ACT 318-2002 Building Code provi
sions. The initial approach to steel
failure was to lump embedded bolts
with headed studs. The proposed ACT
provisions provided the following two
equations for computing the capacity
of an anchorage group governed by
steel failure:
A. For a fastener material with a
well-defined yield point:
Vs
flAsefy
B. For cast-in fasteners without a

well defined yield point:
V
flO.6Asefut
These initial ACT proposed provi

sions set the capacity to the steel yield
strength if the material has a well de
fined yield. If the embedded bolt or
stud material does not have a well-de
fined yield point, which is the usual
case, ACI defines a shear yield point
of 0.6f,.
The Appendix D design procedure
involves calculating a basic connec
tion capacity defined by the steel ma
terial. The WJE study has shown the
steel capacity is governed by the ulti
mate tensile strength of the fastener
material, in this case a steel stud. Al
though the loading on these connec
tions was almost pure shear, shear
yield did not govern the steel capacity.
Under the shear yield criterion from
the Energy of Distortion Theory by
PCI JOURNAL
6.0
Proposed P,y-Out Equation
5.0
Normal Veight Concrete
0
0
o ACI Appendix D Pry-Out

Equation
Ci
CD
08
0
c 2.0
=
55OnA.
5
(f)
S.
1.0
0.0
Embedment Depth Ratio (h

, / d)
0
Fig. 1 7. Comparison of the proposed pryout equation to the AC! Appendix D proposal.
Huber-von Mises-Hencky, the ulti

mate values would be reduced to yield
(F O.58F
) or ii-J if shear yield
1
governed. A reduction due to shear
yield, as implied by V = n0.6AJ
,
1
was not observed in the behavior and
strength capacity prediction in tests re
ported here, and the reduction is not
justified.
provisions in ACI Appendix D

4 are as
follows:
The nominal pryout strength,
shall not exceed:
Vep
The ACT Appendix D requirements

for pryout appear to be somewhat
cumbersome and are based only on the
effective embedment depth, hef. As
shown in this study and in previous
studies, the critical parameter for eval
uating the likelihood of a concrete
pryout failure for headed studs is the
held ratio. The critical value proposed
by analysis of data in this study is 4.5,
which is slightly greater than a value
of 4.2 proposed by Driscoll and Slutter
almost 40 years ago. Fig. 3(a) shows
the WJE test data added to the earlier
results for further reference.
The ACT concrete pryout capacity
load requires the calculation of the ten
sile breakout capacity based on com
puting the effective area of the CCD
physical model breakout surface. The
kcpNcb
where
coefficient for pryout strength
1.0 for hef< 2.5 in.
= 2.0 for hef> 2.5 in.
= nominal concrete pryout
strength (lb)
Nb = nominal concrete breakout
strength in tension of a single
anchor (lb)
The term Nb is the concrete tensile
pull-out strength and is determined in
accordance with the ACT 318 Ap
pendix D requirements.
When the normal weight concrete
push-off data with one y-row and held
<4.5 are evaluated with the above pro
cedure, the ACT predicted results can
be overly conservative for headed
studs, as depicted in Fig. 17. The in
herent conservatism of the ACT equa
tion occurs when the
factor be
comes 1.0, among other factors. Better
predicted capacities were calculated
with Eq. (2) for tests where pryout is a
potential, as was shown in Fig. 3(b).
=
Pryout Capacity
It must be recognized that Eq. (2)

was formulated based on data that
represented some pryout results. Re
gardless, considering that this pushoff database seems to represent the
best information for short, stocky
studs available at this time, the above
Appendix D pryout equation is not
recommended for use with headed
stud anchor connections under those
circumstances.
Because pryout failures will only
govem in those (rare) instances when
short, stocky studs are used (held <
4.5), a simple equation is viewed to be
desirable. For computational ease,
Eqs. (1) or (2) have been shown to be
good prediction equations. Direct con
version of Eq. (1) yields a constant of
977 for Eq. (2), but the constant was
reduced to 800 for a lower bound solu
tion. However, neither equation repre
sents a characteristic capacity based
on a 5 percent fractile value using a 90
percent confidence level. By review
ing the push-off database for stocky
stud (held < 4.5) tests, the proposed
characteristic capacity equation for
headed stud pryout is:
V;;p
550n2A4
(5)
71
where
V = nominal shear strength (Ib)
n = number of anchors in a con
nection group
A
=
specified
compressive strength
f,
of concrete (psi)
= concrete unit weight factor
Predicted capacities based on pro
posed Eq. (5) are plotted in Fig. 17 as
solid diamond shapes. Test-to-pre
dicted ratios exceed 1.0 for the embed
ment depth ratios (held) from 2.0 to
4.0. Moreover, proposed Eq. (5) is
shown as a better predictor of pryout
capacity because the data points are
tightly grouped with less scatter. The
ACI equation, data points plotted as
open circles, generally tracks above Eq.
(5) predictions and has more scatter.
Lightweight Concrete Effect
Eqs. (1) and (2) have been shown to
provide a good lower bound limit on
stud strength in lightweight concrete,
regardless of embedment depth. The
steel shear failure equation in ACI 318
Appendix D does not acknowledge
lightweight concrete as a variable and
that a stud may not develop its full ul
timate tensile strength in lightweight
concrete.
When Eq. (2) is evaluated using the
lightweight aggregate concrete test
data from the push-off tests, Eq. (5)
again is a better predictor equation
for the characteristic value of capac
ity. The proposed Eq. (5) is applica
ble for all stud embedment depths.
Eq. (5) incorporates a concrete unit
weight factor, A, which accounts for
the lightweight concrete effect.
Eq. (5) is thus proposed as the pre
diction equation for V, and V when a
headed stud connection is embedded
in lightweight concrete away from all
edge influences for all embedment
depths. The upper limit on V, will be
V. = l.0A
.
1
F
5
Further research into steel capacity
in lightweight aggregate concrete may
modify this capacity. Therefore, Eq.
(5) is proposed to serve a two-fold ca
pacity: determination of maximum
stud capacity for anchorages located in
lightweight concrete and avoidance of
pryout failure in normal weight con
crete for headed studs with held < 4.5.
72
CONCLUSIONS
AND DESIGN
RECOMMENDATIONS
Given the results of the work re
ported herein, the following conclu
sions and design recommendations are
presented:
Away from a Free Edge (de4)
1.For both the 1/2 and /8 in. (12.7 and
15.9 mm) diameter studs used in
this study, steel stud failure was
achieved at the minimum back edge
distance of 4d. Closer back edge
distances are not practical, consider
ing clear cover requirements.
2. The WJE shear test load application
produced a minimal amount of ec
centricity on the connection, such
that the anchorage was loaded in
practically a pure shear condition.
Small eccentricities on such an an
chorage, as reported in the literature,
may be sufficient enough to alter the
behavior and failure mode due to
the shear-tension interaction.
In-the-Field
1. When an anchorage is placed suffi
ciently away from all edges to
negate all edge influences, termed
in-the-field of the member, the con
nection will achieve a capacity
based on all headed studs failing in
a steel failure mode. Assuming the
weld quality is adequate, the failure
will be ductile with appreciable lat
eral deformation.
2. For the two- and four-stud anchor
ages tested in this program, the x
and y-spacings of 4.5d and 7.Od in
different combinations and load ap
plication direction had no influence
on the failure mode, that is, steel
stud shear, and did not have a sig
nificant effect on the ultimate capac
ity of the anchorage. From a review
of the literature, the proposed ACT
minimum x-spacing in an anchorage
of 4d seems appropriate.
3. Although the WJE tests did not ex
hibit a reduction in anchorage ca
pacity, large y-spacings of the studs,
as reported in the literature, can pro
duce a shear lag effect. Shear lag re
duces the efficiency of the connec
tion failing in a stud steel mode.
Capacity Based on Steel Failure

1. For anchorages governed by steel
stud shank failure, this study shows
the ultimate shear failure load is best
predicted using the ultimate tensile
strength of the stud. Eq. (4b) is rec
ommended as the steel prediction
equation (V
) for headed studs in nor
5
mal weight concrete with held> 4.5.
2. Steel stud failures in this study were
achieved with well-embedded studs.
It is proposed that a well-embedded
stud have a minimum effective em
bedment-to-diameter ratio (held) of
4.5, as determined by reviewing tests
in the literature. The minimum held
ratio of studs used in this study was
5.30.
3. Headed studs with an held less than
4.5 will precipitate a failure mode
known as pryout. This failure mode
produces an ultimate capacity less
than that predicted by 1.0 ASFUt(desjgn).
Eq. (5) is proposed to predict the ca
pacity for short, stocky studs having
held<4.5.
4. A review of limited test data on stud
anchorage failures in lightweight
concrete shows that the stud cannot
develop its full capacity based on 1.0
. From the information available
1
F
5
A
at this time, a maximum stud capac
ity based on Eq. (5) is proposed. Eq.
(5), which is a concrete predictor
equation, is appropriate because the
ultimate connection capacity in
lightweight concrete is dependent on
a combination of local concrete
crushing and increased stud deforma
tion.
5.The minimum connection plate
thickness of one-half the stud diame
ter (0.5d) presently in the PCI Hand
book is slightly conservative for a
pure shear loaded connection. The
practical minimum to achieve steel
stud failure was experimentally de
termined to be 0.37d or (
/8)d. as re
3
ported in the literature. Increased
plate thickness may be required for
bending resistance or to ensure a
more uniform load distribution to the
attached studs. The plate thickness
used in this study was /2 in. (12.7
mm), thus exceeding the minimum
suggested limits.
Based on the results of this study
summarized above, concise design
recommendations for the shear capac
PCI JOURNAL
ity of an anchorage failing in a steel

stud failure mode are conveniently
presented in the flowchart of Fig. 18.
Research Needs
Additional work is needed in deter
mining the ultimate capacity of a con
nection located in-the-field of a
lightweight concrete member. To date,
most work on cast-in headed stud an
chorages has been performed in nor
mal weight concrete. Lightweight con
crete is used by the precast concrete
industry in numerous applications,
many of which use connections with
large distances from member edges.
Connections with a large y-spacing
of studs, or an overall large out-to-out
y-spacing, located in-the-field of a
member need further study with re
spect to shear-lag effects. A determi
nation of the maximum individual or
overall y-spacing permissible to pre
clude shear lag reduction would seem
appropriate. Moreover, an efficiency
factor for connections affected by
shear lag may be required. Guidance
for this influence can be found in re
search for load transfer mechanisms in
long, high strength bolted connections.
ACKNOWL.EDGM[NTS
wishes to express its appreciation to the
for sponsoring this research project. In
particular, the authors wish to thank
PCIs Research and Development Com
mittee and the members of the Industry
Advisory Group (Thomas J. DArcy,
chairman) for their constructive com
ments during the course of this project.
Gratitude is expressed to Harry
Chambers and Nelson Stud Welding,
Elyria, Ohio for their contributions of
technical training, stud material dona
tion, stud welder use, and additional
laboratory support in Ohio. Thanks
C,)
C.)
C
C,)
where
A = cross-sectional area of stud (sq in.)

5
d = nominal stud diameter (in.)
hei effective embedment depth of stud (in.)
= ultimate tensile strength of stud (=65 ksi)
n = number of studs in anchorage
= 1 .0 for normal weight concrete
= 0.85 for sand lightweight concrete
= 0.75 for all lightweight concrete
Fig. 18. Flowchart showing recommended maximum capacity of a headed stud
anchorage.
are also expressed to Thomas Soukup

and Michael Pistilli of Prairie Materi
als, Bridgeview, Illinois for their gra
cious donation of the (consistent)
concrete material for the project,
Kevin Heinert of Williams Form En
gineering for the high-strength coil
rod used as part of the loading frame
work, and Roy Edgar of DaytonRichmond for supplying slab lifting
hardware.
Publications cited in the literature

were oftentimes difficult to locate, espe
cially the push-off literature and reports
from the l960s. Special thanks is ex
tended to Dr. James Baldwin, Civil En
gineering Professor Emeritus, Univer
sity of Missouri-Columbia for locating
and loaning WJE numerous out-of-print
University of Missouri research reports
and engineering experimental station
bulletins.
73
REFERENCES
1. PCI Design Handbook, Fifth Edition (PCI MNL 120-99), Pre
cast/Prestressed Concrete Institute, Chicago, IL, 1999.
2. AU Committee 349, Code Requirements for Nuclear Safety
Related Concrete Structures (ACI 349-97), ACI Manual of
Concrete Practice, Part 4, American Concrete Institute, Farm
ington Hills, MI, 2000.
3. ACI Committee 318, Building Code Requirements for Struc
tural Concrete (ACI 318-99) and Commentary (ACI 31 8R99), American Concrete Institute, Farmington Hills, MI,
1999.
4. ACI Committee 318-B, Appendix D Anchoring to Concrete,
Code CB-30, June 22, 2000, Draft Version, American Con
crete Institute, Farmington Hills, MI, 17 pp.
5. Fuchs, W., Eligenhausen, R., and Breen, J. E., Concrete Ca
pacity Design (CCD) Approach for Fastening to Concrete,
ACI Structural Journal, V. 92, No. 1, January-February 1995,
pp. 73-94.
6. Anderson, N. S., and Meinheit, D. F., Design Criteria for Headed
Stud Groups in Shear, PCI Research Report, PrecastlPrestressed
Concrete Institute, Chicago, IL, under preparation.
7. Viest, I. M., Investigation of Stud Shear Connectors for Com
posite Concrete and Steel T-Beams, Journal of the American
Concrete Institute, V. 27, No. 8, April 1956, pp. 875-891.
8. Driscoll, G. C., and Slutter, R. G., Research on Composite
Design at Lehigh University, Proceedings, AISC National
Engineering Conference (May 11-12, 1961), Minneapolis,
MN, 1961, pp. 18-24.
9. Ollgaard, J. G., Slutter, R. G., and Fisher, J. W., Shear
Strength of Stud Connectors in Lightweight and NormalWeight Concrete, AISC Engineering Journal, V. 8, No. 2,
April 1971, pp. 55-64.
10. Baldwin, Jr., J. W., Composite Bridge Stringers Final Re
port, Report 69-4, Missouri Cooperative Highway Research
Program, Missouri State Highway Department and University
of Missouri-Columbia, MO, May 1970, 62 pp.
11. Buttry, K. E., Behavior of Stud Shear Connectors in
Lightweight and Normal-Weight Concrete, Report 68-6, Mis
souri Cooperative Highway Research Program, Missouri State
Highway Department and University of Missouri-Columbia,
MO, August 1965, 45 pp.
12. Dallam, L. N., Push-Out Tests of Stud and Channel Shear
Connectors in Normal-Weight and Lightweight Concrete
Slabs, Bulletin Series No. 66, Engineering Experiment Sta
tion, University of Missouri-Columbia, Columbia, MO, April
1968, 76 pp.
13. Goble, 0. G., Shear Strength of Thin Flange Composite Spec
imens, AISC Engineering Journal, V. 5, No. 2, April 1968,
pp. 62-65.
14. Chinn, J., Pushout Tests on Lightweight Composite Slabs,
AISC Engineering Journal, V. 2, No. 4, October 1965, pp.
129-134.
15. Hawkins, N. M., The Strength of Stud Shear Connectors, Re
search Report No. Rl41, Department of Civil Engineering, Uni
versity of Sydney, Sydney, Australia, December 1971, 34 pp.
16. Kulak, G. L., Fisher, J. W., and Struik, J. H. A., Guide to De
sign Criteria for Bolted and Riveted Joints, Second Edition,
John Wiley and Sons, New York, NY, 1987, 333 pp.
17. Viest, I. M., Studies of Composite Construction at Illinois and
Lehigh, 1940-1978, Composite Construction in Steel and
Concrete III, Proceedings of an Engineering Foundation Con
ference, Irsee, Germany (June 9-14, 1996), Edited by C. D.
Buckner and B. M. Shahrooz, American Society of Civil Engi
neers, New York, NY, 1997, pp. 1-14.
-
74
18. Seely, F. B., and Smith, J. 0., Advanced Mechanics of Materi

als, Second Edition, John Wiley & Sons, Inc., New York, NY,
1952, .
91
76
pp.
19. AASHO, Standard SpecifIcations for Highway Bridges, Eighth
Edition, American Association of State Highway Officials,
Washington, DC 1961.
20. AISC, Manual of Steel Construction, Eighth Edition, Ameri
can Institute of Steel Construction, Chicago, IL, 1980.
21. Shaikh, A. F., and Yi, W., In Place Strength of Welded
Headed Studs, PCI JOURNAL, V. 30, No. 2, March-April
1985, pp. 56-81.
22. PCI Design Handbook, Fourth Edition (PCI MNL 12092), Precast/Prestressed Concrete Institute, Chicago, IL,
1992.
23. PCI Design Handbook, First Edition, Prestressed Concrete In
stitute, Chicago, IL, 1971.
24. PCI Design Handbook, Second Edition, Prestressed Concrete
Institute, Chicago, IL, 1978.
25. PCI Design Handbook, Third Edition, Prestressed Concrete In
stitute, Chicago, IL, 1985.
26. AISC, Manual of Steel Construction Load & Resistance Fac
tor Design (LRFD), Volume I (Structural Members, Specifica
tions & Codes), Second Edition, American Institute of Steel
Construction, Chicago, IL, 1994.
27. CISC, Handbook of Steel Construction, Fifth Edition, Canadian
Institute of Steel Construction, Willowdale, Ontario, Canada,
1991.
28. AASHO, Standard Specifications for Highway Bridges, Ninth
Edition, American Association of State Highway Officials,
Washington, DC, 1965.
29. AWS, Supplement to AWS D1.O-66, Code for Welding in
Building Construction and AWS D2.O-66, Specifications for
Welded Highway and Railway Bridges on Requirements for
Stud Welding, September 1968, American Welding Society,
Inc., New York, NY, 12 pp.
30. AWS, Structural Welding Code, AWS Dl.1-82, Sixth Edition,
American Welding Society, Miami, FL, 1982.
31. AWS, Structural Welding Code Steel, AWS Dli :2000, 17th
Edition, American Welding Society, Miami, FL, 2000.
32. ASTM, Standard Specification for Steel Bars, Carbon, ColdFinished, Standard Quality (ASTM A108-99), Volume 04.07,
American Society for Testing and Materials, West Con
shohocken, PA, 1999.
33. Kuhn, D. P., and Shaikh, A. F., Pilot Study on Headed An
chor Studs: A Comparison Between PCI and CCD, Interim
Report, Department of Civil Engineering, University of Wis
consin, Milwaukee, WI, August 1997, 29 pp.
34. Wollmershauser, R. E., Anchor Performance and the 5%
Fractile, Hilti Technical Services Bulletin, November 1997,
Hilti, Inc., Tulsa, OK, 5 pp.
35. Linnert, G. E., Welding Metallurgy Carbon and Alloy Steels,
Volume I Fundamentals, Fourth Edition, American Welding
Society, Miami, FL, 1994.
36. Nelson Stud Welding, Hardness Profile Testing of Wiss, Jan
ney, Elstner/Precast Concrete Institute Shear Test Specimen,
Test Report No. 2000-2, Nelson Stud Welding, Elyria, OH,
2000, 5 pp.
37. ASTM, Standard Hardness Conversion Tables for Metals (Re
lationship Among Brinell Hardness, Vickers Hardness, Rock
well Hardness, Rockwell Superficial Hardness, Knoop Hard
ness, and Scleroscope Hardness) (ASTM E140-97e2), Volume
03.01, American Society for Testing and Materials, West Con
shohocken, PA, 1999.
PCI JOURNAL
APPENDIX
effective cross-sectional area of stud anchor (sq in.)

effective cross-sectional area of stud anchor (sq in.)
(Appendix D notation)
d = shaft diameter of headed stud (in.)
del = side edge distance normal to shear load application
direction, parallel to x-axis, taken from center of
anchor shaft to the side concrete edge (in.) (see Fig. 1)
de2 = side edge distance normal to shear load application
direction, parallel to x-axis, taken from center of
anchor shaft to side concrete edge (in.) (see Fig. 1,
de2 is the side edge distance opposite del)
de3 = front edge distance parallel to shear load application
direction and y-axis taken from center of front
anchor shaft to front concrete edge (in.) (see Fig. 1)
de4 = back or rear edge distance parallel to shear load
application direction and y-axis, taken from center of
back anchor shaft to rear concrete edge (in.) (see Fig. 1)
E = modulus of elasticity of concrete (psi)
f = specified compressive strength of concrete (psi)
Fut(actual) = actual ultimate tensile strength of headed stud
Fut(design) = design minimum tensile strength of anchor steel
in tension (psi) (see Table 3)
f = specified ultimate tensile strength of anchor
F = shear yield strength of anchor steel (psi)
F, f
= specified yield strength of anchor steel in
tension (psi)
h = thickness of concrete member in which anchors are
embedded, measured parallel to anchor axis (in.) (see
Fig. 1)
A
NOTATION
effective headed stud embeddment depth taken as
length under head to concrete surface (in.) (see Fig. 1)
= coefficient for pryout stength (from ACT Appendix D)
L = overall length in the y-direction between outermost
anchors in conncetion = y (in.) (from AISC)
number of anchors in connection or group
n
NC,, = nominal concrete breakout strength in tension of
single anchor (Ib) (from ACT Appendix D)
Q = nominal stength of stud shear connector embedded
in solid concrete slab (lb) (from AISC)
= thickness of attachment plate (in.)
t
tf = flange thickness of a structural steel shape (in.)
= nominal concrete pryout stength (ib) (from ACT
Appendix D
nominal shear strength (ib)
V
Vs,Vsteei = nominal shear stength of single headed stud or
group of headed studs governed by steel strength
x = center-to-center spacing of stud anchors in x
direction of Cartesian plane (in.) (see Fig. 1)
i
eccentricity between shear plane and centroidal axis
of connected component (in.) (from AISC)
= center-to-center spacing of stud anchors in the y
y
direction of Cartesian plane (in.) (see Fig. 1)
A = concrete unit weight factor
= 1.0 for normal weight concrete
= 0.85 for sand lightweight concrete
= 0.75 for all lightweight concrete
= one-sided population limit (fractile) factor for a
Ic
normal distribution
= coefficient of friction
= strength reduction factor
hef
75
Influence of Flexure-Shear
Cracking on Strand
Development Length in
Prestressed Concrete Members
Robert J. Peterman, Ph.D., P.E.
Assistant Professor
Kansas State University*
Manhattan, Kansas
*rormerly Post-Doctoral Research
Associate at Purdue University
Julio A. Ramirez, Ph.D., P.E.

Professor and Assistant Head for
Graduate Studies
Purdue University
West Lafayette, Indiana
Jan Olek, Ph.D., P.E.

Associate Professor
Purdue University
West Lafayette, Indiana
76
In this study, 18 development length tests were

carried out on single strand rectangular and
multiple strand T-shaped semi-lightweight beams
having design compressive strengths of 7000 psi
(48 MPa). In the rectangular beam tests, the design
moment capacity was exceeded in all specimens.
However, in the T-beam tests, bond failures at
loads below the design capacity occurred in some
specimens immediately after the formation of a
flexure-shear crack near the loading point.
Additional T-beam tests showed that the bond
failure associated with flexure-shear cracking
could be prevented by increasing the transverse
reinforcement near the point of maximum
moment. The shift in the tension force that occurs
when flexural cracks turn diagonally needs to be
considered when determining if sufficient
anchorage of strands is provided. Therefore, the
authors recommend that the current AASHTO and
AC! requirements for strand development should
be enforced at a critical section that is located a
distance cJ from the point of maximum moment
towards the free end of the strand, where d is the
distance from the extreme compression fiber to
the centroid of the prestressed reinforcement.
PCI JOURNAL
he use of higher concrete

strengths has reduced the cost of
bridge structures by increasing
the maximum span lengths that can be
bridged using standard girder cross
sections. However, with longer span
lengths the self weight of the pre
stressed concrete sections has become
an increasingly larger portion of the
total design load for the bridge. There
fore, in order to reduce the dead load
of the concrete girder, lightweight ag
gregate is often used in the concrete
mix.
In the state of Indiana, lightweight
aggregate (consisting mostly of ex
panded shale) has been used to pro
duce semi-lightweight (SLW) pre
stressed concrete for selected projects
since 1991. The SLW concrete used in
these projects has weighed around 130
) compared to
3
lb per cu ft (2080 kg/rn
) for nor
3
145 lb per cu ft (2320 kg/rn
mal weight concrete. The SLW con
crete is obtained by partially replacing
the gravel or limestone coarse aggre
gate with a lightweight substitute.
PROBLEM STATEMENT
In a recent study sponsored by the
Federal Highway Administration
(FHWA), the applicability of the cur
rent AASHTO
2 equation for calculat
ing development lengths of prestress
ing strands in pretensioned lightweight
concrete beams was evaluated. The
unit weight of the concrete was less
) in
3
than 120 lb per cu ft (1920 kg/rn
dicating that the coarse aggregate had
been replaced in full with a
lightweight ingredient. In the FHWA
study, the current AASHTO develop
ment length equation was found to be
unconservative for estimating devel
opment lengths in girders when
lightweight concrete was used.
In view of these findings, the appli
cability of the current AASHTO de
velopment length equation for strands
in SLW concrete girders was also
questioned. Therefore, the research
presented in this paper, which was co
sponsored by the Indiana Department
of Transportation (INDOT) and the
FHWA, focused on determining the
development length of prestressed
strand when SLW concrete is used.
This was necessary in order to deter
mine the adequacy of existing struc
Strands
#4 Bars x 76 Long
(4 Total)
#3 Ties
2 Bond Break
1.
2
Clear
IIIIIIr
9 Spaces @8=72
(#3 Tie Spacings)
.;
12
r1
24
8
-
Note:
25.4 mm.
Fig. 1. Details of Moustafa pullout specimen used in this study.
tures and to provide recommendations

for the design of future projects using
SLW concrete.
All of the SLW prestressed girder
bridges built in Indiana prior to this
study had a concrete design strength
of 7000 psi (48 MPa) and are rein
forced with 1/2 in. special (13.3 mm)
diameter strand. Therefore, this study
focused primarily on the development
length determination of /2 in. special
(13.3 mm) prestressing strand in SLW
members with a similar concrete de
sign strength.
BACKGROUND
Discussed below is background in
formation on validation of prestressing
strand, assessment of strand surface
condition, and review of existing
strand transfer and development
length equations.
Strand Validation
(Moustafa Test Method)
3 in a special report published
Logan,
in the March-April 1997 PCI JOUR
NAL, concluded that there is a signifi
cant difference in bond performance in
pretensioned concrete beams among
strands produced by different strand
manufacturers. The report recom
mended that all 0.5 in. (12.7 mm) di
ameter strand used in pretensioned ap
plications be required to have a
minimum average pullout capacity of
36 kips (160 kN), with a standard de
viation of 10 percent for a six-sample
group, when embedded 18 in. (460

mm) in concrete test blocks. This test
procedure has become known as the
Moustafa method, named after Dr.
Saad Moustafa who first conducted
pullout tests on similar specimens in
4
the l970s.
The first task of this study presented
herein, therefore, involved the fabrica
tion and testing of a similar pullout
specimen to determine if the strand
used in this study would meet the min
imum average pullout capacity recom
mended by Logan. Following discus
sions with Logan, the transverse
reinforcement used in the pullout
specimens in this study was modified
from that shown in Ref. 3 to provide a
transverse tie next to each strand (see
Fig. 1). Since /2 in. special strand has
a nominal diameter of 0.522 in. (13.3
mm) instead of 0.5 in. (12.7 mm), the
corresponding minimum average pull
out capacity for the /2 in. special
strand (assuming a similar average
bond stress at ultimate load) is 37.6
kips (167 kN).
The girders for all of the semilightweight (SLW) girder bridges in
Indiana were manufactured at CSR
Hydro-Conduit, Lafayette, Indiana.
During the last ten years, Hydro-Con
duit has used strand primarily from
two suppliers. Therefore, at the outset
of this experimental program, it was
decided that the test specimens would
be fabricated using prestressing steel
from those same two suppliers. Refer
ences to strand supplied from these
77
companies will be denoted by the let

ters A and B throughout this
paper. Therefore, the pullout specimen
used to validate the strands in this
study contained nine strands from
each strand supplier.
Surface Condition Assessment

Many observers have noted differ
ences in appearance, color, and draw
ing lubricant residue of strand from
different manufacturers. Therefore, at
tempts were made to document the
initial surface condition of the strand
used in this study. Visual appearance
of the strand, in terms of color and
signs of weathering, were noted for
the strand used in the pullout and
beam specimens.
In addition, every piece of strand
used in the pullout specimen was
wiped with a white paper towel prior
to tying it into the reinforcing bar cage
to remove residue and aid in the visual
assessment of the initial surface condi
tion. This process was also performed
by Logan
3 prior to casting his pullout
specimens.
Transfer and Development Lengths

The transfer length is defined as
the distance required to transfer the
fully effective prestress force in the
strand to the concrete. The transfer
length is not a quantity specified in ei
ther the ACT Building Code
5 or the
AASHTO Specifications.
2 However,
both codes suggest a transfer length of
50 strand diameters when checking
shear provisions. The ACI Commen
tary to the Building Code (Section
12.9) provides a formula for calculat
ing the transfer length that is based on
the expression for development length.
According to this formula, the trans
fer length (Li) is given by:
=
-
Ld
=(fpsfse)
db
(2)
where f is the stress in the

prestressed strand at nominal strength
of the member (in ksi), and fse and db
are the same as in Eq. (1).
Transfer lengths affect structural de
sign considerations in two ways.
First, current code provisions for shear
design of prestressed concrete mem
bers are based on the amount of pre
compression in the member. Since the
effective prestress has been observed
to vary approximately linearly from
zero at the end of the member to be
fully effective at the end of the trans

fer zone,
6 significant deviations in the
actual transfer length from the code
suggested 50 strand diameters could
lead to inadequate estimates of the
members shear strength.
The transfer length can also have a
significant impact on the flexural be
havior of prestressed concrete mem
bers. Kaufman and Ramirez,
7 and Rus
sell and Burns,
6 have found that
anchorage failures were likely when
diagonal shear cracking propagated
through the transfer zone of a preten
sioned strand. Beams with debonded
strands are especially susceptible to
this phenomenon. Therefore, the value
of the transfer length is important in
determining whether flexural cracks
will likely propagate into this zone
prior to the member reaching its nomi
nal capacity.
In practice, development length re
quirements are typically checked,
rather than designed for. When a pre
stressed concrete member is designed,
required longitudinal reinforcement
quantities are based on service load
stresses as well as calculations of
nominal capacities. Both the ACT and
AASHTO Codes prescribe reinforce
ment ratio limits to ensure that ductil
ity is provided through ample yielding
of the prestressed reinforcement at ul
timate loads.
Thus, for flexural considerations,
the designer calculates a nominal
moment capacity of the prestressed
section by estimating a final level of
(1)
where fse is the effective stress (ksi) in

the strand after all losses, and db is the
nominal diameter of the strand in
inches.
The development length is the
bond length required to anchor the
strand as it resists external loads on
78
the member.
6 As external loads are ap
plied to a flexural member, the mem
ber resists the increased moment de
mand through increased internal
tensile and compressive forces. The
increased tension in the strand is
achieved through anchorage to the sur
rounding concrete.
In the current specifications, it is as
sumed that the development length is
equal to the length required to transfer
the effective prestress force (transfer
length) plus an additional length re
quired to develop the increase in strand
tension produced by the external load
demand. This additional length re
quired to develop the maximum stress
in the strand is often referred to as the
flexural bond length. The develop
ment length is specified by both the
ACI and AASHTO Codes as:
Fig. 2. Whittemore points mounted on transfer length specimen.

PCI JOURNAL
stress that will be achieved by the

strand (f). Based on the estimate of
5 the designer calculates a devel
f
opment length (La) by Eq. (2). A
check is then made to ensure that the
strand will have a large enough em
bedment length (Le) in the concrete
to obtain the estimated stress at
nominal capacity (J).
The embedment length is defined as
the bonded length of the prestressed
strand from the beginning of bond to
the critical section. In most design ap
plications, and in the literature, the
critical section is interpreted as the
6 ACT Sec
point of maximum moment.
tion 12.10.2 states that critical sec
tions for development of reinforce
ment in flexural members are at points
of maximum stress and at points
within the span where adjacent rein
forcement terminates... Both the ACT
and AASHTO Codes imply that if the
embedment length is greater than the
development length (Le > Ld), then the
beam should be able to reach its nomi
nal moment capacity in the absence of
shear failure.
Conversely, if the embedment
length is less than the development
length (Le < La), then bond failure
may occur prior to the beam reach
ing its nominal moment capacity and
the design is unsatisfactory. How
ever, research has shown that bond
failures may still occur when (Le >
Ld) if web shear cracking occurs and
propagates into the transfer zone.
6 recommended
Russell and Burns
design procedures which take this
into consideration when normal
weight concrete is used.
TEST PROGRAM
While considerable research has
been published on the experimental
determination of transfer and devel
opment lengths in members utilizing
normal weight concrete, with empha
sis on structural behavior and impli
cations for design, similar work for
members made of semi-lightweight
concrete is essentially absent from
the literature. Therefore, the initial
objective of this experimental pro
gram was to determine the transfer
and development length of pre
stressed strand in semi-lightweight
girders, and to assess the adequacy of
Table 1. Single-strand beam parameters.

Number of
beams
3
A_
Note: 1 in.
Embedment
length
Strand
producer
25.4 mm; 1 ft = 0.305 m; I psi
6ftl/2in.
6 ft
=
1/2
in.
Strand
size
Concrete
strength
psi
/2
ispeciaI
7000 psi
/2
in. special
0.006895 MPa.
current code provisions for the design

of such members.
Measurement of Transfer Lengths
Transfer lengths were experimen
tally determined by measuring con
crete surface strains at the ends of test
specimens. Stainless-steel points were
secured to the specimens at 2 in. (51
mm) spacings prior to detensioning
the strands. The points were mounted
using a five-minute epoxy and were
located at the depth of the strand.
Distances were measured between
points using a Whittemore gauge that
had a 10 in. (254 mm) gauge length
and had a differential reading capabil
ity of 1/i000 in. (0.00254 mm), with a
perceived accuracy of twice this
amount. Therefore, the resolution of
the gauge was about 20 ire. Surface
strain readings were taken prior to de
tensioning, immediately after deten
sioning, and periodically during the
first month after stripping.
Two specimens were fabricated
specifically for measuring transfer
lengths. These specimens had a cross
section that measured 4 x 6 in. (100 x
150 mm) and contained two concentric
strands. One of the specimens con
tained A strands while the other
specimen contained B strands. Whit
temore readings were taken on both
sides and at both ends of each speci
men, which was approximately 7 ft
10/2 in. (2400 mm) long. Fig. 2 shows
the Whittemore points mounted on one
of the transfer length specimens.
Evaluation of Development Lengths
Development lengths must be evalu
ated, rather than determined, in experi
mental programs. This is typically done
using test specimens that are loaded
such that the maximum moment occurs
at the point in the beam where the pro
vided embedment length Le is equal to
the calculated development length L.
This point is commonly referred to as

the critical section.
Development length evaluation in
this experimental program consisted of
testing six single-strand specimens and
six multiple-strand specimens. The
specimens in this investigation had
fully-bonded straight strands and were
tested by applying loads from a hy
draulic actuator that was located at a
distance Ld from the end of the speci
men. Loads were applied incrementally
until failure of the members occurred.
Interpretation of the test results is
straightforward. A flexural failure in
dicates that the embedment length is
adequate to develop the strand, while a
bond failure indicates that the embed
ment length is not sufficient and that
the actual development length is larger
than the calculated value.
Calculation of Development
Lengths for Test Specimens
The ACT and AASHTO develop
ment length equation [Eq. (2)] consid
ers the development length to be a
function of three variables, namely:
fse effective stress in strand after all
losses (ksi)
db: nominal diameter of strand in
inches
: stress in strand at nominal
5
f
strength of member (ksi)
Thus, the code-prescribed develop
ment length is not a single value that
can be evaluated for a given strand. In
stead, it is a function of both the strand
properties and the properties of the
member in which it is cast. Interest
ingly, for a given strand size and mem
ber geometry, the development length
may be calculated to be different values
by different designers, depending on
the assumptions which are made in cal
culating fse and f. From Eq. (2), it can
be seen that the calculated development
length is largest when f is maximized
andfe is minimized.
79
Test at Short End

2 Layers of Mesh (Center Region Only)
Test at Long End
1/2-Special (13.3 mm)

270 ksi (1860 MPa) Strand
5-6
(4725
Fig. 3. Cross section of single-strand

development length beams.
In other words, if the designer over

estimates f while underestimating
(i.e., by overestimating prestress
losses), then the calculated develop
ment length will be longest. While
there may be other implications on de
sign (i.e., member sizing, stress and
camber calculations, etc.), the result of
calculating excessively long develop
ment lengths (in terms of actual bond
performance) is conservative since it
will result in longer required embed
ment lengths.
However, the opposite case may not
be true. If the designer underestimates
f while at the same time overestimat
ing fse (by underestimating prestress
losses), then the calculated develop
ment length will be minimized. This
calculation will, in turn, lead to shorter
required embedment lengths and less
conservative designs (since bond fail
ure could occur if the actual develop
ment length is larger than the provided
embedment length).
Since this research was aimed at
evaluating the validity of the current
code equations for development
lengths when semi-lightweight con
crete is used, it was determined that
the worst-case scenario should be
tested. With this in mind, the experi
mental tests in this study were de
signed so that the shortest develop
ment lengths that might realistically be
calculated by designers would be
tested.
As noted above, the shortest de
velopment lengths are calculated when
80
mm)
Fig. 4. Loading arrangement for the single-strand beams.
f is minimized and fse is maximized.

The stress in the strand at nominal
strength of the member, f, is typi
cally estimated by either direct calcu
lation from code equations or by a
strain compatibility analysis. While
the strain compatibility analysis is
generally considered to be more accu
rate, especially when more than one
layer of steel is provided, the code
equations typically yield a lower esti
mate off.
Using the ACT Code,
5 the stress in
the strand at member nominal capacity
may be estimated by the equations in
Section 18.7.2 of the Code. For mem
bers with bonded prestressing tendons
and no compression reinforcement, the
formula forf reduces to:
fPs=fPu{1[P
f11
ii
(3)
where
f = specified tensile strength of
f3j =
pt,,
prestressed tendons in ksi

factor for the type of prestress
ing tendon used (= 0.28 for
low-relaxation strand)
factor used to enable ultimate
flexural capacity calculations
to be made by representing the
concrete in compression by an
equivalent rectangular stress
block
ratio of prestressed reinforce
ment = AIbd, where
is
the area of the prestressed re
inforcement in the tension

zone, b is the width of the
compression face of the mem
ber, and d is the distance from
the extreme compression fiber
to the centroid of the pre
stressed reinforcement
= specified compressive strength
of concrete in ksi
Thus, in order to test the most se
vere condition, the stress in the
strand at nominal capacity of the test
specimens was calculated using Eq.
(3). In addition, the effective pre
stress after all losses, f, was calcu
lated by assuming only 8 percent
total losses. This value was also used
by Logan
3 and is a practical mini
mum value that might be calculated
by design engineers.
Single-Strand
Development Length Specimens
Six single-strand development
length specimens were fabricated and
tested in this investigation. The pur
pose of the single-strand specimens
was to provide an economical means
to conduct multiple development
length tests with the same concrete
and strand supplier combinations.
Table 1 lists the concrete and strand
parameters of the single-strand test
beams.
The single-strand beams were used
for two development length tests each
one test per end. Shear reinforce
ment was provided only in the central
PCI JOURNAL
portion of the beams to ensure that this

region would be intact after the first
end of the beam was tested. The cen
tral portion of the beam was part of
the loaded span for the testing of both
ends of the single-strand beams.
The shear reinforcement in the cen
ter region of the beams was purpose
fully not centered in the beam. This
was done to provide the ability to in
crease the embedment length at one
end of the test specimens in the event
that initial beam tests based on the cal
culated development lengths would
experience bond failure.
The nomenclature used for the sin
gle-strand specimens is the following:
[Concrete Strength & Type]- [Strand
Typej-[Beam # within Seriesj[Test
End]
Thus, the name 7SLW-A-2L
would refer to a test specimen utiliz
ing 7000 psi Semi-LightWeight con
crete that had a single /2 in. special
(13.3 mm) strand produced by Manu
facturer A, and was the test at the
Long end of specimen number
(of 3). The long end refers to the
beam end with the greatest distance to
the shear reinforcement located in the
central region.
Although all of the development
length beams in this study contained
semi-lightweight concrete, the study
was carried out concurrently with an
other study that utilized normal
weight concrete. Therefore, the term
SLW was (unnecessarily) used in
the naming of all specimens in this
study. All of the test data presented in
this paper are from prestressed beams
with a design concrete strength of
7000 psi (48 MPa).
The single-strand specimens had a rect
angular cross section measuring 8 x 12 in.
(200 x 305 mm), and contained a single
prestressing strand located at a depth d
of 10 in. (255 mm) (see Fig. 3). The
width of 8 in. (200 mm) was slightly
larger than the 612 in. (165 mm) width
used by Logan
3 for his single-strand
specimens, in order to minimize the
depth of concrete in compression and
maximize the strain in the prestressed
strand at the ultimate flexural capacity
of the specimens. The increased speci
men width was needed because /2 in.
special (13.3 mm) diameter strands
were tested in this study, whereas
Fig. 5.
Test setup for
single-strand
beams.
Logan tested specimens containing /2

in. (12.7 nmi) diameter strands.
Using the 8 in. (200 mm) width, the
strain in the strands at nominal flexural
capacity of the test beams in Table 1
was estimated at 2.7 percent based on a
strain compatibility analysis. Although
this value was lower than the 3.5 per
cent value recommended by Buckner
9
for minimum strand strains in develop
ment length specimens, it was larger
than the 2.0 percent value calculated
by Logan for his single-strand beams
that failed by strand rupture.
Fig. 4 shows the loading arrange
ments corresponding to the 7000 psi
(48 MPa) singlstrand rectangular
beams. Loads were incrementally ap
plied to the beams using a hydraulic
ram powered by an electronically con
trolled power unit. Values of load, de
flection at the applied load, and strand
slip at the beam end were recorded
throughout the entire loading sequence
of all 12 tests (two tests per beam for
six beams). Fig. 5 shows the test setup
used for the single-strand beam tests.
Multiple-Strand
Development Length Specimens
The purpose of the single-strand
specimens in this investigation was to
provide an inexpensive means to con
duct numerous development length
tests on beams having the same con
crete and strand supplier combina
tion. To study the effect of multiple
strands (at close spacing) on develop
ment length, full-scale specimens
containing multiple strands were
tested in addition to the six singlestrand specimens. These specimens
were designed based on the analysis
of test data from the single-strand
rectangular specimens.
Results from tests on single-strand
specimens indicated that the current
ACT and AASHTO equations were
appropriate for use with semilightweight concrete. Therefore, the
first group of three multiple strand
specimens was designed with an em
bedment length based on the current
code provisions. These specimens
each contained five bottom strands
81
and had a T-shaped cross section with

an overall depth of 21 in. (535 mm)
and a compression flange width of 36
in. (915 mm) (see Fig. 6).
Two of the three T-beam specimens
contained 1/2 in. special (13.3 mm) di
ameter A strands while the third T
beam had a similar strand pattern but

utilized B strands. All three T
beams had a design compressive
strength of 7000 psi (48 MPa). This
initial group of three specimens were
cast at the same time and in the same
prestressing bed (see Fig. 7).
(1) 1/2-Special (13.3 mm) strand to control

stresses @ release. This strand was bonded only
ember ends and was cut prior to testing.
Fig. 6. Cross section of multi-strand T-beams.
Fig. 7. The T-beams were cast end-to-end so that the strand jacking force for each
beam was identical.

82
This was enabled by splicing all five

strands in the span between the bulk
heads of the beams using A and B
strands (see Fig. 8). Splicing of the
strands ensured that all beams had the
same initial tension in the strands.
These beams are referred to as T
Beam Al, T-Beam A2, and T-Beam B
(see Table 2).
Unlike the single-strand specimens,
which were each tested at both ends,
the multiple-strand T-beams were de
signed with a length that was approxi
mately equal to twice the calculated
development length, so that a point
load applied at midspan would effec
tively test the anchorage at both ends
simultaneously. The actual length of
the T-beams was 6 in. (150 mm)
longer than twice the calculated devel
opment length, because the load was
applied to the beam through a 6 in.
(150 mm) wide steel plate.
This beam length increase ensured
that the length of embedment of the
strand (from the free end of the beam to
the edge of the loading plate) coincided
with the development length calculated
based on the principles discussed earlier
in this paper. Fig. 9 shows the dimen
sions and loading arrangement for the
multi-strand T-beams.
Design of shear reinforcement using
both the AASHTO and ACI Code pro
visions showed that #4 (13 mm) stir
rups at 15 in. (305 mm) spacings
would provide sufficient shear rein
forcement for all T-beams in this
study. However, the transverse rein
forcement provided in T-beams Al,
A2, and B was #4 (13 mm) stirrups at
6 in. (150 mm) spacings, or more than
twice the code-required amount. Fig.
10 shows the vertical stirrup spacing
for the first group of three T-beams.
The T-beams were tested in the Ket
telhut Structural Engineering Labora
tory at Purdue University. Loads were
incrementally applied to the beams
through a 6 in. wide x 24 in. long (150
x 610 mm) steel plate using a 220 kip
(978 kN) capacity MTS hydraulic ac
tuator. Values of load, midspan deflec
tion, and strand slip for all five strands
at both ends of the beam were
recorded during the testing of each T
beam. Figs. 11 and 12 show the test
setup and strand slip measuring device
used for the multi-strand T-beams.
PCI JOURNAL
Importance of Stirrup Spacing on

Longitudinal Steel Stress
The results of the development
length tests will be presented later in
this paper. However, some of the find
ings are mentioned at this point be
cause they affected the fabrication of
the second group of three T-beam
specimens.
T-Beam B experienced bond failure
prior to reaching the nominal moment
capacity. A careful review of a video
tape of this load test (in slow motion)
showed that an inclined flexure-shear
crack occurred immediately prior to
strand slip and subsequent web-shear
cracking. It has been observed that
the initiation of inclined cracks in sim
ply supported beams will cause an in
crease in the tension demand closer to
the support.
The ACT Code accounts for this ten
sion force shift in flexural members
with non-prestressed reinforcement
(ACI 12.10.3) by requiring that longi
tudinal bars in tension be extended for
a distance equal to the effective depth
of the member beyond the point where
they are required to resist flexure.
The behavior of T-Beam B sug
gested that this shift in tension de
mand in the prestressed reinforcement
resulted in a bond failure. This ob
served behavior also suggests that the
critical section for prestressed con
crete members referred to in ACI
12.9.1 may not correspond to the loca
tion of the maximum moment, but
rather a section at some distance
closer to the free end of the strand.
This hypothesis is discussed in the fol
lowing paragraphs.
Fig. 13 shows an idealized bilinear
representation of the stress capacity in
bonded prestressed tendons versus the
distance from the free end of the
strand. For the T-beam specimens, de
velopment length tests were conducted
so that the maximum stress in the
strands (f) was produced at a dis
tance from the end of the beam that
was equal to the ACI prescribed de
velopment length Ld.
In Fig. 13, the stress in the pre
stressed concrete strand must lie
below the bilinear curve at all loca
tions in order for bond requirements to
be satisfied. Note that at a distance (x)
closer to the free end from the devel
Fig. 8.
Splicing of A and
B strands between
T-Beams.
Table 2. Multi-strand T-beam parameters.

Strand
producer
Al
6 ft
11/2 in
7000 psi
/2
6ftl/2in.
7000psi
V2in.special
6 ft
7000
A2
B
Note: I in.
25.4 mm; 1 ft = 0.305 m; 1 psi
Embedment
length
Strand
size
T-beam
1/2 in.
Concrete
strength
in special
/2
in. special
0.006895 MPa.
Fig. 9. Loading arrangement for multi-strand T-beams.
12-6 (3800 mm) Longitudinal Bar
111111111 liii 1111111 liii III

3
(75 mm)
jj
24 Spaces @6(152 mzn)= 12-0 (3650 mm)
(75 mm)
Fig. 10. Spacing of transverse reinforcement for T-Beams Al, A2, and B.
opment length (Ld), the maximum per

missible stress in the strand is equal to
-4f).
2
(f
Fig. 14 (from MacGregor
) shows
10
the internal forces in a cracked beam
without stirrups. For wide cracks, the
aggregate interlocking force V
0 disap
pears, along with Vd, V, and Ci, and

2 = T
T
. In other words, the inclined
1
crack has made the tensile force at
Point C a function of the moment at
Section A-B-D-E, and Point C can
now be assumed to be the critical
section.
83
Fig. 11. Test setup for development length evaluation of

T-beams.
For a beam without stirrups, having

its nominal moment capacity (and cor
responding stress f) demanded at a
distance (Ld) from the free end of the
strand, the onset of flexure-shear
cracking will produce a strand stress
(at the crack location) that lies above
the bilinear curve in Fig. 13. If the
ACT expression for strand develop
ment length represents the actual
length required to develop the strand
stress (J) corresponding to nominal
capacity for the beam, then inclined
cracking may lead to bond failure.
Fig. 15 (from MacGregor
) shows
10
the internal forces in a cracked beam
with stirrups. In this case, the presence
of stirrups will ensure that there will
always be a compression force C] and
a shear force V acting on the part of
the beam below the crack, and there
fore T
2 will be less than T
. However,
1
even though the tension force at Point
C will be less than the tension at Sec
tion A-B-D-E, the strand stress may
still lie above the bilinear curve in Fig.
13 and failure by bond can still occur.
The ACT Commentary to Section
12.10.3 notes that A diagonal tension
crack in a flexural member without
stirrups may shift the location of the
calculated tensile stress approximately
a distance d towards a point of zero
moment. When stirrups are provided,
this effect is less severe, although still
present to some extent. Thus, if a
flexure-shear crack forms, some de
gree of shift in the location of maxi
mum tensile stress will occur.
Fig. 13 illustrates that a change in
strand stress of (4/) equal to (xldb) ksi
84
Fig. 12. LVDTs were used to measure the strand slip at both
ends of the I-beams.
must occur over the distance x from

the point of maximum moment so that
the strand stress (at distance x from the
maximum moment) will not lie above
the bilinear curve. Therefore, when an
inclined crack occurs, extending from
the point of maximum moment to a
point x closer to the free end of the
strand, a reduction in strand stress of
(x/db) ksi must occur over the horizon
tal projection of the crack to preclude
bond failure. This can be accom

plished by providing extra transverse
reinforcement across the crack.
The amount of required transverse
reinforcement to cause the appropriate
reduction in strand stress may be esti
mated using the model in Fig. 16. This
model, which assumes the inclined
crack can be represented by a linear
crack, is the basis for the calculation
of required transverse reinforcement
I;
(fp
From Figure,
Slope
f,)d,
f, f,Af
ff
(fp,f,.)db
Therefore,
4f
(SIope)x)
x
=
Fig. 13. Bilinear variation of steel stress with distance from free end of strand.
PCI JOURNAL
D
Vay
Va
Vdt7
= = = -=. = =
2
T
E
Fig. 15. Internal forces in a cracked beam with stirrups.
Fig. 14. Internal forces in a cracked beam without stirrups.
2
r
v,c.z
=
jd
Rieht Free-Body Diagram
T, =7; AT
Left Free-Body Diaeram
Model Assumptions:
The flexure-shear crack can be represented by a linear crack as shown.

Dowel action is conservatively ignored.
The weight of the beam is negligible.
The line of action of the sum of all aggregate-interlock forces (J,) passes approximately through
point 0. Therefore, the moment due to this force about point 0 is small and can be ignored.
The force in the stirrups crossing the crack can be determined by summing moments about point 0.
Solving...
M,,=0
2A,,jd
Assuming all stirrups crossing the crack are yielding, then the force in the stirrups (Vs) is equal to the total
area of the stirrups (A.,) multiplied the yield stress (f;.,).
Therefore,
Afrv
= 2A,jd
_
For the 7000 psi (48 MPa) T-beams...
2AJd
A,,., = (5)(0.167) = 0.835 sq in.

a = /Jjc = 0.987 in. --from flexural analysis of section
jd=d-a/2 19(0.987)/2 18.51 in.
f., = 60 ksi
ci,, = 0.522 in. for 1/2-Special strand
--
A,
2(0.835)(l 8.5 1)
=0.99 sq in.(640
(60)(0.522)
sq mm)
/stirrup), 3 stirrups are required to cross the crack.

2
Therefore, using #4 (13mm) stirrups (A = 0.40 in
Fig. 16. Calculation of transverse reinforcement required to reduce the tension force across an inclined crack by the amount T.
85
plained in the previous paragraphs,

namely, that increased stirrup spacing
can reduce the tension shift that sup
posedly resulted in bond failure of T
Beam B.
The third additional T-beam had #4
(13 mm) stirrups at 15 in. (375 mm)
on center in the middle portion of the
beam. This corresponded to the ACI
Code-required amount of transverse
reinforcement for shear. The addi
tional three T-beams were cast end-toend, as in Fig. 7, which ensured that
they each had identical levels of pre
stress.
The reinforcement in the flanges of
the additional T-beams was similar to
that in the original three T-beams.
Table 3 lists the parameters of the ad
ditional three T-beams.
--
III I liii 11111111111111111 U

24 Spaces
3
(75
4.I
mm)
@ 6 (152 mm) =
12-0 (3650 mm)
(75
mm)
Vertical Stirrup Spacing For T-Beam B-6
111111 IIIIIIIIIIIIIIIIIIIIIIIIIIIII 111111

5 Spaces
3
(75 mm)
28 Spaces
@ 3 (75
5 Spaces
mm)
6 (155 mm)
6 (155 mm)
3
(75 mm)
1111111111
1
1114
4
Spa.
6 Spaces
TEST RESULTS
1111111
15 (375 mm)
Spa.
@111
(155mm)
(155mm)
3 (75 mm)
3 (75 mm)

Fig. 17. Vertical stirrup
Second Group of
Additional Three T-Beams
Based on the observed behavior
(flexure-shear cracking followed by
bond failure) of T-Beam B, three ad
ditional 7 ksi (48 MPa) T-beam speci
mens were fabricated and tested in
this experimental program. These T
beams each contained strand from
Producer B and were identical in
length, cross section, and longitudinal
reinforcement to the other three T
beams presented in this paper. How
ever, the three additional specimens
86
Pullout Test Results
detailing for the three additional T-beams.
(also in Fig. 16). This calculation indi

cates that, for T-Beam B, at least three
#4 (13 mm) stirrups would have
needed to cross the diagonal crack in
order to cause the appropriate reduc
tion in strand stress and preclude pos
sible bond failure. A post-failure in
spection of T-Beam B indicated that
only one (or perhaps two) stirrup
crossed the inclined crack that pre
ceded failure.
Presented below is a summary of

the strand pullout, strand surface con
dition, strand transfer length, and sin
gle-strand and multi-strand develop
ment length test results.
differed from the original three T

beams in the amount of transverse re
inforcement near midspan. Fig. 17
shows the varying amounts of trans
verse reinforcement used in the addi
tional three T-beams.
One T-beam had #4 (13 mm) stir
rups at 6 in. (152 mm) on center
throughout its entire length. The stir
rup spacing in this beam was identical
to the spacing in T-Beam B (that
failed by bond). A second T-beam had
#4 (13 mm) stirrups at 3 in. (75 mm)
on center in the middle portion and
was used to test the hypothesis ex
The pullout specimen was cast and

tested at two days and at four days. The
strands were pulled out of the block
using a hydraulic ram furnished by
CSR Hydro-Conduit. The load was
recorded using a load cell that was
placed between two steel plates. The
lower plate had two steel angles welded
to the bottom side to allow the same
contact area with the concrete block
that was specified by Logan.
3 Fig. 18
shows the load cell configuration used
to detennine the pullout values.
The maximum load occurring during
a given pullout test was stored auto
matically by a data acquisition system.
Also, the load at first slip was ob
Table 3. Parameters of additional three T-beams.

T-beam
name
Strand
producer
Embedment
length
Concrete
strength
B-3
6 ft 1V2 in.
7000 psi
B-6
6 ft 1/z in.
7000 psi
f in. special
B-15
6 ft 1/2 in.
7000 psi
/2
Note: 1 in.
25.4 mm; 1 ft
0.305 m; 1 psi
Strand
size
1/2
in. special
in. special
0.006895 MPa.
PCI JOURNAL
Fig. 18.
Load cell arrangement used to
measure pullout force.
tamed by placing a small piece of tape

on the strand at the point where it en
tered the concrete. As soon as a motion
of the tape was detected, the corre
sponding value of load was recorded.
Tables 4 and 5 show the pullout
loads and first slip values, while
Fig. 19 shows the positioning of the
strands in the pullout specimen. As in
dicated, the average pullout capacity
for each strand type tested exceeded
the minimum value of 37.6 kips (167
kN), indicating that the bond quality
of both strands was acceptable. Tables
4 and 5 show that the pullout and firstslip values were slightly higher at four
days than at two days, as expected,
due to the higher concrete strength and
presumed higher modulus of elasticity.
These tables also show that, although
both strands had similar ultimate pull
out capacities, the B strand began to
slip at a significantly lower load.
the strands. The towels from the B

strands have considerably more no
ticeable residue than those corre
sponding to the A strands.
In performing the towel wipes, it
was noted that it was easier to wipe
Table 4. Results from pullout tests at two days (concrete compressive

strength was 4800 psi).
From the beginning of this study it

was noted by several observers that
the two 1/2 in. special (13.3 mm)
strands used had a markedly different
appearance. The A strand had a
bluish hue about it that might best be
described as a gun-steel blue color.
The B strand, on the other hand, had
a yellow/brassy tint.
Prior to casting the pullout speci
mens, the strand samples were wiped
with a paper towel to remove residue
and help assist in the visual assess
ment. Fig. 20 shows the towels used to
wipe the strands for the pullout speci
men. This figure shows there is a dis
tinct difference between the residue on
Strand B
Strand A
---.-...--.--
-.------
Strand
designation
Max. pullout
force (Ib)
Load at
first slip (Ib)
Strand
designation
A-9
36,450
31,300
A-8
37,300
A-7
Load at
Max. pullout
force (Ib)
first slip (Ib)
B-9
38,450
20,800
29,500
B-8
38,000
16,200
41,000
32,000
B-7
38,800
15,100
A-6
41,000
36,000
B-6
37,800
19,100
Average
38,940
32,200
Average
38,260
17,800
Std. Dev.
2,080
2,380
Std. Dev.
390
2,270
Note: 1 lb
Surface Condition
Assessment Results
the B strands than the A strands,

as there was much more of a ten
dency to bind or tear the towels on
the A strands when applying equal
pressure. In general, the residue cor
responding to the A strands was
brown or rust colored while the
residue corresponding to the B
strands was black. The chemical
composition of this residue was not
determined.
It should be noted that all towel
wipes in this study were performed
by the same person. Also, the two
strand rolls in this study (one roll of
A strand and one roll of B strand)
were placed indoors when they were
received from the producers in an at
tempt to minimize weathering during
the course of the study. Both strand
rolls were received at the precasting
plant during the summer of 1997.
4.448 N; 1 psi
0.006895 MPa.
Table 5. Results from pullout tests at four days (concrete compressive

strength was 5200 psi).
Strand B
Strand A
Strand
designation
---.---
Max. pullout
force (Ib)
Load at
first slip (ib)
Strand
designation
Max. pullout
force (Ib)
Load at
first slip (lb)
-.---
A-S
39,350
33,800
B-S
42,700
20,900
A-4
41,450
31,600
B-4
38.000
19,200
A-3
43,200
36,700
B-3
40,550
23,200
A-2
42,450
41,600
B-2
38.550
22,000
A-i
41,100
37.200
B-I
37,450
Average
41,510
36,180
Average
39.450
21,640
Std. Dev.
1.3 10
3,390
Std. Dev.
1,933
1,460
Note: 1 lb
4.448 N; 1 psi
22,900
0.006895 MPa.
87
Transfer Length
Measurement Results
A-I
B-2
A-3
B-4
A-5
8-6
A-7
B-8
A-9
Values of surface strains were mea

sured on specimens similar to those in
Fig. 2. The transfer length can then be
inferred directly from the recorded
values of strain. For a concentrically
B-I
A-2
B-3
A-4
B-5
A-6
B-7
A-S
B-9
prestressed member supported along

the entire length, such as the one in
Fig. 2, the member will be under sim
Fig. 19. Positioning of strands in pullout specimen.
pie axial loading when the prestress

force is transferred to the concrete.
Thus, the compressive stresses (and
hence strains) in the member will vary
from zero at the free end to a nearconstant value corresponding to the
P/A
stress. The length it takes for
the strand to bond to the concrete and

transfer the full tension in the ten
don to the concrete is called the trans
fer length.
Fig. 21 shows the results of surface
strain measurements taken for the
7000 psi (48 MPa) SLW transferlength specimen with two
/2
in. spe
cial (13.3 mm) B strands. Results

from the specimen containing A
strands were similar. Measurements
were taken immediately after transfer
Fig. 20. Towels used to wipe strands before placing them in pullout specimen.
20
40
of prestress, and at 3, 14, 36, and 66
60
100
80
0
I
I
I
-200
-400
At Transfer of Prestress
3 Days After Transfer
*-
ysAfterTransfer,
-600
-800
-1000
-1200
...._
-1400
-1600
-1800
Distance Along Beam (in.)
Fig. 21. Surface strains for

88
/2
in-Special (13.3 mm) B strand in transfer-length specimens. Note: 1 in.
25.4 mm.
PCI JOURNAL
days thereafter. The vertical dashed

lines are drawn at the approximate
breaks between the sloping portion
of the curves where strains are in
creasing and at the flat portion cor
responding to the region under con
stant stress of P/A. The distance from
the end of the transfer length specimen
to the dashed line is the transfer
length.
Results from both transfer length
specimens indicated that the measured
transfer lengths were less than 50
strand diameters (the amount sug
gested by both AASHTO and ACT
when checking shear provisions) ex
cept in the case of one end where
splitting of the concrete was noted. In
this case, the measured transfer length
was close to 70 strand diameters.
Single-Strand
Development Length
Test Results
A total of twelve single-strand de
velopment length tests were conducted
in this study (six beams tested at both
ends). As discussed previously, the
single-strand beam specimens pro
vided a cost effective means of con
ducting several load tests. The failure
loads and deflections corresponding to
the maximum sustained load are
shown in Table 6.
Table 6 shows the test results for the
single-strand rectangular beams con
taining A and B strands. This
table lists the maximum moment in
the beams, occurring at a distance Ld
from the end of the beam and deter
mined from the measured values of
applied load. In every case, the maxi
mum moment exceeded the AASHTO
nominal moment capacity (Ma), indi
cating the beams strands were ade
quately developed at the point of max
imum moment (which occurred at a
distance equal to the code-prescribed
development length Ld from the beam
end).
This finding is consistent with the
results from measurements of strand
end-slip during testing, which revealed
that slip did not occur in all but one
specimen, namely, 7SLW-A-1S. In
this specimen. a strand slip of 0.051
in. (1.3 mm) was recorded on the data
scan prior to failure. However, this
minimal slip occurred after the nomi
Table 6. Results from the single-strand rectangular beam tests.

Beam
Max. load
(Ib)
Max. moment*
(kip-ft)
Deflection at max.
load (in.)
Failure mode
7SLW-A-1 S
10,740
40.0
2.1
Shear
7SLW-A-IL
10,370
39.8
3.1
Shear
7SLW-A-2S
10,600
-
39.5
---
2.7
_ff_____________
Flexure
-
Strand rupture
7SLWA-2L
10,000
38.5
>3.0
Flexure
7SLW-A-3S
10,540
39.2
2.6
Flexure
7SLW-A-3L
10,350
39.8
2.4
Flexure
7SLW-B- IS
11,000
40.9
2.2
Flexure
7SLW-B-IL
10,680
40.9
>3.0
Flexure
7SLW-B-2S
10,980
40.8
1.7
Flexure
7SLW-B-2L
10,530
40.4
2.6
7SLW-B-3S
11,180
41.5
2.0
Flexure
7SLW-B-3L
10,350
39.7
2.9
Flexure
Strand rupture
Strand rupture
Strand rupture
Strand rupture
Strand rupture
Strand rupture
Shear, then Strand rupture
Strand rupture
Strand rupture
Note: 1 lb=4.448 N; I kip-ft= 1.357 kN-m; 1 in. = 25.4 mm.

*AASHTO Nominal Moment capacity (M,) = 34.6 kip-ft (J = 260 ksi).
nal moment capacity had been ex

ceeded by over 10 percent.
Table 6 also indicates that all fail
ures occurred after considerable de
flections had occurred. Each of the
7000 psi (48 MPa) beams deflected
more than l/2 in. (40 mm) in a 15 ft
3 in. (4.65 m) span prior to reaching
its ultimate capacity. Nine of the
twelve beams failed in flexure when
the strands ruptured. The other three
specimens failed in shear.
While these shear failures occurred
when the shear stress on the section
(V/bd) was less than 92 psi (631
they occurred well
MPa), or 1.1
after yielding of the prestressing steel
had occurred and after considerable
ductility had been exhibited. It is pos
sible that when the prestressing steel
yielded, the effects of dowel action di
minished and the lightly reinforced
beams (without stirrups) became sus
ceptible to shear.
Tn summary, test results from the
twelve single-strand development
length specimens indicated that the
code-required development lengths
were sufficient to develop the capacity
of a single prestressed strand in a
member cast with 7000 psi (48 MPa)
semi-lightweight (SLW) concrete.
Therefore, as previously mentioned,
the T-beam specimens containing

multiple strands were also designed
using the ACT and AASHTO develop
ment length expressions.
Development Length
Test Results on Multi-Strand
T-Beams
Table 7 summarizes the test data, in
cluding the failure load, deflection
corresponding to the maximum sus
tained load, and maximum moment
for each T-beam. The first group of
three T-beam specimens cast, namely,
T-Beams Al, A2 and B were identical
in detail. Table 7 shows that both T
beams containing A strands (Al and
A2) exceeded the AASHTO nominal
moment capacity for the section and
failed by strand rupture. Thus, the
AASHTO and ACT development
length expressions were conservative
for these beams.
Both of these failures were ductile,
/t
as midspan deflections exceeded 1 3
in. (36 mm) prior to the ultimate ca
pacity being attained. T-Beam B had
essentially the same ultimate capacity
as T-Beam Al of 120.5 kips (536 kN).
Interestingly, though, this load corre
sponded to a much lower deflection
for the B beam [0.82 in. versus 2.02
89
Table 7. Results from the multi-strand T-beam tests.

T-beam
Maximum
load (kips)
Maximum
moment*
Deflection at
max. load (in.)
Al
120.5
362.9
2.02
Flexure
A2
124.0
373.1
1.42
Flexure
120.5
362.9
0.82
B-3
129.9
390.5
1.39
Flexure
B-6
110.0
332.0
0.31
Bond/Web shear failure
302.6
0.24
Bond/Web shear failure
B-iS
Failure mode
Strand rupture
Strand rupture
Bond
Strand rupture
Note: 1 lb =4.448 N; I kip-ft= 1.357 kN-m; 1 in. =25.4 mm.

*AASHTO Nominal Moment Capacity (LVI,,) = 340.3 kip-ft.
in. (20 mm versus 51 mm)] and it

failed by bond, as the strands slipped
with respect to the surrounding con
crete and pulled in from the end of the
beam (see Fig. 22).
After reviewing videotape of the
failure of T-Beam B, the investigators
Fig. 22.
Strands pulled in
from the ends of
T-Beam B.
Fig. 23.
Flexure-shear
cracking, and
subsequent splitting
in T-Beam B.
90
noted that a flexure-shear crack de

veloped just prior to collapse (see
Fig. 23). This observation led to the
hypothesis that the flexure-shear
crack caused an increase in tension
nearer the beam end and effectively
shifted the critical section from the
section at the point load to the place

where the flexure-shear crack inter
sected the strand. This was discussed
earlier in this paper.
It is important to note that the crack
ing that occurred in T-Beam B prior to
failure was not dissimilar to the crack
patterns that developed in T-Beams
Al and A2, yet significantly different
modes of failure occurred. It is likely
(based partly on pullout specimen be
havior and towel-wipe tests) that the
bond quality of the A strand may
cause it to develop over a shorter dis
tance than the equivalent-sized B
strand.
If this were true, then a flexure
shear crack which shifts the tension
demand closer to the support may not
be critical in the case where A
strand was used. In general, this ten
sion shift would only lead to sudden
collapse upon cracking if the actual
distance required to develop a strand
lies between the point of maximum
moment and the point where the diag
onal crack intersects the strand.
In order to test the hypothesis out
lined above, three additional 7000 psi
(48 MPa) T-beam specimens utilizing
1/2 in. special (13.3 mm) B strand
were fabricated and tested, each hav
ing the same dimensions and test con
figuration as the original three T
beams. Table 7 shows the results from
the additional three 7000 psi (48 MPa)
T-beam tests.
T-Beam B-6 was tested first. This
beam had a constant 6 in. (152 mm)
stirrup spacing which was identical to
that in T-Beam B. As Table 7 indi
cates, T-Beam B-6 failed by
bond/shear at a load of 110.0 kips
(489.2 kN), corresponding to 97.6 per
cent of the AASHTO nominal moment
capacity for the section. Strand slip
data showed that all strands had small
values of slip prior to reaching the
maximum sustained load.
At the time of failure, the load was
held constant and crack patterns were
being recorded. Therefore, it is likely
that additional slip of the strands oc
curred during the time period when
the load was held constant (but endslip readings were not continuously
recorded). While it cannot be proven,
it is plausible that additional slip of the
strands resulted in a reduced prestress
PCI JOURNAL
force and, therefore, a loss in shear ca

pacity. Fig. 24 shows the failure
cracks for T-Beam B-6.
T-Beam B-3 was the next beam
tested. This beam had #4 (13 mm)
stirrups at 3 in. (75 mm) on center in
the middle portion of the beam. Table
7 shows that this beam failed by
strand rupture at a load of 129.9 kips
(577.7 kN). This load corresponded to
a maximum moment in the beam that
was 14.7 percent larger than the
AASHTO nominal moment capacity
for the section.
Review of strand-slip data for this
beam showed that slip was essentially
zero at the time of failure. Nine of the
ten ends measured had a recorded slip
at failure that was less than 0.001 in.
(0.03 mm). The other strand had a
recorded slip of 0.046 in. (0.12 mm).
T-Beam B-3 had the same strand
and concrete batch used in T-Beam
B-6, which experienced bond failure
at a load of only 110.0 kips (489.2
kN). In other words, with stirrups
spaced at 3 in. (75 mm) on center, T
Beam B-3 was able to withstand an
applied load that was 18.1 percent
larger than the failure load for T
Beam B-6. Fig. 25 shows the failure
crack and corresponding strand rup
ture for T-Beam B-3.
T-Beam B-15 was the last beam
tested in the series. This beam had a
stirrup spacing in the central region
of the beam of 15 in. (375 mm),
which corresponded to the AASHTO
and ACT Code minimum amount re
quired for shear. As expected, this
beam experienced bond/shear failure
at only 100.0 kips (444.9 kN), the
lowest load for all the 7000 psi (48
MPa) T-beam specimens tested (see
Table 7). Fig. 26 shows the shear fail
ure that occurred after strand slip ini
tiated in the member.
In summary, both of the 7000 psi
(48 MPa) T-beams containing A
strand (Al and A2) experienced flexu
ral failures (by strand rupture). Each
had #4 (13 mm) stirrups at 6 in. (152
mm) throughout the entire length of
the beam. Three of the four T-beams
utilizing B strand (B, B-6, and B15) experienced bond failure and, in
two cases (believed subsequent), shear
failure when loaded at a distance from
the end of the beam equal to the
Fig. 24.
Failure of
T-Beam B-6.
Fig. 25.
Failure
occurred by
strand rupture
in T-Beam
B-3.
Fig. 26. Failure of T-Beam B-i 5.

91

lengths.
These failures occurred suddenly,
and without much warning, at signifi
cantly smaller deflections (see Table
7). Flexural failure (by strand rupture)
was achieved in a T-beam using B
strand when #4 (13 mm) stirrups were
provided at 3 in. (75 mm) centers in
the middle portion of the beam. This
spacing provided a stirrup area that
was five times greater than the amount
required by the AASHTO and ACT
shear provisions.
MAJOR FINDINGS
This study has yielded credible evidence that there is an interaction between the shear carried by a prestressed concrete member near the
point of maximum moment and the
length required to sufficiently anchor
the longitudinal reinforcement. Although the findings of this study were
made in the context of tests on members with semi-lightweight concrete,
which typically have a lower modulus
of rupture and would thus be more
susceptible to flexure-shear cracking,
the principles discussed herein should
also be applicable for members cast
with normal weight concrete.
Measurements of concrete surface
strains indicated that the transfer
lengths associated with both A and
B strand in 7000 psi (48 MPa) SLW
concrete were less than the AASHTO
and ACT Codes assumed 50 strand diameters in the absence of longitudinal
splitting at transfer. In the end where
splitting occurred, the measured transfer length was close to 70 strand diameters. These measurements also
showed that the transfer lengths remained essentially unchanged during
the first 60 days following transfer of
prestress.
Tests on single-stranded rectangular
beams and multiple-stranded T-beams
revealed that the length required to develop the tensile capacity of a strand
in concrete is, in some cases, dependent on the member geometry and
loading configuration. Strand rupture
(associated with flexural failure) occurred for both A and B strands
when cast in single-strand rectangular
specimens containing 7000 psi (48
MPa) concrete.
92
When the same combinations of

strand and concrete were tested in the
multi-stranded T-beams, however, the
results were mixed. In particular, the
combination of 1/2 in. special (13.3
mm) B strands and a 7000 psi (48
MPa) concrete mix resulted in bond
failures for three of the four T-beam
specimens tested. Two other T-beam
specimens, which had the same con
crete mix and the same diameter A
strands, resulted in flexural failures by
strand rupture.
The review of the failure of T-beam
B in this study showed that the bond
failure was preceded by a flexure
shear crack. It is postulated that the
onset of cracking resulted in a shift of
the maximum tensile stress (i.e., the
critical section) in the strand from the
point of maximum moment towards
the free end of the strand.
Although similar crack patterns
were noted for the T-beams containing
the 1/2 in. special (13.3 mm) A
strands, that failed by strand rupture, it
is surmised that the actual development length for the A strand may
have been considerably less than the
code value that was tested. In this
case, then, a shift in the critical sec
tion would still result in an embedment length (to the critical section)
that is larger than the actual development length for the strand, and col
lapse would not occur.
On the other hand, if the actual de
velopment length of the B strand
was close to the code-determined
value that was tested, then a shift in
the critical section could lead to col
lapse. The above hypothesis is consis
tent with the earlier slip measured for
the B strand in the Moustafa pullout
specimen tests.
To test this theory, three additional
T-beam specimens, containing the
same concrete mix and strand combi
nation as T-Beam B, were fabricated
and tested. Each of these additional T
beams had different amounts of trans
verse reinforcement near midspan. For
the T-beam with the largest amount of
transverse reinforcement at the point
of maximum moment, namely, T
Beam B-3, bond failure was pre
vented (presumably by minimizing the
shift in the location of the critical sec
tion) and the mode of failure was flex-
ure by strand rupture.

The amount of transverse reinforce
ment required to prevent bond failure
represents between 2.5 to 5 times the
amount required by shear design
(Note: T-Beam B-6 failed by bond
while T-Beam B-3 failed by strand
rupture). While such increases in the
amount of transverse reinforcement
are not practical for most design situa
tions, the designer must consider that
the critical section may shift in the
event of inclined cracking.
CONCWSIONS
Based on the work carried out in
this investigation, the following con
clusions can be drawn:
1. Both of the 1/2 in. special (13.3
mm) strands used in this study met
the requirement for minimum aver
age pullout force (37.6 kips) accord
ing to the Moustafa procedure.
Therefore, according to the Moustafa
test, both strands had acceptable
bond characteristics.
2. Twelve load tests on rectangular
single-strand beams indicated that the
lengths of:
Ld
(f
fse)
db
provided sufficient embedment to

develop the full capacity of a single
strand in the SLW concrete mixes
used.
3. The onset of flexure-shear crack
ing resulted in a shift of the maximum
tensile stress (i.e., the critical section)
in the strand from the point of maxi
mum moment towards the free end of
the strand.
RECOMMENDATIONS
The following recommendations are
made based on the results of this
study.
1. Since a shift in the location of the
critical section may occur due to flex
ure-shear cracking, the authors recom
mend that the current AASHTO and
ACI requirements for strand develop
ment length be enforced at a critical
section that is located a distance d
from the point of maximum moment
PCI JOURNAL
towards the free end of the strand,

where d
, is the distance from the ex
1
treme compression fiber to the cen
troid of the prestressed reinforcement.
2. For beam sections other than the
point of maximum moment, strand
anchorage should be provided so that
the required prestress force at a given
section may be theoretically devel
oped (based on linear interpolation
of Fig. 13 or equivalent) at a section
located a distance d from the point of
interest towards the free end of the
strand.
3. In place of the above recommen
dations, the designer may elect to pro
vide enough transverse reinforcement
to minimize the shift in tensile de
mand that will occur in the event of
diagonal cracking. The amount of re
quired transverse reinforcing steel
may be calculated using the model
presented in Fig. 16.
Discussion
Recommendations 1 and 2 may ap
pear to be too conservative at first
glance. However, the implications for
most design situations will be small. For
shallow members, the effective depth is
clearly small, so checking development
length requirements at a small distance
of d from the current point will not be
overtaxing on design. For larger mem
bers with fully bonded strands, the issue
of development length is seldom a criti
cal factor in design.
The implications for the design of

members containing blanketed strands
is beyond the scope of this work.
However, the more stringent develop
ment check would certainly be conser
vative. Since the termination of blan
keting is usually staggered, then a shift
in the location of maximum tensile
stress may not be as critical as in this
study due to the inherent redundancy
provided by strands with different
debonded regions.
Finally, the recommendation to
check development length require
ments at a distance d from the point
of interest towards the free end of
the strand is based on the assump
tion that the initiating crack will
have an orientation of 45 degrees
with respect to the longitudinal
member axis. While more sophisti
cated models may be used to esti
mate the most likely crack angle
(considering beam geometry and re
inforcing ratios), the authors believe
the simple and conservative assump
tion of 45 degrees is highly desirable
from a design viewpoint.
NEED FOR
ADDITIONAL RESEARCH
Since this study involved only
bonded strand in SLW concrete, addi
tional research would be beneficial to
quantify the:
1. Effect of flexure-shear cracking
on members with unbonded tendons.
2. Effect of different concrete

mixes, and corresponding aggregate
interlock, on the tension shift noted in
this study. The model used to calculate
required transverse reinforcement
(Fig. 16) in this paper conservatively
ignored the role of aggregate interlock
by assuming the resultant force passed
through the point about which mo
ments were summed.
ACKNOWLEDGMENTS
Appreciation is expressed to the In
diana Department of Transportation
(INDOT) and the Federal Highway
Administration (FHWA) for providing
funding for this research. Specifically,
INDOT personnel Tommy Nantung
and Hasmukh Patel and FHWA per
sonnel Tom Saad were key resource
persons within these organizations.
The authors would also like to
thank the personnel at CSR Hydro
Conduit, Inc., for their support in
meeting the difficult scheduling de
mands of this project. Specifically,
the assistance given by William
Yoder and Rick Yoder is especially
appreciated. In addition, the authors
are appreciative of Jon Jonsson and
Dr. Prasad Rangaraju for their many
hours of service in the design of the
SEW concrete mix and fabrication of
the test specimens.
Finally, the authors want to thank
the PCI JOURNAL reviewers for their
valuable and constructive comments.
93
REFERENCES
1. Zena, D., Albrecht, P., and Lane, S., Investigation of Transfer
and Development Length of Lightweight Prestressed Concrete
Members, Federal Highway Administration (Eisenhower
Grants for Research Fellowships) Preliminary Report, Febru
ary 1995, 271 pp.
2. AASHTO, Standard Specifications for Highway Bridges, Six
teenth Edition, as amended by the 1998 Interim Revisions,
American Association of State Highway and Transportation
Officials, Washington, DC, 1998.
3. Logan, D., Acceptance Criteria for Bond Quality of Strand
for Pretensioned Prestressed Concrete Applications, PCI
JOURNAL, V. 42, No. 2, March-April 1997, pp. 52-79.
4. Moustafa, S., Pull-Out Strength of Strand and Lifting Loops,
Concrete Technology Associates Technical Bulletin, 74-B5,
May 1974. Available from the Precast/Prestressed Concrete In
stitute, Chicago, IL.
5. ACT Committee 318, Building Code Requirements for Struc
tural Concrete (ACI 318-99) and Commentary (ACT 31 8R99), American Concrete Institute, Farmington Hills, MI,
1999.
APPENDIX
a
depth of equivalent rectangular stress block

area of prestressed reinforcment in tension zone
= area of total shear reinforcement crossing crack
width of compression face of member [calculation of
in Eq. (3)1; also, horizontal distance from beam re
action force to section of interest (see Fig. 16)
= resultant compressive force in concrete above inclined
crack (see Figs. 14, 15, 16)
= resultant compressive force in concrete below inclined
crack (see Figs. 14, 15, 16)
= nominal diameter of strand (in.)
= distance from extreme compression fiber to centroid
of prestressed reinforcement
= specified compressive strength of concrete (ksi)
= stress in strand at nominal strength of member (ksi)
= specified tensile strength of prestressed tendons (ksi)
= effective stress in strand after all losses (ksi)
= yield strength of shear reinforcement
= lever arm between resultant tension and compres
sion forces at nominal flexural capacity of section
= development length
= embedment length
= transfer length
= vertical reaction force at end of simply supported
beam (see Fig. 16)
=
1
C
C
db
d
f
f
fse
jd
Ld
Le
L
94
6. Russell, B., and Burns, N., Design Guidelines for Transfer,

Development and Debonding of Large Diameter Seven
Wire Strands in Pretensioned Concrete Girders, Texas De
partment of Transportation in cooperation with Federal
Highway Administration Research Report 1210-SF, January
1993, 286 pp.
7. Kaufman, M. K., and Ramirez, J. A., Reevaluation of the Ul
timate Shear Behavior of High Strength Concrete Prestressed
I-Beams, ACI Structural Journal, V. 85, No. 3, May-June
1988, pp. 295-303.
8. Peterman, R. J., Ramirez, J. A., and Olek, J., Evaluation of
Strand Transfer and Development Lengths in Pretensioned
Girders with Semi-Lightweight Concrete, Final Report, Joint
Highway Research Program FHWAIIN/JTRP-99/3, Washing
ton, DC, August 1999, 183 pp.
9. Buckner, D., A Review of Strand Development Length for
Pretensioned Concrete Members, PCI JOURNAL, V. 40, No.
2, March-April 1995, pp. 84-105.
10. MacGregor, J., Reinforced Concrete: Mechanics and Design
(Third Edition), Prentice Hall, NJ, 1997, pp. 180-195.
NOTATION
resultant of all aggregate interlock forces along in
clined crack
V = horizontal component of aggregate interlock forces
along inclined crack
Vay = vertical component of aggregate interlock forces
along inclined crack
V = force in stirrups crossing crack =
= resultant shear force in concrete above inclined crack
(seeFigs. 14,15,16)
V = resultant shear force in concrete below inclined crack
(seeFigs. 14,15,16)
x = horizontal projection of inclined crack
13i = factor used to enable ultimate flexural capacity calcu
lations to be made by representing the concrete in
compression by an equivalent rectangular stress block
4f = change in strand stress along horizontal projection of
inclined crack
= change in prestress force along horizontal projection
of inclined crack
= factor for type of prestressing tendon used (= 0.28 for
low-relaxation strand)
p, = ratio of prestressed reinforcement = A/bd
Va
PCI JOURNAL
Order Form for CTA TECHNICAL BULLETINS

Volume I
LI
LI
LI
LI
LI
CTAI
CTA2
Production and Construction Techniques
Workability of Concrete (74B 1), January 1974, 13 pp., $10
Accelerated Curing (73B9), 1973, 25 pp., $10
CTA2O
CTA21
Precast Concrete Products
Flange Supported Tees (74B 12), December 1974, 84 pp., $30
Composite Flange Supported Tees (76B 1), January 1976,
$10
LI
CTA4
Accelerated Curing Using Electrically Heated Systems, Part
I, Internal Curing (76B9/10(I)), September 1976, 36 pp., $15
CTA22
Untopped Flange Supported Tees (80B 1), January 1981,
76
O
3
pp.,$
LI
CTA5
Accelerated Curing Using Electhcally Heated Systems, P
II, External Curing (76B9/lO(H)), October 1976,43 pp., $20
CTA23
Performance of Double Tees Without Web Reinforcement
(73B4), April 1973, 28 pp., $10
LI
LI
A24
LI
CTA26
CTA6
Measurement Devices Used for Prestressed and Precast Concrete Production (73B8), August 1973, 23 pp., $10
LI
CTA7
Non-Destructive Testing of Concrete (Including Supplement:
Methods for Evaluating the Concrete Strength of Dry-Cast Products)
(76B3), March 1976, 60 pp. plus 16 pp. supplement, $30
LI
LI
LI
CTA3
Accelerated Curing Using Electrically Heated Forms (73B2),
February 1973. 26 pp., $10
LI
LI
Volume III
CTA9
Volume U
ACI Code
LI
LI
CTA1O
LI
LI
LI
LI
CTA12
Materials, Connections, and a Review of the

High-Strength Concrete (73B5), May 1973, 50 pp., $20
CTA11
Epoxies for Preeast Concrete (75B1), January 1975,
50 pp., $20
CTA13
CTA14
Prestress Losses (73B7), July 1973, 19 pp., $10

Prestress Losses (Phase 1) (84B2), April 1985, 121 pp., $40
CTA15
Prestressed Concrete Pile Splicing (74B2), February 1974,
34. $15
LI
CTA16
LI
CTAI7
Concrete Corbels Attached to Precast Concrete Columns
(80B2), March 1981,28 pp., $10
LI
CTA18
Tube-in-Tube Corbels for Precast Concrete Columns
(80B4), October 1981, 39 pp., $15
LI
CTA19
Ductile Pullout Connections (74Bl 1). November 1974,
$10
Load Bearing Insulated Walls (77B3), February 1979,
CTA28
Precast, Biaxially Prestressed Concrete Bridge Decks
(79B7), April 1980, 38 pp., $15
Design: Flexure, Shear, and Precast Diaphragms
LI
CTA29
Flexural Bond Performance (75B 10/11), September-Octo
ber 1975, 34 pp., $15
LI
LI
CTA3O
Partial Prestress Techniques (75B4), April 1975,46 pp., $20
CTA31
Partial Prestressing for the Precast Industry (81B4), March
1983,65 pp., $25
CTA32
Design Procedure for Unstressed Strand (83B2), April 1984,
85 pp., $30
LI
LI
CIA33
Shear Strength of Prestressed Concrete Beams (76B1 1/12),
November-December 1976,45 pp., $15
LI
CTA34
Web Shear Strength of Prestressed Concrete Members
(85Bl), March 1987, 36 pp., $15
LI
CTA35
Composite Systems Without Roughness (74B6), June 1974,
40 pp., $15
CTA36
Composite Systems Without Ties (76B4), April 1976,
39 pp., $15
LI
LI
CTA37
Shear Diaphragm Capacity of Precast Floor Systems
(73B6), June 1973, 51 pp., $20
LI
CTA38
Shear Diaphragm Capacity of Untopped Hollow Core Floor
Systems (80B3), August 1981, 36 pp., $15
LI
Low Relaxation Strand (85B3), May 1987, 66 pp., $30
CTA27
Thin Shell Roof Components (77B4/5), May 1978,
82 pp., $30
Volume IV
Design of Precast Shear Walls (76B6), June 1976, 60 pp., $25

CTA25
Design of Ledger Beams (Including Supplement) (76B7/8),
July-August 1976, 60 pp. plus 43 pp. supplement, $40
LI
Grouting Precast Floor Systems (82B2), August 1983,
34 pp., $15
s pp., $20
CTA8
Waterproofing Joints in Untopped Decks (79B5), December
1979, 33 pp., $15
Review of ACI Code (82B3), September 1983, 40 pp., $15
BOUND CTA TECHNICAL BULLETIN VOLUMES
LI
LI
CTAV1
pp. $70
Volume I: Production and Construction Techniques, 310
CTAV2
Volume II: Materials, Connections, and a Review of the
ACT Code, 485 pp., $90
LI
LI
CTAV3
Volume III: Precast Concrete Products, 560 pp., $100
CTAV4
Volume IV: Design: Flexure, Shear and Precast
Diaphragms, 490 pp., $90
LI Order the Complete Set of Four Volumes
Only $250
Subtotal from above
Shipping in formation:
Shipping and handling (see table below):
Name______________________________________________
U.S. addresses:
Book rate
PCI pays 4-6 weeks
First class/UPS 25%
ASAP
.
Outside U.S.:
Surface rate.
Air mail
PCI pays
65%
Illinois residents add 8.75% sales tax
Total
2-4 months
ASAP
Company
Address
City______________________ State
Zip
LI Visa/MasterCard
(only)
Tel.:
LI Check
Expiration Date
Payment must be made in U.S. dollars and drawn on a U.S. bank.

Payment must accompany order.
Fax:
Mail or Fax to: PCI, 209 W. Jackson Blvd., Chicago, IL 60606 Tel.: (312) 786-0300 Fax: (312) 786-0353
Strand Development and Transfer

Length Tests on High Performance
Concrete Box Girders
P. Benson Shing, Ph.D.
Daniel E. Cooke
Professor
Department of Civil, Environmental
& Architectural Engineering
University of Colorado
Boulder, Colorado
Engineer, Former Graduate Student

Boulder, Colorado
Mark A. Leonard, P.E.

Senior Structural Engineer
Staff Bridge
Colorado Department of
Transportation
Denver, Colorado
Dan M. Frangopol, Ph.D.

Professor

Boulder, Colorado
Michael L. McMullen, P.E.

Senior Structural Engineer
Staff Bridge
Transportation
Denver, Colorado
Werner Hutter
Research Engineer
Research Branch
Transportation
Denver, Colorado
96
..;
-._:..
J,) ..
Presents the results of an investigation on the

transfer and development lengths required for
Grade 270, 0.6 in. (15.2 mm) diameter prestressing
strand spaced at 2 in. (51 mm) on center in high
performance concrete prestressed box girders.
Three 411 in. (10400 mm) long box girders with
composite topping slabs were tested. The specified
compressive strength of the girder concrete is
10,000 psi (69 MPa). The average transfer length
required for these girders was determined to be
23.4 in. (593 mm). The required development
length was measured using an iterative testing
process involving six flexural tests. The
development length for these girders was
determined to be 60 in. (1500 mm). The
AASHTO/ACI formulas overestimate the transfer
length of the girders by 18 percent and the
development length by 53 percent. The average
pullout strength attained in strand pullout tests is 48
kips (215 kN). The average strand slip measured
right after stress transfer is 0.06 in. (1.52 mm).
PCI JOURNAL
he study presented in this paper

was part of a showcase project
on the replacement of an old
bridge at Interstate 25 over Yale Av
enue in Denver, Colorado, with a new
structure utilizing high performance
concrete (HPC) and Grade 270, 0.6 in.
(15.2 mm) diameter prestressing
strand in side-by-side box girders. The
project was one of the HPC showcases
instituted by the Federal Highway Ad
ministration (FHWA) and carried out
with the cooperative efforts of FHWA,
state transportation departments, and
universities.
The old bridge was 215 ft (65 m)
long and had four spans with cast-inplace T-girders. One design objective
for the new bridge structure was to re
duce the superstructure depth to attain
the desired vertical clearance beneath
the bridge with a minimum change to
the existing vertical alignments.
Side-by-side pretensioned box gird
ers with a composite topping slab
were employed in the new bridge to
minimize construction costs and opti
mize the erection process. The new
bridge has spans of 113 and 98 ft (34.4
and 29.9 m). The maximum span-todepth ratio of the bridge girders is
about 40.
The use of 10,000 psi (69 MPa)
concrete and large diameter strand was
necessary to achieve the desired long
spans and maintain the high span-todepth ratio. The concrete for the top
ping slab was specified to have a de
sign compressive strength of 5100 psi
(35 MPa), for which a concrete mix
design of 5800 psi (40 MPa) was
adopted.
In 1988, the FHWA issued a memo
randum which required that the devel
opment length for prestressing strand
be increased to 1.6 times the develop
ment length stipulated in the
AASHTO Specifications. In addition,
the minimum spacing of prestressing
strand must be four times the strand
diameter and the use of 0.6 in. (15.2
mm) diameter strand was disallowed.
These requirements were instituted be
cause the AASHTO formulas are
based on tests with lower strength ma
terials and smaller strand diameters
and because recent test results had led
to the questioning of the conservatism
of these formulas.
2
diameter strand in 10,000 psi (69

MPa) concrete.
Accompanying the girder tests, ma
terial tests and strand pullout tests
were conducted to characterize the
concrete materials used and the bond
quality of the prestressing strand. The
test program and results are summa
rized in this paper, while the details of
the study can be found in the report by
3
Cooke et al.
At the time this project was initi

ated, the use of 0.6 in. (15.2 mm) di
ameter strand at 2 in. (51 mm) spacing
was prohibited by FHWA unless there
was experimental verification demon
strating the adequacy of the transfer
and development lengths provided.
For this reason, three girders of identi
cal designs were tested at the Univer
sity of Colorado to comply with this
requirement.
The test girders were scaled down
versions of the actual girders to be
used in the bridge at Interstate 25. The
strand size and spacing, concrete
strength, and strand stress at the ulti
mate flexural capacity were approxi
mately the same as those in the actual
girders. This study provides additional
data on the transfer and development
lengths required for 0.6 in. (15.2 mm)
EXPERI MENTAL
PROG RAM
The test program consisted of three
411 in. (10400 mm) long girders,
which provided six transfer length
measurements and six flexural tests
for determining the development
length. The design of the girders, test
411
t1
1
EEl El IEEE EEZE EEl El lEE ZEE
TOP VIEW
COMPOSITE TOPPING SLAB..
LI
BOX GIRDER SECTION
SIDE ELEVATION
C.)
#3 REBAR @ 6
O.C. MAX. (TYP.)
@6 O.C
PRESTRESSING STRANDS
END DETAIL
Fig. 1. Test girder. Note: 1 in.
243 STIRRUPS @ 6 o.C. FOR

FIRST 6 FROM EACH END. 1 -#3
STIRRUP @6 O.C. THEREAFTER
25.4 mm.
97
program, and instrumentation are

briefly summarized below.
Girder Specimens
The girder specimens were reduced
scale models of the box girders used in
the bridge structure at Interstate 25. The
scaling was determined such that the
test girders would have the same perfor
mance as the actual bridge girders in
terms of the stresses at the top and bot
tom of the section, the percentage of the
compression area provided by the top
ping slab, and the strain in the prestress
ing strand at flexurai failure.
The specimen design is shown in
Fig. 1. Each girder specimen con
sisted of a box section with a com
posite topping slab. The girder sec
tion was prestressed with nine Grade
270, seven-wire, 0.6 in. (15.2 mm)
diameter tendons at 2 in. (51 mm)
spacing. The jacking stress was spec
ified to be 204 ksi (1407 MPa). The
concrete for the main box section
was specified to have a compressive
strength of 6.5 ksi (45 MPa) at stress
transfer and 10,000 psi (69 MPa) at
56 days. The mix design for the top
ping slab had a specified compres
sive strength of 5800 psi (40 MPa) at
28 days.
The girder was 34.25 in. (10.4 m)
long with a span length of 33.4 ft
(10.2 m). This was selected to ensure
that development length tests could
be carried out on both ends of each
girder. The girders were designed to
sustain the high shear forces ex
pected during development length
tests by using the maximum shear re

inforcement allowed in the AASHTO
Specifications.
The reinforcement ratio of the gird
ers is 0.007 1, which is close to the
maximum reinforcement ratio of
0.0079 permitted by the AASHTO
Specifications. This results in a strand
strain at ultimate flexural capacity of
approximately 0.011, which is well
below the permissible strand elonga
tion before rupture. The moment ca
pacity of the girders calculated with
the simplified method in AASHTO
and specified material properties is
619 kip-ft (839 kN-m), and that calcu
lated with the strain compatibility
method is 644 kip-ft (872 kN-m).
All three girders were fabricated in
the same casting bed at Rocky Moun
tain Prestress, Boulder, Colorado.
Each tendon was stressed to a value
slightly over the jacking force speci
fied in the design to compensate for
losses due to anchorage slip. The overstress was based on the calibration ob
tained by the fabricator. After the
completion of the prestressing opera
tion, concrete placement began. The
girder concrete was batched on-site at
the fabricators batch plant. The fabri
cator developed the mix design, which
is shown in Table 1. A slump of 4.5
in. (114 mm) was measured.
Upon the completion of concrete
casting, the girders were covered by a
tarp and cured overnight. They were
heat cured for approximately eight
hours. The maximum temperature
measured by thermocouples during the
curing process is about 176F (80C).
Table 1. Girder concrete mix design.

Quantity
lb per Cu yd (kWm
)
3
Material
Type III cement
800 (474)
4
Water
Coarse aggregate
Note: I in.
98
(3/s
263 (156)
in. pea gravel,_CooleY)L
1570 (930)
Pine aggregate (Sand, Cooley)
1320 (782)
Silica fume
30 (14)
Water reducer (Polyheed 997)
12.5 oz/100 lb cement (0.354 kg/lOO kg cement)
Water reducer (Rheo 1000)
15-25 oz./100 lb cement

(0.425-0.709 kgIlOO kg cement)
25.4 mm.
Stress transfer took place two days

after casting with flame cut. The gird
ers are designated numerically 1, 2,
and 3, with the ends of each girder la
beled east (F) and west (W).
Test Setup and
Instrumentation
The test setup and instrumentation
plan are shown in Fig. 2.
Transfer Length Measurement
Transfer length was determined by
measuring strains on girder surface
with a Whittemore gauge. This gauge
had a 7.87 in. (200 mm) gauge length.
Threaded target points were cast into
both sides of each girder at the same
level as the center of gravity of the
strand. As shown in Fig. 2, these tar
get points were located at 3.94 in. (100
mm) spacing over the first 63 in.
(1600 mm) from each end of a girder,
and at 7.87 in. (200 mm) spacing in
the rest of the girder.
The distances between the target
points were measured immediately be
fore and after stress transfer, immedi
ately before and after topping slab
casting, at 14 and 28 days after girder
casting, right before each development
length test, and during the develop
ment length test. These measurements
were used to evaluate the concrete
strains, which were in turn used to de
termine the transfer length, strand
stresses, and prestress losses.
Development Length Tests
For
each test, an estimate of the develop
ment length required for the girder
was made. The member was then
loaded until the peak load was
reached, using a point load at a dis
tance from the end of the member
equal to the estimated development
length. Based on whether a bond fail
ure or flexural failure occurred, the es
timated development length was re
vised for the next test. This process
was repeated until the maximum em
bedment length at which bond failure
would occur was found. This embed
ment length would be the development
length required for the strand.
After eliminating extreme values
from previous research data, it was es
timated that the development length
would most probably fall between 58
and 116 in. (1.47 and 4.22 m). The av
erage of these experimental values
PCI JOURNAL
1! ILRMOCQI JlI..l(S
Cl ilki ) POINTS
AN1) IN PN1.) III .O(KS
(4 iCYI I
4.75
EMBEDMENT LENGTH
5 MECH. GAGE.
34 MECH. GAGE
POINTS @
3.94 O.C.
-[
LOAD
LVDT AT MIDPOINT
AND UNDER LOAD
DURING DEVELOPMENT
LENGTH TESTS
Fig. 2. Test setup and instrumentation. Note: 1 in.
4.75
POINTS @
3.94 O.C.
2VSG
401
was calculated to be 85 in. (2.16 m),

which was used as our first estimate of
development length. The embedment
length at which each test was eventu
ally conducted is listed in Table 2.
The development length tests were
performed in the Structures Labora
tory of the University of Colorado at
Boulder. In each test, the girder was
simply supported on specially fabri
cated supports as shown in Fig. 2.
These supports allowed rotation of
the girder ends through a roller ar
rangement. Teflon pads were placed
between the girder and supports to
allow for free displacement of the
girder end by minimizing possible
friction.
A picture of the test setup is shown
in Fig. 3. The load was applied with a
200 kip (890 kN) capacity hydraulic
jack. An inline pressure sensor was
used to record the applied load. This
device was connected to a data acqui
sition system that was linked to a per
sonal computer. Linear voltage differ
ential transducers (LVDTs) were used
to measure the strand slips and girder
deflection.
As shown in Fig. 4, LVDTs were at
tached to the tendons at the girder end
being tested. These LVDTs were con
stantly monitored in an effort to detect
impending strand slips. LVDTs were
also attached to the girder at midspan
MECH. GAGE,
POINTS @ 7.98 O.C.
SUPPORT WITH
ROLLER AND
TEFLON PAD
/
LVDTS ATIACHEDTO
STRANDS DURING
DEVELOPMENT LENGTH
TESTING
25.4 mm.
and beneath the load point to measure

deflection during the tests. Concrete
strains at the center of gravity of the
strand were measured at selected load
intervals using a Whittemore gauge.
Other Measurements
Bolts were attached to
Camber
one side of each girder at both ends at
the level of the center of gravity of the
girder section. A fishing line was
stretched tight between these bolts and
a reference mark was made on the
girder at midspan. The tension in the
line was maintained the same in mea
suring camber. The camber for each
girder was measured immediately
after prestress release, immediately
before and after the topping slab cast
ing, 14, and 28 days after girder cast

ing, and on the day of development
length testing.
Strand slip
End Slip at Transfer
measuring devices, developed at
FHWA s Turner-Fairbanks Highway
Research Center,
4 were attached to the
tendons at both ends of each girder be
fore stress transfer. These devices
were essentially a section of an alu
minum channel whose web was
clamped to a tendon at approximately
3.5 in. (90 mm) from the end face of
the girder. A Brown & Sharpe DigitCal Mark IV caliper was used to mea
sure the distance from the far flange of
the channel to the girder end. Mea
surements were taken immediately be
fore and after stress transfer.
Table 2. Development length tests.

Test designation
Test date
Embedment length
in. (mm)
1-E
10/2/96
85 (2159)
1-W
10/11/96
81 (2057)
2-W
10/21/96
76 (1918)
2-E
10/28/96
65 (1651)
3-E
11/4/96
60 (1524)
3-W
11/11/96
59 (1497)
99
Fig. 3. Picture of test setup.
Fig. 4. LVDTs attached to tendons.
Strand Pullout Tests

Strand
pullout tests were performed using the
method developed by Moustafa.
5 A
total of eight strand samples, approxi
mately 6 ft (1.83 m) long, were em
bedded to a depth of 18 in. (457 mm)
in a 24 in. deep x 36 in. long x 24 in.
wide (610 x 914 x 610 mm) concrete
block. Two such blocks were cast and
allowed to cure for two days before
the commencement of the pullout
tests. The block design is shown in
Fig. 5. Out of the eighteen tendons in
the two blocks, tests were conducted
only on nine tendons in view of the
consistency of the results.
Curing Temperature
Four ther
mocouples were placed in each girder,
one at each third point of the span in
the bottom flange of the girder, and
one centered in each end block (see
Fig. 2). These devices were read every
30 minutes for approximately a week
after casting. The temperature histo
ries of the three girders were very sim
ilar. The temperatures in Girder 1
within the first 20 hours after casting
are shown in Fig. 6.
facturer, the modulus of elasticity of

the strand is 28,700 ksi (198 GPa) and
the ultimate strength is about 275 ksi
(1.89 GPa).
Chemical tests were conducted to
determine the amount of phosphate
coating on the strand wires. From 21
approximately lI4 in. (44.5 mm) long
wire samples, the average phosphate
content was determined to be 5.7 g/m
2
with a standard deviation of 1.5 gIm
.
2
Girder Concrete
Material tests were conducted fol
lowing ASTM standards wherever ap
plicable on both moist and air cured
specimens. The air cured specimens

were initially heat cured together with
the girders. Tests to determine the
compressive strength, modulus of
elasticity, modulus of rupture, shrink
age behavior, and split cylinder
strength were conducted at the Univer
sity of Colorado (CU).
Additional tests to determine the
shrinkage and creep behavior of air
cured specimens were performed by
Commercial Testing Laboratories
(CTL-Thompson). In general, the air
cured specimens yielded higher
strengths at an early age due to heat
cure and consistently lower strengths
after seven days when compared to the
-#3 STIRRUPS
MATERIAL PROPERTIES
The mechanical properties of the
prestressing strand and concrete used
in the test girders are summarized
below.
Prestressing Strand
The prestressing strand used in the
test girders was manufactured by Insteel Wire Products. A little rust was
observed on the strand surface. Based
on information provided by the manu
100
#3-18
n4
Fig. 5. Pullout test block configuration. Note: 1 in.
25.4 mm.
PCI JOURNAL
sponding to the day of stress transfer

for the girders. The test results are
plotted in Fig. 11. Shrinkage measure
ments were made on specimens cured
under the same environment to take
out the shrinkage effect from the creep
measurement. The shrinkage data are
plotted in Fig. 10 as well. Further
more, creep was calculated using the
7
following empirical law:
200
150
.100
5o
0
0
3 4
9 101112131415
Hours After Girder Casting
60
to.
(2)
6
10 + t
Fig. 6. Temperatures of Girder 1 during curing.
where
moist cured specimens. The compres
sive strength of concrete is plotted
against age in Fig. 7.
The modulus of rupture, fr, and
modulus of elasticity, E, data are plot
ted against
in Figs. 8 and 9. Lin
ear fits to these data have been per
formed and are shown with the
equations representing these lines.
Also shown are the equations given by
ACT 318,6 which aref = 7.5-L and
E=57,000fJ.
As can be seen from the graphs,
the ACT expression for the modulus
of rupture is somewhat conservative,
while the expression for the modulus
of elasticity yields values that are
higher than the test data. In fact, for
normal weight concrete with com
pressive strength between 3000 and
12,000 psi (20.7 and 82.7 MPa), the
following equation is recommended
for the modulus of elasticity based
on research conducted at Cornell
7
University:
(40,000
w
ioooooo)___j
following two days, and air cured at

room temperature thereafter. The two
lime-water cured specimens were
cured in lime saturated water for the
first 28 days and were air cured at
room temperature thereafter. The
shrinkage data are presented in Fig. 10.
Creep tests were performed at
CTL-Thompson on 4 x 8 in. (102 x
203 mm) cylinders under 2800 psi
(19.3 MPa) compression. Due to
measurement problems encountered
in the specimens originally cast with
the girders, the results presented here
are from another batch of specimens
that were cast several weeks after the
girders. However, they had the same
mix design and curing conditions as
the originally cast cylinders.
During the creep tests, the speci
mens were air cured in the laboratory
at a temperature of 73F (23C) and a
humidity of 50 percent. Loading
began two days after casting, corre
unit creep strain at time t

time in days after loading
= ultimate creep strain
Based on research at Cornell Uni
7 an ultimate creep strain of
versity,
0.28 x 10.6 per psi (41 x 10.6 per MPa)
has been suggested for 10,000 psi (69
MPa) concrete. The values based on
the above formula are plotted in Fig.
11 as well. As can be seen from the
graph, the creep exhibited by the spec
imens is much higher than what would
be expected with the above formula.
=
Topping Slab Concrete

For the topping slab concrete, air
cured cylinders reached a compressive
strength of 7900 psi (54 MPa) in 72
days, which was considerably higher
than the specified strength. Moist
cured cylinders showed a little higher
compressive strength than the air
cured ones.
14.0
1.5
12.0
(1)
10.0
in which w is the unit weight of the

hardened concrete in lb per cu ft.
It can be seen in Fig. 9 that the Cor
nell expression fits the modulus of
elasticity data from the moist cured
specimens very well.
At CU, the shrinkage of four air
cured specimens and two lime-water
cured specimens was monitored. The
four air cured specimens were initially
heat cured together with the girders,
cured in lime saturated water for the
8.0
.
6.0
4.0
2.0
0.0
0
I0
20
30
40
50
70
60
80
90
100
Age (days)
Fig. 7. Girder concrete compressive strength. Note: 1 ksi
6.89 MPa.
101
Strand Slip at Transfer

12(8)
800
0
600
0
(8
0
0
-I
70
75
80
85
90
Moist Cuted
Ct
AirCured
--Aaeression
95
Fig. 8. Modulus of rupture data. Note: f is in psi; 1 psi
00
105
110
6.89 x 1O MPa.
Strand slip was measured at the end

of each girder right after stress trans
fer. The elastic shortening which oc
cuffed over the strand between the end
of the girder and the reference point
was calculated and subtracted from the
apparent strand slip to give the actual
strand slip. The average end slip ob
tained is 0.060 in. (1.49 mm).
Strand slip can be theoretically re
lated to the required transfer length.
9
With this knowledge, one can deduce
the maximum allowable strand slip to
comply with the transfer length re
quirement of AASHTO/ACI. Based
on this information and their experi
mental data, Anderson and Anderson
9
proposed the following semi-empiri
cal formula for the maximum allow
able slip:
_fSd
5000
max
(8
70
75
80
85
90
100
95
lOS
110
Jj
Results from the strand pullout

tests, transfer length measurements,
and development length tests are pre
sented here, together with the pre
stress losses estimated from the strain
measurements. Important observa
tions from the development tests are
also presented.
Strand Pullout Strength

Pullout tests were conducted two
days after the casting of the pullout
blocks. The concrete compressive
strength at the time of the tests was
measured to be 8200 psi (56.5 MPa).
The results are shown in Table 3. The
average load at which first slip oc
cuffed is 30 kips (134 kN) and the av
102
where db is the nominal strand

diameter in inches and f
5 is the initial
prestress in ksi.
For the girders tested here, f
5 = 184
ksi (1265 MPa) and db = 0.6 in. (15.2
mm), which result in a maximum al
lowable slip of 0.116 in. (3.0 mm).
The average slip measured is well
below this limit.
Concrete Strain
Fig. 9. Modulus of elasticity data. Note: f is in psi; 1 psi
EXPERIMENTAL RESULTS
950
6.89 x 1 0 MPa.
erage maximum pullout load is 48

kips (215 kN).
Based on a recent study by Logan,
8
the AASHTO/ACI transfer and devel
opment length requirements can be
considered satisfactory for 0.5 in. (13
mm) diameter strand that has a pullout
capacity exceeding 36 kips (160 kN).
Such an empirical criterion is not
available for 0.6 in. (15 mm) diameter
strand. However, by extrapolating the
pullout strength for 0.5 in. (13 mm) di
ameter strand, one may deduce that
the minimum pullout strength required
for 0.6 in. (15 mm) diameter strand is
about 43 kips (192 kN). With this con
sideration, the 48 kips (215 kN) aver
age pullout strength exhibited by the
strand used in this project is more than
adequate.
Concrete strains after transfer were

calculated using the measurements
taken with the Whittemore gauge dis
cussed previously. The initial dis
tances between the embedded target
points in the girders were measured
immediately before stress transfer.
These readings were then subtracted
from subsequent measurements to
yield the change in distance between
the target points. Knowing this change
in distance, the strain was calculated
by dividing this value by the initial
distance between the points.
In this manner, the strains were cal
culated at discrete points along both
sides of the three girders, and the
strains for the corresponding points on
both sides were averaged to obtain one
set of readings along the length of
each girder. For each girder, strains at
the following times were measured:
right after prestress release; seven
PCI JOURNAL
days after girder casting, immediately

before and after topping slab casting;
28 days after girder casting; right be
fore development length testing; and
during development length testing.
0.6
0.5
0.4
0.2
--CU Air Cured Specimens
.-...-.--,-,-,--
---.-
..-
-Pt.
0.3-
Transfer Length
Transfer length was determined
using the 95 percent average maxi
mum strain plateau method. The
strain data collected were examined
and the region in which the strains lev
eled out, or reached a plateau, was
identified. The strain values in this re
gion were then averaged to determine
the magnitude of the average maxi
mum strain.
This value was then reduced by 5
percent to obtain the value of the 95
percent average maximum strain
plateau. Fig. 12 shows the curve used
to determine the transfer length at the
west end of Girder 1. Both the 100 and
95 percent average maximum strain
plateaus are shown as horizontal lines
in the figure.
The transfer lengths obtained at both
ends of the three girders are shown in
Table 4. The average transfer length
measured immediately after stress
transfer is 23.4 in. (593 mm), while
that at 28 days is a little higher.
14
28
42
56
70
84
98
112
126
140
154
Age (Days)
Fig. 10. Average shrinkage strain for girder concrete.
20
40
60
80
100
120
140
160
Age (Days)
Prestress Losses
Using the strain measurements dis
cussed previously, prestress losses in
each of the girders at different times
were calculated. The strain values
throughout the plateau regions mea
sured at each time instance were sub
tracted from the corresponding initial
strains to obtain strain changes along
the girder length. The changes in
strain due to the application of the top
ping slab were calculated from the
measurements taken immediately be
fore and after the topping slab was
cast. This effect is taken out from the
strains measured thereafter. The
strains obtained at each time for dif
ferent girders were then averaged and
multiplied by the modulus of elasticity
of the strand to obtain the cumulative
prestress losses due to the creep and
shrinkage of concrete.
Stress relaxation, fs,rel for low re
1
laxation strand can be estimated by the
following equation:
7
Fig. 11. Unit creep strain for girder concrete. Note: 1 psi
6.89 x 1 0 MPa.
Table 3. Pullout test results.

Strand
number
Load at first slip

kips (kN)
Maximum load
kips (kN)
Ram extension at
maximum load
in. (mm)
28.8 (128.1)
47.1 (209.5)
1.38 (35.1)
27.8 (123.7)
42.5 (189.0)
1.50 (38.1)
32.7 (145.4)
43.2 (192.2)
2.44 (62.0)
30.1 (133.9)
42.5 (189.0)
2.00 (50.8)
24.9 (110.8)
49.1 (218.4)
2.00 (50.8)
29.4 (130.8)
50.4 (224.2)
2.00 (50.8)
+
7
22.9 (101.9)
54.3 (241.5)
2.13 (54.1)
27.5 (122.3)
50.4 (224.2)
2.50 (63.5)
46.4 (206.4)
55.6 (247.3)
1.44 (36.6)
Average
30.1 (133.7)
48.3 (215.0)
1.93 (49.1)
103
I
)
(4)
900
::
where
steel stress at time
= yield stress
f = initial prestress
t = time in hours after stressing
The relaxation calculated with the
above formula is added to the pre
stress losses calculated from the mea
sured strains to obtain the total losses.
The average total losses obtained
from the three girders are shown in
Table 5.
Prestress losses were also calculated
using a time-step procedure. The re
suits of these calculations, using the
measured creep strain and the creep
strain generated from the empirical ex
pression shown in Fig. 11 are pre
sented in Table 5 as well.
It can be seen that the prestress
losses calculated using the empirical
creep equation are much closer to the
measured prestress losses than those
calculated using the measured creep
data. Therefore, it is suspected that the
measured creep might not reflect the
actual creep of the girder concrete.
Camber
A time-step procedure was used to
estimate the camber of the girders,
based on the creep and shrinkage
properties measured (see Figs. 10 and
11) and the steel relaxation given by
Eq. (4). Calculations were repeated
with the creep estimated with the em
pirical formula shown in Fig. 11.
The calculated cambers are com
pared to the measured values in Table
6. It can be seen that the calculated
cambers using the measured creep
data are significantly higher than the
measured values. However, the cam
bers calculated with the creep for
mula are very close to the measured
values.
These results and also the results on
prestress losses indicate that the gird
ers exhibited a creep which is proba
bly much lower than that reflected by
the creep tests. This discrepancy could
be attributed to different curing condi
tions. Even though efforts were taken
to ensure that the concrete cylinders
were cured in the same way as the
104
-:
=24;4In:
1
5L
-_
T-H---
400
f-h
500
iT
300
--
200/
100
------
0--- Measured
Strains After Release
Average Strain Value for 100% Strain Plateau
Average Strain Value for 95% Strain Plateau
0...
0
25
50
75
100
125
150
175
200
Distance from West End (in.)
Fig. 12. Transfer length after release by 95 percent average maximum strain plateau
method (1-W). Note: 1 in. = 25.4 mm.
girders, the girder concrete retains

heat longer due to a larger volume of
concrete in a girder.
It has been shown that the creep of
high strength concrete can be very
sensitive to the curing temperature and
duration. Hence, further studies are
needed to characterize the actual creep
property of high strength concrete
girders in order to make a more reli
able assessment of cambers and pre
stress losses.
Development Length
Development length testing began
50 days after girder casting. The em
bedment length with which each test
was conducted is shown in Table 2.
Only one end was loaded in each test.
Both Girders I and 2 failed in flex

ure, while the failure of Girder 3 was
caused by bond slips at both ends. In
Tests 2-W and 2-E, extremely small
strand slips were observed at the ends
of Girder 2, even though the ultimate
load capacities of this girder at both
ends were governed by the compres
sion failure of the topping slab in the
vicinity of the load point. The ob
served slips could be caused by the de
velopment of shear cracks in the trans
fer zone, which led to localized
increases in strand stress.
The ultimate load capacities of
Girder 3 were governed by strand slips
at both ends. For Test 3E, cracking
was first observed at a load of approx
imately 100 kips (445 kN), which
corresponds to a maximum moment of
Table 4. Transfer lengths measured.

-
1 by 95 percent average maximum strain plateau method

L
Girder end
Immediately after release

in. (man)
28 days after girder casting

in. (mm)
1-E
24.2 (615)
28.3 (720)
1-W
24.4 (620)
26.4 (670)
2E
22.0 (560)
22.0 (560)
2-W
23.8 (605)
25.6 (650)
3-E
23.8 (605)
25.6 (650)
3-W
21.9 (555)
21.7 (550)
Average
23.4 (593)
24.9 (633)
PCI JOURNAL
425 kip-ft (577 kN-m). All nine

strands began to slip at a load of ap
proximately 120 kips (534 kN).
This slip increased rapidly as load
was increased, ultimately reaching a
maximum slip of 0.09 in. (2.29 mm).
A sudden and violent shear failure oc
curred at a load of approximately 151
kips (672 kN), which corresponds to a
maximum moment of 613 kip-ft (831
kN-m). The maximum shear at failure
was 132 kips (587 kN).
It is evident in this test that strand
slips led to a reduction of the shear
capacity and thereby a sudden shear
failure at the end of the girder. The
failure pattern in this test is shown in
Fig. 13.
Results of the development tests are
summarized in Table 7. In summary,
strand slip was first observed in Test
2-E at an embedment length of 65 in.
(1650 mm). Nonetheless, no bond-slip
failure was observed to occur in this
test. Bond-slip failure was first ob
served in Test 3-E, which had an em
bedment length of 60 in. (1500 mm).
From these observations, it can be
concluded that the required develop
ment length for the strand should be
between 60 and 65 in. (1500 mm and
1650 mm).
The development length can be
more conclusively determined from
the variation of the flexural stresses in
the prestressing strands along the gird
ers with different embedment lengths.
In calculating the stresses in the
strands, the strand strain before stress
transfer was assumed to be equal to
0.0711, which corresponds to the jack
ing stress, after anchorage losses, of
204 ksi (1407 MPa). To calculate
strain changes in the strands, it was
necessary to assume that the change in
strain measured with the embedded
surface points was reflective of the av
erage change in strain in the strands.
This assumption is good as long as
no slip occurs between the strand and
the concrete. The change in strain due
to stress transfer, calculated from the
mechanical gauge measurements, was
subtracted from the initial strand strain
to yield the strand strain right after
stress transfer.
Nevertheless, because slip occurs
in the transfer region immediately
after stress transfer, the change in
Table 5. Average measured and calculated prestress losses.

Days
after
release
Average prestress
losses
measured
ksi (MPa)
Calculated prestress losses

using measured
creep data
ksi (MPa)
Calculated prestress losses

using empirical creep
equation
ksi (MPa)
23.1 (159)
22.7(157)
22.7 (157)
29.7 (205)
48.4 (334)
34.6 (239)
14
315(217)
54.7 (377)
39.0 (269)
28
34.6 (239)
60.7 (419)
44.0 (303)
90
43.5 (300)
69.0 (476)
50.8 (350)
Table 6. Measured and calculated cambers.
Time
days after release
Calculated midspan
camber using
measured creep
in. (mm)
Calculated midspan
camber using
creep equation
in. (mm)
Measured
midspan
camber
in. (mm)
1.02 (26.0)
1.02 (26.0)
0.86 (21.7)
7 (before topping)
1.94 (49.3)
1.29 (32.7)
1.19 (30.2)
7 (after topping)
1.84 (46.7)
1.19(30.1)
1.15(29.1)
14
2.02 (51.4)
1.28 (32.4)
1.21 (30.7)
2.17 (55.1)
1.38 (34.9)
1.22 (31.0)
2.43 (61.6)
1.56 (39.6)
NA
28
90
Fig. 13.
Shear failure at east
end of Girder 3.
concrete strain cannot be assumed

equal to the change in strand strain in
the transfer region. Thus, a linear in
terpolation was performed assuming
that the strand strain varied linearly
over the transfer distance determined
with the 95 percent average maxi
mum strain plateau method. The
strand strain was assumed to vary
from zero at the free end of the strand
to the previously determined strand
strain right after transfer at the end of
the transfer region.
The resulting strain was then multi

plied by the modulus of elasticity of
the strand to obtain the variation in
strand stress with distance along the
transfer zone right after stress transfer.
The results are plotted in Fig. 14 for
the east end of Girder 1. The elastic
modulus of the strand was determined
to be 28,700 ksi (198 GPa) from the
stress-strain curve provided by the
manufacturer.
The changes in strand strain which
occurred in the time period between
105
Table 7. Summary of development length test results.

Test designation
Test
i.E
1-W
2-W
2-E
3-E
3-W
10/2/96
10/11/96
10/21/96
10/28/96
11/4/96
11/11/96
50
59
69
76
83
90
Girder
compressive
strength
ksi (MPa)
11.0
(76.0)
11.1
(76.5)
11.2
(76.9)
11.2
(77.1)
11.2
(77.3)
11.2
(77.5)
Topping slab
compressive
strength
ksi (MPa)
7.9
(54.7)
8.0
(54.9)
8.0
(55.1)
8.0
(55.2)
Embedment
length
in. (mm)
85
(2159)
81
(2057)
76
(1918)
65
(1651)
Failureload
kips (kN)
118
(525)
120
(534)
135
(601)
Failure moment*
925
(682)
907
(669)
Maximumshear
kips(kN)
97
(431)
Maximum
deflection
under load
iii (mm)
Test date
Concrete age
(days)
kip-ft (kN-m)
8.0
(55.3)
COMPARISON OF
EXPERIMENTAL RESULTS TO
CODE SPECIFICATIONS
60
(1524)
59
(1497)
135
(601)
151
(672)
145
(645)
967
(714)
857
(632)
895
848
(660)
(626)
100
(444)
113
(504)
117
(521)
132
(587)
128
(568)
2.8
(71)
(69)
2.8
(70)
2.1
(53)
2.5
(64)
Flexure
Flexure
Flexure and shear strengths were

calculated for the girders using the
AASHTO Specifications.
1 The mate
rial properties measured at the time of
each test were used in these calcula
tions. In Table 8, the calculated
strengths are compared to the maxi
mum moment and shear values ob
tained in the tests.
In three of the six tests, the measured
maximum moment values are less than
the calculated values. In Girder 3, it is
the bond failures causing the lower
than predicted failure moments. The
failure moment for Test 2-E is lower
than that predicted probably due to a
premature compression failure of a
poorly compacted section of the top
ping slab near the load point. The max
imum shear forces measured in all the
girder tests are lower than the calcu
lated values. Despite this, both ends of
Girder 3 failed, in whole or in part, by
shear following strand slips.
The required transfer length and de
velopment lengths for the girders have
been calculated in accordance with the
AASHTO Specifications shown in
Eqs. (5a) and (5b):
Maximum slip
in. (mm)
8.0
(55.3)
2.3
(58)
-_
0.005
(0.127)
0.007
(0.178)
Flexure/Slip
Flexure/Slip
0.09
(2.29)
0.08
(2.03)
Slip/Shear
Slip/Flexure
Failure type
Shear
*Jncludes girder self weight.
stress transfer and the day of the test

were then subtracted from the after re
lease strain values to yield the strand
strain before the test. These values
were subsequently used to calculate
the average strand stress, which is
plotted in Fig. 14 as well.
Strain measurements were also
taken immediately after flexural fail
ure. These values were also converted
to strand strain and then to steel stress
using the stress-strain plot for the pre
stressing strand. These values are plot
ted as the heavy solid line in Fig. 14.
The strain values used to develop
this plot do not account for steel re
laxation or strand slip during test
ing. However, these effects are
small. The increases in strand stress
due to the applied loading can be
clearly seen. It is apparent that these
106
is still longer than the required devel

opment length.
Finally, Fig. 16 shows that the flex
ural bond stress propagated to the end
of the transfer zone at the east end of
Girder 3. Hence, one can conclude
that the embedment length of 60 in.
(1500 mm) used at the east end of
Girder 3 should be very close to the
development length required for the
strand.
increases did not reach the transfer

region and that the embedment
length tested was apparently greater
than the required, development
length for the east end of Girder 1.
It is also clear that an increase in
strand stress had occurred in the
transfer region. This is most likely
due to shear cracks that had propa
gated through or near the strand an
chorage region.
The approximate strand stresses cal
culated for the east ends of Girders 2
and 3 are shown in Figs. 15 and 16.
Compared to Girder 1, Fig. 15 shows
that the flexural bond stress moved
closer to the end of the transfer zone at
the east end of Girder 2, but it was still
away from the end of the transfer
zone. Hence, one can conclude that
the embedment length used in this test
(5a)
Ld
(f
fse) db
(5b)
where
stress in ksi in the strand at the
nominal flexural capacity
fse = effective prestress
db = nominal strand diameter
The calculated transfer length is 28
in. (700 mm), which is about 18 per-
f=
PCI JOURNAL
cent greater than the measured values.

The calculated development length is
92 in. (2330 mm), which is about 53
percent longer than the measured
value.
It is interesting to compare the exper
imental results with the transfer length
deduced from the measured end slip
using the slip theory
9 given as follows:
L,
2E
250
Wi.,
150
(6)
0
where
measured end slip
jacking stress immediately be
fore stress transfer
= modulus of elasticity of the
strand
For the test girders, f
0 = 204 ksi
(1407 MPa), E = 28,700 ksi (198
GPa), and the average measured end
slip is 0.06 in. (1.49 mm) as reported
in a previous section. With these val
ues, Eq. (6) results in a transfer length
of 16.6 in. (422 mm), which is about
30 percent less than the average trans
fer length measured immediately after
stress transfer.
15
30
45
6)
75
9)
105
l3
150
135
165
LstivLe Akiig Bean (in)
Fig. 14. Estimated strand stress at east end of Girder 1. Note:

1 in. = 25.4 mm.
The following conclusions can be

drawn from this investigation:
1. The AASHTO/ACI formulas for
transfer and development lengths
overestimate the required transfer
and development identified in the
tests. The formula for transfer length
yields a value 1.18 times the average
value measured immediately after
stress release. The formula for devel
opment length yields a value 1.53
times the measured value. However,
it must be emphasized that the strand
strain at flexural capacity of the gird
ers considered in this study is rela
tively low and the strand exhibited a
good quality bond based on the
strand pullout tests.
2. The average pullout strength for
the untensioned strand is 48.3 kips
(215 kN). Even though no benchmark
value has been established for 0.6 in.
(15.2 mm) diameter strands, this
strength is substantially higher than
the minimum pullout strength of 36
kips (160 kN) recommended by
8 for 0.5 in. (12.7 mm) strand.
Logan
ksi
6.89 MPa;
250
CM
200
CM
CM
150
-
Before Release
1-
----
100
CI)
CONCWSIONS
ffr
AfterRelease
Before Test
-
50
100 kips
After Failure (135 kips)
0
0
15
30
45
60
75
90
105
120
135
150
165
Distance Along Girder (in.)

Fig. 15. Estimated strand stress at east end of Girder 2. Note: 1 ksi
1 in. = 25.4 mm.
6.89 MPa;
250
__r
150
fT
Before Release
I,
It.
if
0
0
After Release
Before Test
lOOkips
147 kips
15
30
45
60
75
90
105
120
135
150
165
Distance Along Girder (in.)

Fig. 16. Estimated strand stress at east end of Girder 3. Note: 1 ksi
1 in. = 25.4 mm.
6.89 MPa;
107
Table 8. Comparison of calculated flexure and shear strengths to the

maximum values obtained in the tests.
kip-fL (kN-iii)
Mn,AASHTO
st
1
v
kips (kN)
Vn,A.ASHTO
kip-ft (kN-ni)
682 (925)
643 (872)
1.06
97 (432)
141 (628)
1-W
669 (907)
644 (874)
1.04
100 (444)
144 (639)
2-W
714 (967)
645 (874)
1.11
113 (504)
149 (661)
2-F
632 (857)
645 (874)
0.98
117 (521)
158 (703)
3-E
660 (895j
645 (874)
1.02
132 (589)
164 (729)
3-W
626 (848)
645 (874)
0.97
128 (568)
164 (729)
AASHTO
Test
1-F
3. The ACT equation for the modu

lus of rupture of concrete is conserva
tive for the 10,000 psi (69 MPa) con
crete used in this study, while the
equation for the modulus of elasticity
tends to yield values higher than the
experimental values. For the modulus
of elasticity, the equation proposed by
researchers at Cornell University pro
vides a good correlation with the ex
perimental data.
4. The camber calculated with the
time-step method, based on the mea
sured creep data, is much higher than
the measured cambers. A much better
correlation between the measured and
calculated cambers is obtained when
the creep strain is calculated with an
empirical equation based on the ulti
mate unit creep strain recommended
by researchers at Cornell University
for 10,000 psi (69 MPa) concrete. This
indicates that the curing conditions for
the girders and concrete cylinders,
such as the temperature and duration,
108
kips (kN)
could be different and creep is sensi

tive to these conditions.
5. The estimation of prestress losses
also depends on the creep property of
concrete, and, therefore, on the curing
conditions. Hence, a good control of
the curing conditions is crucial for a
good estimation of prestress losses and
camber. Further studies are needed to
find better ways to characterize the
creep property of pretensioned high
strength concrete girders.
RECOMMENDATIONS
Results of this study show that the
use of 0.6 in. (15.2 mm) strand at 2 in.
(51 mm) spacing is acceptable for the
given strand, concrete strength, and
design conditions. Furthermore, both
the strand slips measured during stress
transfer and the pullout strengths of the
strand reflect that the AASHTO/ACI
transfer and development length re
quirements are adequate for these
girders. However, general design rec

ommendations cannot be drawn based
on the reported tests alone.
Factors such as the surface condi
tion of prestressing strand and con
crete strength may affect the required
development and transfer lengths. Fur
thermore, it is possible that the re
quired transfer and development
lengths may increase with the increase
of strand strain at failure. Further stud
ies are needed to investigate the pull
out strengths of strands with different
surface conditions and diameters, and
the required transfer and development
lengths for girders with different
strand pullout strengths, concrete
strengths, and reinforcement ratios.
ACKNOWLEDGMENTS
This study was sponsored by the
(FHWA). However, opinions expressed
in this paper are those of the authors
and do not necessarily reflect those of
the sponsor or the Colorado Depart
ment of Transportation (CDOT). The
authors also appreciate the assistance of
Nat Jansen of FHWA, David Price of
the CDOT, and a number of undergrad
uate students, Sam Scupham, Michael
Meiggs, and Ann Grooms, in the exper
imental work.
The constructive input provided by
Susan Lane of FHWA and Don Logan
of Stresscon in this study is also grate
fully acknowledged. Finally, the au
thors would like to thank the review
ers of this article for their constructive
comments.
PCI JOURNAL
REFERENCES
1. AASHTO, Standard Specifications for Highway Bridges, Fif
teenth Edition, American Association of State Highway and
Transportation Officials, Washington, DC, 1992.
2. Buckner, C. Dale, A Review of Strand Development Length
for Pretensioned Concrete Members, PCI JOURNAL, V. 40,
No. 2, March-April 1995, pp. 84-105.
3. Cooke, Dan E., Shing, P. Benson, and Frangopol, Dan M.,
Colorado Study on Transfer and Development Length of Pre
stressing Strand in High Performance Concrete Box Girders,
Report No. CDOT-DTD-R-98-7, Colorado Department of
Transportation, Denver, CO, 1998.
4. Russell, H., Implementation Program on High Performance
Concrete
Guidelines for Instrumentation of Bridges, Re
port No. FHWA-SA-96-075, U.S. Department of Transporta
tion, Federal Highway Administration, Washington, DC, 1996.
5. Moustafa, S., Pull-Out Strength of Strand and Lifting Loops,
Concrete Technology Associates Technical Bulletin, 74-B5,
1974. Available from Precast/Prestressed Concrete Institute,
Chicago, IL.
APPENDIX
db
fr
nominal strand diameter
modulus of rupture of concrete
=
=
=
specified compressive strength of concrete

stress in prestressing strand
yield strength of prestres sing strand
strand stress immediately before stress transfer
initial prestress immediately after stress transfer

effective prestress
= strand stress at nominal moment capacity
fse
6. ACI Committee 318, Building Code Requirements for Rein

forced Concrete (ACT 318-89), American Concrete Institute,
Farmington Hills, MI, 1989.
7. Nilson, A. H., Design of Prestressed Concrete, Second Edi
tion, John Wiley & Sons, Inc., New York, NY, 1987.
8. Logan, D. R., Acceptance Criteria for Bond Quality of Strand
for Pretensioned Prestressed Concrete Applications, PCI
JOURNAL, V. 42, No. 7, March-April 1997, pp. 52-90.
9. Anderson, A. R., and Anderson, R. G., An Assurance Crite
rion for Flexural Bond in Pretensioned Hollow Core Units,
ACI Journal, V. 73, No. 8, August 1976, pp. 457-464.
10. Russell, B. W., and Burns, N. H., Measured Transfer
Lengths of 0.5 and 0.6 in. Strands in Pretensioned Con
crete, PCI JOURNAL, V. 41, No. 5, September-October
1996, pp. 44-65.
11. Roller, J. J., Russell, H. G., Bruce, R. N., and Martin, B. T.,
Long-Term Performance of Prestressed, Pretensioned High
Strength Concrete Bridge Girders, PCI JOURNAL, V. 40,
No. 6, November-December 1995, pp. 48-59.
NOTATION
t =time
w = unit weight of concrete
0 = modulus of elasticity of concrete
E
E = modulus of elasticity of prestressing strand
L = transfer length of prestressing strand
Ld = development length of prestressing strand
unit creep strain at time t
=
ultimate unit creep strain

slip of prestressing strand at end of girder
109
Transfer Length of CFRP/CFCC

Strands for Double-T Girders
Nabil F. Grace, Ph.D, P.E.

Professor and Director of Structural
Testing Center
Civil Engineering Department
Lawrence Technological University
Southfield, Michigan
Presents the results of an experimental investigation addressing the

variations in transfer lengths in two types of carbon fiber reinforced
polymer (CFRP) prestressing strands. Parameters such as the level of
prestress at release, creep, and the rate and method of release of
prestress are considered. In all, eight double-T (DT) prestressed
girders fabricated with high strength concrete were tested. Four DT
girders were prestressed with CFRP Lead/me tendons while the other
four DT girders were prestressed with carbon fiber composite cable
(CFCC) strands. Based on the experimental results, a proposed
modification in an available model for predicting the transfer length
of CFRP Leadline tendons and CFCC strands is suggested. A
comparison of results for the transfer length obtained using the
proposed modification with those of others is made. It is shown that
the calculated and measured transfer lengths are in close agreement
with those obtained by others in the case of Leadline tendons only. It
is also noted that the level of release of prestress has no significant
effect on transfer length.
reinforced polymer (FRP)

prestressing strands have been
used in several structures and
3 because of their high
bridges
strength to weight ratio and noncorro
sive characteristics. It is essential to
understand the behavior of prestressed
concrete members using strands made
of carbon fiber reinforced polymers
(CFRP).
The transfer length is an important
parameter for pretensioned, pre
stressed concrete structural design,
which facilitates determination of the
Fiber
location of the critical section for

checking stresses near the ends of
members and for checking the flexural
and shear strength of the prestressed
member. Thus, adequate knowledge of
the transfer length for CFRP prestress
ing strands is absolutely essential so
that designers can use it easily for the
safe and efficient design of preten
sioned, prestressed structural systems.
The transfer length is defined as the
length required to transfer the full pre
stressing force from the pretensioned
strands to the concrete. In other words,
PCI JOURNAL
the transfer length is the length of the

bond from the free end of the strand to
the point where the prestressing force
is fully effective. This is a function of
several factors that depend on the
properties of both the prestressing ten
dons and the concrete.
These factors include the size and
type of strand, the level of prestress at
transfer, the strength of the concrete at
transfer, the surface condition of the
strand and the method of transfer or
release of the prestress. It is well es
tablished (Mitchell et al.
) that a pre
4
stressing steel strand with a rough ir
regular surface will require a shorter
transfer length than one that is clean
and smooth. Also, a transfer length for
a steel strand will be 6 to 30 percent
greater if the release of the strand is
sudden rather than gradual.
There are many equations available
in the literature, which can be used to
determine the transfer length of steel
prestressing strands. Most of these
equations are empirical and are based
primarily on interpretations of experi
mental results. However, there are
only a few equations which deal with
transfer lengths of FRP strands.
A brief discussion of the equations
and results available in the literature
for transfer lengths of steel and FRP
strands is presented in the work of
Soudki et al.
5 However, for the sake of
completeness, some of the equations
and results are discussed in the section
that follow:
Transfer Length of Stee Strands
Some of the equations for transfer
length of steel strands are summarized
below:
ACI 318-M98/CSA-A23.3-94
The provisions of the ACT 318-98 and
CSA23.3-94 Codes
7 are essentially
6
the same. The equation for transfer
length in these codes is as follows:
In U.S. customary units:
where
Ut = plastic transfer bond stress
fse = effective prestress in the strand
B = bond modulus
= cross-sectional area of the
strand
Shahawy et al.,
8 and Buckner
9 sug
gested that this equation was inade
quate and that fpe should be replaced
with fi the initial stress in the strands.
Alternatively, Russell and Bums also
proposed the same ACT transfer length
equation but with a modification by
introducing a multiplier factor of
(3/2). This revision was found to pro
vide a better correlation with experi
mental data for small strand diameters.
Model by Cousins, Johnson and
Zia
Cousins et al.
11 developed ana
lytical equations for transfer length as
suming elastic and plastic bond zones
within the transfer length:
Using the authors recommended

values for steel strands of B = 300
psi/in. (82 x 10 MPaJmm) and U, =
7 psi (0.55 J7 MPa), this equa
6.7L
tion reduces to Eqs. (3a) and (3b) as
given below:
Ifl U.S. customary units:
0.5 (U,/B) + (fSAp

)I(Jrdb U,) (2)
2
0.012
J7 +
0.048(A, /db)(fse
/f) in.
(3a)
Table 1. Characteristics of LeadlineTM tendons.

23
Characteristics
Specification
Matrix
Epoxy
Carbon fiber volume fraction (percent)
65
Tensile strength, ksi (kN/mm

)
2
328 (225)
Youngs modulus, ksi (GPa)
21320 (147)
Extension at break (percent)
1.5
Specific gravity
1.6
Relaxation ratio (percent)
2.3
Thermal expansion (1/deg. C)
0.68 x 106
Effective cross-sectional area of

8 mm diameter tendon, sq in. (mm
)
2
0.071 (46.1), [0.111 (71.8)1*
Guaranteed tensile strength for

8 mm diameter tendon, kips (kN)
23 4 (104) [36.4 (162)J
*Refe to the tendon of 10 mm diameter.
Table 2. Characteristics of CFCC strands.

18
Lt=/3fpedb in.
(la)
where
L = transfer length of strand
fpe = effective prestress in strand
db = nominal diameter of strand
In metric (SI) units:
Lt=0.O48fpedb mm
(lb)
Characteristics
1 x 7, 10.5 mm
1 x 7, 12.5 mm
Nominal diameter, in. (mm)
0.41 (10.5)
0.49 (12.5)
Actual diameter, in. (mm)
0.43 (10.9)
0.50 (12.7)
Effective cross-sectional
area, sq in. (mm-)
009 (55.7)
0.12 (76)
Linear density, lblft (glm)
0.07 (112)
0.10 (152)
Tensile strength,
ks, (1c\/mm
331 (2.28)
Elongation (percent)
1.7
1.6
Breaking load, kips (kN)
21.6 (96)
31.96 (142)
327 (2.215)
I
111
In metric (SI) units:

1 =3.36[j
L
O.58(Apc/dh)(fccIf) mm (3b)
where f is the concrete strength and

f is the strength of the concrete at
strand release (transfer).
Transfer Length of FRP Strands
There are three types of fibers avail
able for FRP prestressing strands.
These are aramid fibers, glass fibers
and carbon fiber strands. Each of the
available FRP prestressing systems
12
and their anchorages are substantially
different. As a result, there are signifi
cant differences in how load is trans
ferred from the anchor into the FRP
tension element and from the tension
and that these fibers showed little slip

page when compared to steel strands.
4 conducted
Taerwe and Pallemans
a transfer length study using 0.3 and
0.2 in. (7.5 and 5.3 mm) diameter Ara
pree aramid fiber rods. The 0.3 in. (7.5
mm) diameter rods were coated with
Expancel coating to enhance mechani
cal interlock. The 0.2 in. (5.3 mm) di
ameter rods had three different surface
finishes, namely, sanded, Expancel
coating with a smooth surface, and
Expancel coating with a sanded sur
face. The Expancel coating was used
to absorb part of the radial compres
sion of the bars. Based on tests on 29
specimens of varying cross sections, a
transfer length of 16 times the nominal
diameter of Arapree aramid fiber rod
was suggested.
Ehsani et al.
2 carried out tests for
transfer length on Arapree [0.39 in.
(10 mm)], FiBRA [0.41 in. (10.4
element to the concrete. Most of the

early research on transfer length of
FRP strands focused on glass and
aramid fibers. More recently, stud
9 can be found in the literature re
5
ies
garding the transfer length of carbon
fiber strands.
Nanni et
Aramid FRP Strands
3 examined the transfer length of
al.
braided epoxy-impregnated aramid
fiber reinforced polymer tendons.
They conducted an experimental study
on 25 beams. The specimens, with
varying numbers of tendons, were
used with tendon sizes varying from
I,6 to /8 in. (8 to 16 mm). They ob
served that the transfer lengths ranged
from 12 to 20 in. (300 to 500 mm) for
low levels of prestress and from 10 to
22 in. (250 to 550 mm) for high levels
3 concluded
of prestress. The authors
was
the predominant
friction
that
in aramid fibers,
mechanism
bonding
..j
CrossBeam
(8 x
1.5
5
li.
I 8
CrossBeam
(12 s 12) (Typ.)
121
I I
LPre_Cast 13.5
DoubleT
54
[.._12
Elevation
CFRP 4 x 4
Mesh
J CFRP
Mesh
4 x
.__
4
4.5
lb
Tj
75
13.5l
ii
1395
\\
II
2 x 2 CFRP Mesh
12 x 12 CrossB
8 mm Draped Tendons
10 mm CFRP Tendon
8 mm CFRP Draped Tendons

10 mm CFRP tendon
1E1
Girders CDTI-1. 2.3 and 4
1 x 7 (12.5 mm 0)
cFcC Stirrups 0 6
CrossBeam
1 x 7 (12.5 mm 0)
CFCC Draped Strands
1 x 7 (12.5 mm 0)
CFCC Strand
1 x 7 (12.5 mm 0)
CFCC Strand
@ Mid-Span
1 a 7 (7.5 mm 0)
CFCC Longitudinal
Reinforcement
@ Support
Girders CDT2-1. 2.3 and4
Fig.
1. Details of pretensioned, prestressed concrete DT girders.

PCI JOURNAL
TRANSVERSE BARS
[j
LONOIFUDINAI.
TRANSVERSE BAR SPACING

0 4 FOR FIANCE
& 2 FOR WEB
f--iA
BAR WDTH
F-i
LONGIIUO4AL BAR SPACING
4 FOR FLANGE
& 2 FOR WEB
SECtiON AA
1ruar ORrtc rQ_a_
CTILAI.
2 (lIWIhIUM)
3mm
hIAX. TENSILE LOAD (GUARANTE)_....a.2 KFS
BAR
ULTWTE LOAD..................
_1O.O KPS PER BAR
1ENS1L 173 KS
KS
PERCENT OF FERS bT vuit........4
mm)] and Technora [0.29 in. (7.4

mm)] tendons. The Arapree tendons
were round with a sand-impregnated
surface, the FiBRA tendons were
braided, and the Technora tendons
were deformed with a spiral indenta
tion. The transfer length was found to
be 33 bar diameters for FiBRA, 43 bar
diameters for Technora, and 50 bar di
ameters for Arapree. It was also noted
that the transfer length for the Arapree
tendons was significantly affected by
the level of prestress.
Glass FRP Strands
Issa et al.
5
conducted a comparative study of the
transfer lengths of fiberglass and steel
pretensioned members. Seven speci
mens pretensioned with two /s in. (9.5
mm) S-2 glass epoxy strands and five
specimens pretensioned with a single
/2 in. (12.7 mm) steel strand were
tested. They concluded the transfer
lengths to be equal to 28 times the
nominal diameter of the strand. Also,
they observed that the fiberglass had
better bond characteristics than the
steel due to better adhesion and inter
lock at transfer. Additionally, the re
searchers concluded that the transfer
length of fiberglass could be reason
ably predicted using the Cousins et
= 12L.
1
modelwithU
tt
al.
Carbon FRPs
Two types of
CFRP strands are most commonly
used as prestressing strands. These are
CFRP LeadIine tendons* and car
bon fiber composite cables (CFCC)
strands. A brief discussion on the

transfer length results and expressions
developed by a few researchers for
CFRP Leadline tendons and CFCC
strands is presented below.
6 reported transfer
Abdelrahman
length data for eight T-beams pre
stressed with 5/ in. (8 mm) diameter
CFRP Leadline tendons. These trans
fer lengths are based on a prestress
level varying from 40 percent [152 ksi
(980 MPa)] to 60 percent [200 ksi
(1380 MPa)] of the specified guaran
teed strength. The strength of concrete
at release was between 5.4 and 7.3 ksi
(37 and 50 MPa), and the method of
release applied was gradual. The
transfer length was found to be 14.2
in. (360 mm) or 46 bar diameters
when the stress after release was 138
ksi (950 MPa) and 19.7 in. (500 mm)
or 64 bar diameters when the stress
after release was 190 ksi (1310 MPa).
Thus, the reported results indicate that
the transfer length depends on the
level of prestress in Leadline tendons
after release.
7 examined the transfer
Domenico
length and bond characteristics of 20
pretensioned concrete beams pre
stressed by a seven-wire CFCC
9
strand. The CFCC diameter varied
from /2 in. to 5/5 in. (12.5 to 15.2 mm).
His program considered different con
crete covers [2 to 3 in. (50 to 75 mm)],
concrete strengths [5.36 to 10 ksi (37
to 70 MPa)], and prestressing levels
Fig. 2.
Typical NEFMAC
two-dimensional
grid.
(50 to 70 percent of guaranteed

strength). He found the measured
transfer length [varied from 5.5 to
15.7 in. (140 to 400 mm)] to be pro
portional to the diameter of the CFCC
strand and the prestress level.
2 also conducted a
Ehsani et al.
transfer length study on Leadline [0.31
in. (8 mm)] tendons and CFCC [0.33
in. (8.3 mm)] strands. They used the
same specimens for these tests as they
used for tests on Aramid tendons and
found the transfer lengths to be equal
to 54 times the bar diameter for Leadline tendons and 50 times the bar di
ameter for CFCC strands.
Most recently, Mahmoud et al.
9
studied the transfer and development
lengths of CFRP tendons in 43 preten
sioned concrete beams. In their experi
mental study, 17 beams were preten
sioned with Leadline tendons and 26
beams were prestressed with CFCC
/2, and
strands. Three diameters,
/8 in. (10.5, 12.5, and 15.2 mm), were
considered for the CFCC strands and
/,e
in. (8 mm) diameter Leadline ten
don was used.
The concrete strength at transfer
ranged from 3.2 to 5.1 ksi (22 to 35
MPa). The prestressing level varied
from 58 to 80 percent of the guaranteed
strength. The measured transfer length
IN refers to the trade mark name.
* Manufactured by Mitcubishi Oreniical Corporation, Japan.
t Manufactured by Tokyo Rope, Japan.
113
was found to vary from 17.7 to 27.6 in.

(450 to 700 mm) for Leadline tendons
and from 11.4 to 25.6 in. (290 to 650
mm) for the CFCC strands. Based on
the experimental results, Mahmoud et
9 recommended the following equa
al.
tion for the transfer length of CFRP
Leadline tendons/CFCC strands:
L,
1 db) I (a, f
(f,,
)
67
mm
(4)
GR FOR DYFlANGE HFORCEh4ENr

St RAR
LONGfl1JDIkAL RAR
WRIICAL 8A FOR
4EM INFORCDIUIT
flCAL ORID
where
f,,, = prestress at transfer

a, = transfer length coefficient
= 1.9 for Leadline tendons
a, = 4.8 for CFCC strands
6 also reported that
Abdelrahman
the transfer lengths of 0.3 in. (8 mm)
diameter Leadline tendons are 14.2
and 19.7 in. (360 and 500 mm) for
prestressing levels of 50 and 70 per
cent of the guaranteed ultimate
strength, respectively. When com
pared to the above proposed model for
Leadline, the results are within 7 to 9
percent of the predicted values using
Eq. (4). Soudki et al.
reported that
2
the transfer lengths of 0.3 in. (8 mm)
diameter Leadline tendons are 25.6 to
28.5 in. (650 and 725 mm) for pre
stress levels of 50 and 70 percent of
the guaranteed strength, respectively.
These reported values are 20 percent
higher than those predicted using the
proposed model in Eq. (4).
z
Fig. 3. Three-dimentional CFRP cage arrangement for flexural and shear
reinforcement for DT Girders CDT1 -1, 2, 3 and 4.
I x 7 (1.5 mm
0)
CFCC
Flenunil Ruinforceinento 6
in tranevirie Direction
13
I x 7 (7.5 mm
CFGC
Flaszal Ruinforcemant 3
In I.angltudlnol Direction
40
(Mdth of DtGdnr)
Fig. 4. DT-flange reinforcement.
TEST PROGRAM
The major purpose of this experi
mental program was to determine the
transfer length of CFRP Leadline ten
dons and CFCC strands in double-T
(DT) girders. Eight DT prestressed
girders fabricated with high strength
concrete were tested. Four DT girders
were prestressed with CFRP Leadline
tendons while the other four DT gird
ers were prestressed with carbon fiber
composite cable (CFCC) strands.
1
7 (7.5 mm
CFCC Oaii/ Web
Construction of Test Girders

Two groups of four DT girders were
fabricated. The first group of girders is
designated as CDT1-1, CDT1-2,
CDT 1-3 and CDT 1-4 and prestressed
with Leadline tendons, while the sec
ond group of girders is designated as
CDT2-1, CDT2-2, CDT2-3 and
1 a 7 (12.5 nun 0)
CFCC Preinneluned
Sband/ Web
Web reinforcement
Fig. 5. Details of CFCC flexural and shear reinforcement for DT Girders CDT2-1, 2, 3
and 4.
PCI JOURNAL
CDT2-4 and prestressed with CFCC

strands.
These girders are designed as sim
piy supported girders. However, they
are currently being used for the con
struction of continuous
21 prestressed
concrete bridge models; hence, the no
tation CDT, which stands for continu
ous double-T bridge model. The first
digit refers to the type of strand used
(1 for Leadline tendon and 2 for
CFCC strand); the last digit refers to
the DT girder number. In this paper,
DT girders are simply supported.
Basic characteristics of the material
for Leadline
23 tendons and CFCC
18
strands are presented in Tables 1 and
2, respectively. Fig. 1 shows the de
tails of Girders CDT1-l, 2, 3, 4 and
CDT2-l, 2, 3, 4 with their elevation,
cross sections, and positions and sizes
of tendons/strands. The span of each
girder is 19.33 ft (5890 mm). Each DT
girder is 13.5 in. (345 mm) deep and
its flange is 40 in. (1015 mm) wide.
The center-to-center distance be
tween the two webs of each DT girder
is 20 in. (510 mm) and the thickness
of flange is 1.5 in. (38 mm). The
widths of each web at the bottom of
the DT girder and at the flange are 3.5
and 4.5 in. (90 and 115 mm), respec
tively. Six tendons/strands per DT
girder were pretensioned prior to plac
ing the concrete. Two top (draped)
Leadline tendons are of 0.3 in. (8 mm)
diameter while the bottom tendon in
each web (straight) is of 0.4 in. (10
mm) diameter. The diameter of each
CFCC strand is 0.5 in. (12.5 mm).
:__-_
Fig. 6.
Arrangement at live
end for prestressing DT
Girders CDT1-1, 2,3
and 4.
Fig. 7.
Hold-up arrangement
for CFRP Leadline
prestressing tendons.
Reinforcement Details
Arrangements of reinforcements for
girders CDT1 and CDT2 are described
below:
Girders CUT1
For this group of
girders, four CFRP reinforcing cages
using NEFMACTM* CFRP grids of 2 x
2 in. (50 x 50 mm) spacing were fabri
cated for flexural and shear reinforce
ments. A typical NEFMAC two-di
mensional grid with its specification is
shown in Fig. 2.
Note that the size of each grid in the
flange is 4 x 4 in. (100 x 100 mm)
while the size of each grid in the web
*NEFMACTh1 denotes trade mark name; provided by

Autocon Equipment Incorporated, Canada
24
Fig. 8. CFCC strands with anchor head and CFRP spirals.

115
Fig. 9. Girders CDT2-1, 2 prior to casting concrete.
is 2 x 2 in. (50 x 50 mm). The threedimensional CFRP cage arrangement

is shown in Fig. 3.
Girders CDT2
For this group
of girders, CFCC strands of 0.3 in.
(7.5 mm) diameter are used to pro
vide flexural reinforcement in the
flanges. The strands are laid in trans
verse and longitudinal directions
with a spacing of 6 and 3 in. (152
and 76 mm), respectively, as shown
in Fig. 4. Two-legged 0.5 in. (12.5
mm) diameter CFCC stirrups with a
spacing of 6 in. (152 mm) are pro
vided for shear reinforcement in
each web. In addition, four longitu
dinal CFCC strands of 0.3 in. (7.5
mm) diameter are provided on each
side of each web. The details of the
shear reinforcement are shown in

Fig. 5.
Details of the arrangement of pre

stressing strands in the DT girders are
given elsewhere.
21 For prestressing a
girder, each Leadline tendon/CFCC
strand was tensioned (before casting
of concrete in the fonn) by a hydraulic
jack at the live end of the girder and
anchored to the bulkhead using a spe
cial anchorage system.
22
The arrangement for prestressing
DT Girders CDT1-1, 2, 3, and 4 at the
live end is shown in Fig. 6. In this fig-
Fig. 10. Sudden release of LeadlineTM

prestressing tendons using hand-held saw.
116
Prestressing Setup,
Release of Prestressing Force
and Strain Measurements
ure, it is shown that the straight Leadline tendon (bottom of Web B) is con
nected to the hydraulic jack for ten
sioning. The connection between the
hydraulic jack and the tendon is made
by means of a connecting steel rod,
which passes through a coupler that is
screwed on to the Leadline anchor
head.
To provide the desired draped pro
file of the tendons, internal holddown and hold-up arrangements, fol
lowing recommendations of earlier
22 were installed. A
investigations,
typical hold-up arrangement for
CFRP Leadline prestressing tendons
is shown in Fig. 7.
A set of CFCC strands with pre
attached anchor heads is shown in Fig.
8. These anchor heads are placed on
the dead end side of the girders and
are connected to the load cells (posi
tioned between the lock nut and bulk
head). It is also shown in Fig. 8 that
these CFCC strands are provided with
CFRP spirals around them within the
transfer length.
These spirals around the strands aid
in preventing the development of
cracks at the transfer of prestress to
the concrete. A typical view of Girders
CDT2-1 and 2, prior to casting the
concrete, is shown in Fig. 9. This fig
ure also shows the details of CFCC re
inforcements for flexure and shear in
Girders CDT2- 1 and 2.
The concrete mix was cast into the
girder forms and after it had gained a
strength of 7000 psi (48 MPa) (at 7
Fig. 11. Typical CFCC strands after sudden release of prestressing forces.
PCI JOURNAL
days), the tendons/strands were re

leased by means of a hand-held saw as
shown in Figs. 10 and 11. Fig. 10
shows a typical case for the sudden re
lease of prestressing forces in Leadline
rods using a hand-held saw, while
Fig.1l shows the sudden release of
prestressing forces in CFCC strands.
Measurements of concrete strain at
the centroid of the prestressing strands
were made by two persons using de
tachable mechanical gauges with
equally spaced DEMEC target points
(see Fig. 12). The average spacing be
tween each pair of target points at
midspan was 7.9 in. (200 mm). Closer
spacing between target points was
used at the two ends of each girder.
The concrete strain recorded was de
termined by averaging the readings
taken by both persons.
The prestressing force in each ten
don was monitored using load cells at
the dead ends of the girders. The aver
age prestressing force in the draped
Leadline tendons was approximately
17 kips (76 kN). In the straight Leadline tendons, the average prestressing
force was approximately 19 kips (85
kN). Therefore, the average stresses in
the 0.3 and 0.4 in. (8 and 10 mm)
Leadline tendons were 239 and 170
ksi (1650 and 1170 MPa), respec
tively, as shown in Table 3.
The data in Table 4 indicate that the
average prestressing force in the
CFCC strands was approximately 16.6
kips (74 kN), producing an average
stress of 140 ksi (971 MPa). It is also
shown in Table 3 that the average
elongations of the 0.3 and 0.4 in. (8
and 10 mm) Leadline tendons were
approximately 3.24 and 2.24 in. (80
and 60 mm), respectively. In Table 4,
the data show that the average elonga
tion of the CFCC strands was 2.7 in.
(69 mm).
RESULTS AND DISCUSSION

The measured transfer lengths at the
live and dead ends of the girders and
the effects of parameters such as level
of release of prestress, time elapsed
after transfer of prestress and longterm effect on transfer lengths of
CFRP Leadline tendons/CFCC strands
(represented by the CDT1/CDT2
groups of girders) are presented in this
section. Finally, a model for predicting
points.
the transfer length of CFRP Leadline
tendons/CFCC strands based on a re
gression analysis is proposed and a
comparison of the measured results
with the theoretically predicted values
is made.
METHOD OF RELEASE
The last two columns of Table 3
show the average transfer length for
Leadline tendons at the live and dead
ends, respectively, for Girders CDT1
1, 2, 3, and 4. Note that these results
correspond to the sudden release of
the prestressing force, except in the
case of Girder CDT1 -1. In this girder
(CDT1-1), the prestress was released
slowly at the live end and released
suddenly at the dead end.
It is observed that the transfer
lengths of the CFRP Leadline tendons
at the live end of the girders are
slightly larger than the transfer lengths
at the dead end. Furthermore, slow re
lease results in a larger transfer length
than sudden release (70 times the av
erage diameter of Leadline tendons
versus 55 times the average diameter).
Table 4 shows a similar difference in
transfer lengths of CFCC strands in
Girder CDT2.
Examination of this table also sug
gests that slow detensioning of CFCC
strands results in a larger transfer
length in comparison to sudden release
(39 times the diameter of CFCC
strands versus 34 times the diameter).
-
Furthermore, it is worth mentioning

that the difference in transfer lengths
for slow and sudden release of CFRP
Leadline tendons is about 28 percent;
the corresponding difference for
CFCC strands is about 15 percent.
Thus, the difference in transfer
lengths obtained for slow release and
sudden release of CFRP strands is in
contrast to what is well known for
steel strands (Mitchell et al.
). In their
4
results, flame-cutting of the steel
strands resulted in transfer lengths of
about 6 to 30 percent greater than
those determined for slowly released
steel strands.
Effects of Level of Release
of Prestress Force
In order to examine the effect of
level of release of prestress force on
the transfer lengths of Leadline ten
dons and CFCC strands, the variations
of concrete strain along the span of
Girders CDT1-4 and CDT2-1 for three
levels of release of prestress (i.e., 33,
67, and 100 percent) are shown in
Figs. 13 and 14, respectively. Here, 33
percent release refers to the top two
(draped) tendons being released at the
same time; 66 percent release refers to
the top four draped tendons being re
leased and 100 percent release refers
to all tendons being released.
Note that the Leadline tendons are
released suddenly at both ends while
the CFCC strands are released slowly
117
at the live end and suddenly at the

dead end. Figs. 13 and 14 show that
the concrete strain profile along the
girders is not following a straight-line
pattern. This is attributed to the
draped profile of the tendons/strands
due to the two internal hold-down and
hold-up arrangements at the bulkhead.
Note that the two hold-downs were
cut before the release of the

tendons/strands.
From Figs. 13 and 14, transfer length
is determined by the 95 percent AMS
(Average Maximum Strain) method;
25
that is, by measuring the distance from
the end of the member to the point of
intersection of a line representing the
strain variation in the transfer zone and
a horizontal line representing 95 per

cent of the average of the strains con
tained within the strain plateau of the
effective prestressing force. It is shown
that there is no appreciable change in
transfer length with the increase in
level of release of prestressing force for
both the CFRP Leadline tendons and
the CFCC strands.
Table 3. Details of prestressing forces, elongation, end slip at release and transfer lengths of
Girders CDT1-1, -2, -3, -4.
Girder
designation
Web
Tendons
Prestressing
force
kips (kN)
Stress
ksi (GPa)
Elongation
in. (mm)
Al
18 (80.1)
252 (1.74)
3.6 (90)
0.21 (5.2)
N/A
A2
19 (84.5)
266 (1.83)
3.8 (97)
N/A
N/A
A3
21 (93.4)
189 (1.30)
2.4 (60)
0.07 (2.2)
N/A
Bl
16 (71.2)
224 (1.54)
3.2 (81)
0.11 (2.8)
N/A
B2
16 (71.2)
224 (1.54)
3.1 (78)
N/A
N/A
B3
22 (97.9)
198 (1.36)
2.5 (64)
N/A
N/A
CDT1-1
End slip_______
Live end
Dead end
in. (mm)
in. (mm)
Measured transfer lengths

Live end
Dead end
in. (mm)
in. (mm)
480
(
)t
25.0 (635)*
18.9
22.4 (570)*
17.5 (445)
18.9 (480)
18.0 (460)1
r
A
Al
17 (75.6)
238 (1.64)
3.2 (82)
N/A
N/A
A2
18 (80.1)
252 (1.74)
3.5 (89)
N/A
N/A
A3
18 (80.1)
162 (1.12)
2.3 (57)
N/A
CDT1-2
N/A
______
Bi
16 (71.2)
224 (1.54)
3.0 (76)
N/A
B2
17 (75.6)
238 (1.64)
3.4 (86)
N/A
B3
18(80.1)
162(1.12)
2.3(57)
N/A
N/A
Al
17 (75.6)
238 (1.64)
3.1 (79)
0.31 (7.9)
0.21 (5.3)
A2
17 (75.6)
238 (1.64)
3.0 (76)
0.13 (3.3)
0.12 (3.0)
A3
17(756)
153 (1 05)
23 (57)
003 (08)
009(23)
20 (89.0)
280 (1.93)
3.0 (76)
0.08 (2.0)
0.17 (43)
B2
16 (71.2)
224 (1.54)
3.3 (83)
0.04 (1.0)
0.04(1.0)
B3
15 (66.7)
135 (0.93)
2.1 (54)
0.05 (1.3)
0.06 (1.5)
Al
17 (75.6)
238 (1.64)
3.1 (79)
0.08 (2.0)
0.15 (3.8)
A2
17 (75.6)
238 (1.64)
3.4 (86)
0.20 (5.1)
0.16 (4.1)
A3
20 (89.0)
180 (1.24)
2.0 (51)
0.10 (2.5)
0.15 (3.8)
Bl
18 (80.1)
252 (1.74)
3.0 (76)
0.09 (2.3)
0.13 (3.3)
B2
14 (62.3)
196 (1.35)
3.1 (79)
0.13 (3.3)
0.05 (1.3)
B3
20 (89.0)
180 (1.24)
2.0 (51)
0.07 (1.8)
0.09 (2.3)
N/A
-
N/A
510
(
20.0 )t
19.5
495
(
)t
CDT1 3
Bi
B
,,
420
(
)t
16.0 (410)i
495
(
19.5 )t
18.9 (480)1
20.0 (5l0)
19.5
16.5
CDT1-4
(495)1
Note 1: Al, A2, BI and B2 are 8mm draped tendons.

Note 2: A3 and B3 are 10 mm bottom straight tendons.
*Slow release.
tSUdden release.
PCI JOURNAL
Variation of Prestressing Force

The variation of prestressing force
with time, measured at the dead end,
during the slow release of the prestress
ing force in Strand B 1 at the live end of
Girder CDT2-1 is shown for all six
CFCC strands in Fig. 15. Also, the vari
ation in prestressing forces, measured at
the dead end, during sudden release of

Strand B 1 at its dead end is shown in
Fig. 16. Hence, in these figures the pre
stressing forces are not the actual pre
stressing force in the strands within the
concrete but are, in fact, a measure of
forces in the portion of the strands lying
between the end of the girder and the
bulkhead at the dead end.
Thus, these figures represent the

variation of apparent prestressing
forces in the strands with time. Exami
nation of Fig. 15 suggests that when
Strand B 1 is released slowly at the live
end, the apparent prestressing force in
Strand B 1 (the released strand), as
well as the other unreleased strands,
decreases during the slow release
Table 4. Details of prestressing forces, elongation, end slip at release and transfer lengths of
Girders CDT2-1, -2, -3, -4.
Girder
designation
Web
Tendons
Prestressing
force
kips (kN)
Stress
ksi (GPa)
Elongation
in. (mm)
Al
15.8 (70.3)
134.1 (925)
2.5 (64)
0.04 (1.0)
N/A
A2
154(685)
1307 (901)
3 1 (79)
008 (20)
N/A
A3
15.0 (66.7)
127.3 (878)
2.4 (61)
0.09 (2.3)
0.04 (1.0)
B1
16.3 (72.5)
138.4 (954)
2.5 (64)
0.04 (1.0)
0.03 (0.8)
B2
16.3 (72.5)
138.4 (954)
2.9 (74)
0.01 (0.3)
N/A
B3
15.6 (69.4)
132.4 (913)
2.4 (61)
0.03 (0.8)
0.04 (1.0)
Al
16.3 (72.5)
138.4 (954)
2.8 (71)
0.05 (1.3)
0.10 (2.5)
A2
16.4 (72.9)
139.2 (959)
2.9 (74)
0.04 (1.0)
0.04(1.0)
A3
16.1 (71.6)
136.7 (942)
2.4 (61)
0.07 (1.8)
CDT2-1
End slip
Live end
Dead end
in. (mm)
in. (mm)
Measured transfer lengths

Live end
Dead end
in.(mm)
in. (mm)
21 0 (533)*
470
(
185 )t
20.0 (508)*
330
(
13.0 )t
(495)*
470
(
18.5 )t
18.9 (480)*
457
(
18.0 )t
584
(
23.0 )t
18.0
16.5 (419)
406
(
16.0 )t
-_____
CDT2-2
--
0.02 (0.5)
-_________
BI
15.6 (69.4)
132.4 (913)
2.4 (61)
0.04 (1.0)
0.02 (0.5)
B2
15.8 (70.3)
134.1 (924)
2.8 (71)
0.05 (1.3)
N/A
B3
16.7 (74.3)
141.8 (977)
2.9 (74)
0.02 (0.5)
0.09 (2.3)
Al
17.4 (77.4)
147.7 (1018)
2.9 (74)
0.06 (1.5)
0.09 (2.3)
A2
16.5 (73.4)
140.1 (966)
2.5 (64)
0.04 (1.0)
0.05 (1.3)
16.8 (74.7)
142.6 (983)
2.5 (64)
0.02 (0.5)
N/A
A3
CDT2-3
-------
19.5
457
(
)T
_____
BI
16.6 (73.8)
140.9 (972)
2.9 (74)
0.06 (1.5)
N/A
B2
17.1 (76.1)
145.2 (1001)
2.6 (66)
0.05 (1.3)
N/A
B3
17.7 (78.7)
150.3 (1036)
2.9 (74)
0.03 (0.8)
N/A
Al
16.8 (74.7)
142.6 (983)
2.8 (71)
0.05 (1.3)
N/A
A2
18.7 (83.2)
158.7 (1095)
2.8 (71)
0.04(1.0)
N/A
A3
17.0 (75.6)
144.3 (995)
2.5 (64)
0.03 (0.8)
N/A
B1
18.1 (80.5)
153.6 (1059)
2.9 (74)
0.05 (1.3)
N/A
B2
17.5 (77.8)
148.6 (1024)
2.9 (74)
0.05 (1.3)
N/A
B3
17,4 (77.4)
180 (1.24)
2.5 (64)
0.05 (1.3)
N/A
_.
18.9
480
(
)T
16.0
406
(
)t
CDT2-4
16.0 (406)
15.5 (394)
Note 1: Al. A2, A3, Bl, B2 and B3 are 12.5 mm CFCc strands.
Note 2: N/A: Results are not available.
*Slow release.
v5udd release.
119
time. This is attributed to the loss of

force in the strands due to the move
ment of the girder towards the dead
end of the strands (after the release of
Strand B 1 at its live end).
Nonetheless, it is seen from Fig. 16
that when Strand B 1 is suddenly re
leased at its dead end, the apparent
prestressing force in the unreleased
strands starts increasing as the force in
Strand Bi (the released strand) goes to
zero. This is also attributed to the ten
dency of the girder to move towards
the live end (in this case) after the re
lease of Strand B 1 at its dead end. Due
to this girder movement, a portion of
the unreleased strands between the end
of the girder and the bulkhead are
stretched further; hence, the load cells
mounted at the dead end record an in
creased force in the strands.
To examine the time taken to re
lease the prestress force in a tendon!
strand during slow and sudden release
conditions, the variation of the pre
stressing forces in all six CFCC
strands with time is shown in Figs. 17
and 18, during slow and sudden re
lease, respectively. The total time over
which the force in the strand shows
variation is taken as the time of release
of the prestressing force.
The average release time is com
puted by dividing the sum of the total
time of release of prestressing forces
in each strand by the number of
strands, It is observed that the aver
age time of release of prestressing
force when the strand is released
slowly is 34 seconds, while the aver
age time of release of prestressing
force when it is released suddenly is
only one second.
C
C
r-)
Span (inches)
Fig. 13. Transfer length at release of prestressing forces of Girder CDT1 -4, Web A.
0.00030
0.00025
0.00020
0.00015
0.00010
0.00005
Span (inches)
Fig. 14. Transfer lengths at release of prestressing forces of Girder CDT2-1, Web A.
20000
19000
Long Term Transfer Length
18000
To study the effect of time on the

transfer length of Leadline tendons
after the release of the prestressing
force, the distribution of concrete
strain along the span of Girder CDT14 is shown in Fig. 19 for different pe
riods (i.e., at transfer, 14, 28 and 300
days after transfer of prestress to the
concrete). As expected, it can be seen
that the concrete strain increases with
time elapsed after the transfer of pre
stress, due to the combined effect of
shrinkage, creep and relaxation of
stress in the tendons.
17000
120
16000
U
0
15000
Ca
C
14000
C-
13000
12000
11000
10000
0
10
15
20
25
Time (Sec.)
30
35
40
45
50
Fig. 15. Slow release of prestressing strand Bi at live end of Girder CDT2-1.
PCI JOURNAL
Nevertheless, it is shown that there

is no significant effect of time on the
transfer length of the tendons; that is,
the increase in transfer length 300
days after transfer is only about 7.8
percent. Similarly, to examine the ef
fect of time on transfer length of
CFCC strands, the variation in con
crete strain for different periods of
measurement (i.e., at transfer, 7, 122,
214 and 391 days after transfer of pre
stress to the concrete) along Girder
CDT2-l is shown in Fig. 20. It is ob
served that the concrete strain for
Girder CDT2-l also increases with
time.
It is also observed that there is, on
average, a 7 percent increase in the
transfer length 391 days after pre
stress transfer. Thus, it may be noted
that the long-term effect (i.e., the ef
fect of elapsed time after transfer of
prestress to concrete) on the transfer
lengths of CFCC strands and CFRP
Leadline tendons is almost the same
as that reported for steel strands
(Mitchell et al.
).
4
20000
18000
16000
14000
4)
12000
10000
8000
6000
4000
2000
0
0.5
1.5
2.5
Time (Sec.)
3.5
Fig. 16. Sudden release of prestressing strand Bi at dead end of Girder CDT2-1.
20000
18000
16000
14000
.e
12000
10000
Effect of Creep on Concrete Strain
8000
To examine the long-term creep ef

fect, the strain distribution at the
midspans of Girders CDT 1-1, 2, 3,
and 4 was monitored for 393 days,
while for Girders CDT2- 1, 2, 3, and 4
it was monitored for 431 days. The
strain distributions for Girders CDT11 and CDT2-l are shown in Figs. 21
and 22, respectively. In the case of
Girder CDT1-1, 393 days after the
transfer of prestress, the average strain
in the concrete at midspan of the
girder increased to about 3.5 times that
of the strain at the time of transfer.
In the case of Girder CDT2-1,
there is an increase in strain of about
4.8 times after a monitoring period
of 431 days. This difference in con
crete strains in Girders CDT1 and
CDT2, prestressed with CFRP Leadline tendons and CFCC strands, re
spectively, may be attributed to the
axial rigidity (EA, i.e., modulus
times area of the tendons/strands)
and to the different bond characteris
tics of Leadline tendons and CFCC
strands with the concrete. Thus, it is
noted that the creep effect is more
significant in the case of Girder
CDT2-1, which was prestressed with
a
6000
4000
2000
0
0
10
15
20
30
25
Time (see.)
35
40
45
50
Fig. 1 7. Slow release of prestressing strands at live end of Girder CDT2-1.
16000
release time IsSe

\cverage
4000
2000
0
0.5
1.5
Time (see)
2.5
Fig. 18. Sudden release of prestressing strands at dead end of Girder CDT2-1.
3.5
121
CFCC strands, than that in the case

of Girder CDT1-l, which was pre
stressed with Leadline tendons.
Figs. 23 and 24 show the nonlinear
variation of concrete strain with time
at the top and bottom of midspan for
Girders CDT1 -1 and CDT2- 1, respec
tively. It is observed that for a fixed
time, the strain at the bottom of the
web is higher than at the top, irrespec
tive of the girder types. This is at
tributed to the presence of tendons at
the bottom, which increase the net
compressive strain in the concrete. It
is also shown that the concrete strain
is increasing with time at the top as
well as at the bottom for both groups
of girders (CDT1 and CDT2 types).
In the above, it is worth noting that
the rate of increase in strain is more
significant (for both groups of girders)
immediately after the release of pre
stressing force. As time passes, the
strain levels out to a constant value.
By examining the slope of the nonlin
ear variation of concrete strain with
time for both girders, it is observed
that the time rate of increase in con
crete strain is larger in the case of
Girder CDT2 than in the case of
Girder CDT 1.
Transfer Length Correlation
0.0012
0.0010
95% Average Maximum Strain
0,0006
.
S
0
0.0004
0.0002
,Vi20.Oin
VT
0.0008
,V,VVVV
i J....
(a, f
) mm
67
a1
122
=
=
it
200
150
100
50
Span (inches)
Fig. 19. Time effect on transfer length of Girder CDT1-4, Web A.
0.00 12
0.00 10
0.0008
C
,
C,,
0.0006
S
0
C.)
0.0004
100
Span (inches)
Fig. 20. Time effect on transfer length of Girder CDT2-1, Web A.
At Transfer
154
538
35 Days
371
984
230 Days
153 Days
581
393 Days
723
605
1502
1398
1711
Web A
WebB
WebA
104
257
540
503
597
(4)
501
where
..\
LJ
)
5
(a
(fr, db) I
...
Hold down of internal

draped CFRP rods. ( T9p.)
0.0002
To examine the applicability of the

currently available empirical model
9
for determining the transfer length for
Leadline tendons and CFCC strands in
DT girders made of high strength con
crete, the relationships between the
measured transfer length and other pri
mary variables such as prestress at
transfer, diameter of the strands and
concrete strength at transfer are shown
in Figs. 25 and 26. Fig. 25 corre
sponds to Leadline tendons while Fig.
26 corresponds to CFCC strands.
Based on a regression analysis, the
following modifications in the transfer
length coefficient
of the model de
veloped by Mahmoud et al.
9 for the
transfer length of Leadline tendons
and CFCC strands are proposed:
V V
.LA
O5In
..!
Transfer
l4Days
28Days
V 300 Days
926
1354
1259
1566
Web B
1.95 for Leadline tendons

2.12 for CFCC strands
Fig. 21. Strain distribution at midspan of Girder CDT1-1
u in. per in.).

PCI JOURNAL
At Transfer
Z.Pays
122 Days
391 Days
431 Days
46
175
291
493
541
394
218
298
917
684
Web A
WebB
WebA
60
It should be noted that the proposed

value of the constant (a,) for Leadline
tendons is close to the value devel
19 but differs
oped by Mahmoud et al.
significantly for CFCC strands. In this
modified model, the transfer length
coefficient (a,) for CFCC strands is
2.12, while the transfer length coeffi
cient is 4.8 in the model of Mahmoud
et al.
19
146
286
453
780
410
392
COMPARISON OF
TRANSFER LENGTHS
1073
1055
Web B
Fig. 22. Strain distribution at midspan of Girder CDT2-1
(j.t
in. per in.).
1800
1600
1400
1200
1000
Cs
800
600
400
200
100
200
Time (Days)
300
400
Fig. 23. Rate of increase in strains at top and bottom of Girder CDT1 -1, Web A.
Figs. 27 and 28 represent a bar chart

comparison of the transfer length data
predicted by the proposed modifica
tions and those of other re
25 for Leadline tendons
6
searchers
and CFCC strands, respectively. In
these two figures, Letters A and B
with the girder numbers represent
Webs A and B of a particular girder.
Each column in this chart represents
the value of the transfer length of the
strand used in a corresponding web
using the appropriate model.
By comparing the obtained values,
it is evident that ACT 3l8986 and the
25 models (devel
Russell and Burns
oped for steel strands) highly overesti
mate the transfer lengths of CFRP
Leadline tendons and CFCC strands.
Furthermore, the results obtained by
the model with a modified coefficient
9
are closer to those of Mahmoud et aL
only for Leadline tendons. However,
in the case of CFCC strands, the re
sults are on an average 1.5 times
9
higher than those of Mahmoud et al.
CONCLUSIONS
1800
1600
1400
.
1200
1000
800
600
400
200
100
200
300
400
Time (Days)
Fig. 24. Rate of increase in strains at top and bottom of Girder CDT2-1, Web A.
Based on the results of this investi

gation, the following conclusions can
be drawn:
1. Transfer lengths of CFRP Leadline tendons vary from 66 to 73 times
the diameter of the tendon for slow re
lease and from 47 to 59 times the di
ameter of the tendon for sudden re
lease. The corresponding range of
transfer lengths for CFCC strands is
found to be 33 to 47 and 27 to 38
times the diameter of the strand for
slow and sudden release, respectively.
2. Slow release results in larger
transfer lengths for Leadline tendons
and CFCC strands (about 28 and 15
percent larger, respectively) in com
123
parison to those determined for similar

strands when they are released sud
denly. The differences in transfer
lengths for slow and sudden release of
CFRP Leadline tendons/strands is in
contrast to that for steel strands, i.e.,
sudden release of steel strands results
in a transfer length of about 6 to 30
percent greater than that determined
for similar strands released gradually.
The average time taken for slow re
lease of the prestressing force in a
strand is 34 seconds while for sudden
release it is one second.
3. There is no appreciable change in
the transfer lengths of CFRP ten
dons/CFCC strands with the variation
in the level of release of prestressing
force.
4. Long-term effects result in ap
proximately a 7 percent increase in
transfer lengths of CFRP tendons!
strands over a period of one year. This
result is almost the same (i.e., 6 per
cent) for steel strands.
5. The bond characteristics and axial
rigidity (EA) of CFRP Leadline tendons
and CFCC strands with the concrete
have a significant effect on the concrete
strain in the DT girders. The concrete
strain is increased about 3.5 times the
strain at transfer in the girders pre
stressed with CFRP Leadline tendons
while the corresponding increase in
strain in the girders prestressed with
CFCC strands is 4.8 times of that at the
transfer of prestress.
6. The time rate of increase in con
crete strain is significant immediately
after release of prestress and eventu
ally becomes constant after about a
year, irrespective of the type of
strand.
7. The proposed modification in the
currently available model for predict
ing the transfer length of CFRP Leadline tendons results in a good agree
ment with the measured values
obtained in this investigation as well
as limited data reported in the litera
ture. However, in the case of CFCC
strands, it differs significantly with
those of other researchers.
9
RECOMMENDATIONS
In view of the highly unconserva
tive results
25 (recommended for
6
steel strands) for the transfer length
of CFRP Leadline tendons and
124
700
600
500
a
a
400
:I300
-
200
100
0
200
400
600
f,,1,/
800
1200
1000
/1
67
f,O
Fig. 25. Transfer length correlation for LeadIineTM prestressing tendons.
C
V
F-
200
400
600
800
1000
1200
/1
l
6
fO
Fig. 26. Transfer length correlation for CFCC strand,
CFCC strands, as well as the much

lower transfer lengths (in compari
son to those obtained by the pro
posed modification) for CFCC
strands obtained by Mahmoud et
9 the following recommendations
al.,
are offered:
1. The ACI 3l8986 and Russell
and Burns
25 models should not be
used for predicting the transfer length
of CFRP Leadline tendons and CFCC
strands, as they will result in an inac
curate estimate.
2. The transfer length for CFRP
Leadline tendons can be reasonably
predicted by the modified equation as

given below:
1
L
)
67
(fr, db) / (1.95f
3. The accuracy of the proposed

model, as given in Eq. (4), for the
transfer length of CFCC strands needs
to be evaluated and corroborated by
further research.
ACKNOWLEDGMENT
This investigation is supported by a
consortium of the National Science
PCI JOURNAL
LB ACI-318 R-95
Russel and Burns
Mabmoud ct a!.
Proposed model
1200
1000
800
.0
bS
C
U
600
400
200
0
CDTI-IA
CDTI-1B
CDTI-2A
CDTI-3A
CDTI-3B
CDTI-4A
CDTI-4B
Fig. 27.
Comparison of
transfer length data
for LeadlineTM
tendon.
LB ACI-318 R-95
J Russel and Burns
Mahmoud et al.
Proposed model
1200
1000
800
600
400
200
CDT2-1A CDT2-IB CDT2-2A CDT2-2B CDT2-3A CDT2-3B CDT2-4A CDT2-4B
Foundation (Grant No. CMS 9900809),

Wright Patterson Air Force Base,
Michigan Department of Transporta
tion, ACT Concrete Research Council,
Holnam, Inc., Mitsubishi Chemical
Corporation and Sumitomo Corporadon, and Tokyo Rope Manufacturing,

Inc., and Mitsui Corporation.
The support of Dr. V. J. Gopu and
Dr. John Scaizi, program directors at
Fig. 28.
Comparison of
transfer length data
for CFCC strand.
the National Science Foundation are

greatly appreciated. The author also
wants to thank the reviewers of the
PCI JOURNAL for their constructive
comments.
125
REFERENCES
1. Abdelrahman, A. A., Tadros, G., and Rizkalla, S. H., Test
Model for First Canadian Smart Highway Bridge, ACI Struc
tural Journal, V. 92, No.4, July-August 1995, PP. 451-458.
2. Shehata, F., Abdelrahman, A., Tadros, G., and Rizkalla, S. H.,
FRP for Large Span Highway Bridge in Canada, Proceed
ings of the U.S.-Canada-Europe Workshop, Recent Advances
in Bridge Engineering, U. Meier and R. Betti (Editors), Zurich,
Switzerland, 1997.
3. ACT Committee 440, Stateof-the-Art Report on Fiber Rein
forced Plastic Reinforcement for Concrete Structures, Ameri
can Concrete Institute, Farmington Hills, MI, 1996, 153 pp.
4. Mitchell, D., Cook, W. D., Khan, A., and Tham, T., Influence
of High Strength Concrete on Transfer and Development
Length of Pretensioning Strand, PCI JOURNAL, V. 38, No.
3, 1993, pp. 52-64.
5. Soudki, K. A., Green, M. F., and Clapp, F. D., Transfer
Length of Carbon Fiber Rods in Precast Pretensioned Concrete
Beams, PCI JOURNAL, V. 42, No. 5, September-October
1997, pp. 78-87.
6. ACT Committee 318, Building Code Requirements for Struc
tural Concrete (ACI 318-98) and Commentary (ACI 318 R
98), American Concrete Institute, Farmington Hills, MI,
1998.
7. C. S. A. Standard CAN 3-A23.3-94, Design of Concrete
Structures for Building, Canadian Standards Association,
Rexdale, Ontario, Canada, 1994.
8. Shahawy, M. A., Issa, M., and Batchelor, B. de V, Strand
Transfer Length in Full Scale AASHTO Prestressed Concrete
Girders, PCI JOURNAL, V. 37, No. 3, May-June 1992, pp.
84-96.
9. Buckner, C. D., A Review of Strand Development Length for
Pretensioned Concrete Members, PCI JOURNAL, V. 40, No.
2, March-April 1995, pp. 84-99.
10. Russell, B. W., and Burns, N. H., Measured Transfer Lengths
of 0.5 and 0.6 in. Strands in Pretensioned Concrete, PCI
JOURNAL, V. 41, No. 5, September-October 1996, pp. 44-65.
11. Cousins, T. E., Johnson, D. W., and Zia, P., Transfer and De
velopment Length of Epoxy Coated and Uncoated Prestressing
Strands, PCI JOURNAL, V. 35, No. 4, July-August 1990, pp.
92- 103.
12. Ehsani, M. R., Saadatmanesh, H., and Nelson, C. T., Transfer
and Flexural Bond Performance of Aramid and Carbon FRP
Tendons, PCI JOURNAL, V. 42, No. 1, January-February
1997, pp. 76-86.
APPENDIX
A
13. Nanni, A., Utsunomia, T., Yonekura, H., and Tanigaki, M.,
Transmission of Prestressing Force to Concrete by Bonded
Fiber Reinforced Plastic Tendons, ACI Structural Journal, V.
89, No. 3, May-June 1992, Pp. 335-344.
14. Taerwe, L., and Pallemans, I., Free Transfer of AFRP Bars in
Concrete Prisms, Proceedings of the Second International
RILEM Symposium, Ghent, Belgium, 1995, Pp. 154-163.
15. Issa, M., Sen, R., and Amer, A., Comparative Study of Trans
fer Length in Fiber Glass and Steel Pretensioned Concrete
Members, PCI JOURNAL, V. 38, No. 6, November-Decem
ber 1993, pp. 52-63.
16. Abdelrahman, A., Serviceability of Concrete Beams Pre
stressed by Fiber Reinforced Plastic Tendons, Ph.D. Thesis,
Department of Civil Engineering, University of Manitoba, Win
nipeg, Manitoba, Canada, 1995, 331 pp.
17. Domenico, Bond Properties of CFCC Prestressing Strands in
Pretensioned Concrete Beams, M.S. Thesis, Department of
Civil Engineering, University of Manitoba, Winnipeg, Mani
toba, Canada, 1995, 160 pp.
18. Tokyo Rope Manufacturing Co. Ltd., Technical Data on
CFCC, JAPAN, October 1993, 100 pp.
19. Mahmoud, Z. I., Rizkalla, S. H., and Zaghloul, E. R., Transfer
and Development Lengths of Carbon Fiber Reinforced Poly
mers Prestressing Reinforcements, AC! Structural Journal, V.
96, No. 4, July-August 1999, pp. 594-602.
20. Soudki, K. A., Lee, C., and Green, M. F., Transfer Length of
CFRP Rods in Prestressed Concrete Beams, Proceedings of
First CSCE Structural Speciality Conference, Edmonton, Al
berta, Canada, May29 to June 1, 1996.
21. Grace, N. F., Innovative System, Continuous CFRP Prestressed
Concrete Bridges, AC! Concrete International, V. 21, No. 10,
October 1999, pp. 42-47.
22. Grace, N. F., and Abdel-Sayed, G., Behavior of Externally
Draped CFRP Tendons in Prestressed Concrete Bridges, PCI
JOURNAL, V. 43, No. 5, September-October 1998, pp. 88-101.
23. LeadlineTM Carbon Fiber Tendons/Bars, Product Manual, Mit
subishi Chemical Corporation, Tokyo, Japan, October 1994.
24. Erki, M. A., and Bakht, B., Nonmetallic Shear Reinforce

ments for Reinforced Concrete Beams, First Structural Spe
ciality Conference, Alberta, Canada, May 29 to June 1, 1998,
pp. 159-165.
25. Russell, B. W., and Burns, N. H., Measurement of Transfer
Lengths on Pretensioned Concrete Elements, Journal of Struc
tural Engineering, V. 123, No. 5, May 1997, Pp. 541-549.
NOTATION
cross-sectional area of member

cross-sectional area of prestressing strand
B = bond modulus
db = nominal diameter of strand
E = Youngs modulus of elasticity
f = specified compressive strength of conrete
f = compressive strength of concrete at strand release (transfer)
126
effective prestress in strand

prestress at transfer
1 = transfer length of strand
L
= plastic transfer bond stress
= transfer length coefficient
= 3.14159
.ve =
PCI JOURNAL
Manual for the Design of Hollow Core Slabs

Second Edition
or the architect or consulting engineer, the newly revised Second Edition of the Manual for the
Design of Hollow Core Slabs is intended to serve as a guideline for working with hollow core
F slabs, a guide for the use and application of hollow core slabs and an indication of some of the
unique properties of hollow core slabs. For the plant engineer, the manual presents backup and refer
ence material for dealing with everyday design problems.
Included in the Manual are extensively illustrated
chapters on:
Hollow Core Slab Systems
Design of Hollow Core Slabs
Special Design Considerations
Diaphragm Action with Hollow Core Slabs
Connections in Hollow Core Slabs
Fire Resistance of Assemblies Made with
Hollow Core Slabs
MANUAL
For the Design of
HOLLOW
SLABS
2nd Editj
0
CORE
Acoustical Properties of Hollow Core Slabs

Guide Specifications for Precast, Prestressed
Hollow Core Slabs.
Also included is a selected list of references and
index.
PCI Members $40.00

Non-Members $80.00
-
Shipping Information
Name
Manual(s) will be shipped as soon as pos

sible
Book Rate by U.S. Postal Service
unless otherwise requested.
Shipping and handling charges are as fol
lows (check one):
Firm___
Address
(If P.O. Box, also give street address for shipping.)
City
Shipment
To U.S. addresses:
Book Rate
FirstCIass/UPS
Next day service..
pays
$10/book
$30/book
4-6 weeks
ASAP
Overnight
Outside U.S.:
Surface rate
Air Mail
$30/book
$75/book
2-4 months
ASAP
....
Charge
ci
State
Zip____________ Country
Delivery
All payments must be U.S. denominated

and drawn on a U.S. bank.
Telephone
Fax
(required)
Form of payment (Payment must accompany order): D Visa/MasterCard(on/y)

Credit Card #
Exp.
# of books_____ x Price per book
Check
Date
Illinois residents add 8.75% sales tax $

Shipping and handling charges $
Total $
Send order to: PCI, 209 W. Jackson Blvd., Chicago, IL 60606
Phone 312-786-0300
Fax: 312-786-0353
OPEN FORUM
PROBLEMS AND SOLUTIONS
The comments and opinions expressed herein are those of the contributing
authors and do not necessarily reflect official PC! policy. Some of the provided
answers may have alternate solutions. Reader comments are invited.
Fonnwork Issues
Ql: Are all steel form designs the same?
Al: The answer to this question is no, because there are
different types of steel forms depending on the application.
In general, there are two types of custom steel forms,
namely, self-stressing forms and non-self-stressing, or freestanding forms.
Designing non-self-stressing forms includes designing
the form skin, longitudinal skin stiffeners and transverse
gussets such that the form will resist the hydrostatic pressure of the concrete. The deflections of the form must not
exceed the allowable dimensional tolerances of the concrete product.
Self-stressing forms require that the forms be designed to
resist the hydrostatic and compressive forces introduced
with prestressing.
Twenty-five years ago, the standard design for selfstressing forms was to make use of compression bars to
carry the prestress forces. The skin was designed to resist
Fig. 1. Original standard design for self-stressing forms.

128
only the hydrostatic forces from the concrete (see Fig. 1).
After much thought and analysis, we determined that in
some cases the compression bars could be eliminated by increasing the thickness of the skin and increasing the size
and quantity of the longitudinal stiffeners. We called this a
"stress-skin design" and began incorporating this design in
1978 (see Fig. 2). Subsequently, we contracted with The
Consulting Engineers Group (CEG) to carry out a load test
to determine the actual stress levels in a double tee form
using this design.
Upon completion of the test, CEG furnished us with a
summary and analysis of the test results. This information
allowed us to further refine our design of self-stressing
forms. The elimination of stressing bars (when possible) allowed us to build more economical forms.
In some cases, such as inverted-tee beam forms, a combination of skin and stressing bars are used to carry the prestressing forces.
Fig. 2. Current stress-skin design for self-stressing forms.

PCI JOURNAL
Table 1. Checklist of questions before ordering forms.
When ordering a form or comparing quotes from different form manufacturers, always be specific about the
form' s intended use and then review the following partial checklist:
Is the form self-stressing or non-self-stressing?
If the form is self-stressing what is the required capacity?
What is the skin thickness?

What is the gusset material thickness and spacing?
What size chamfers are required?
Are there any extra holes required in the form skin? If
so, how are they plugged when not in use and by
whom?
Are all form sections matched and marked?
Is Vibrotrack included for external vibration?
Are any heat holes required?
If top and under ties are required, what are the spacings?
On double tee forms, will strands be deflected? If yes,
with a top or bottom depressor system?
Are product headers required and/or included?
Are stem fillers required and/or included?
So, what is a standard form design today? We have almost always used 3!16 in. (4.76 mm) thick gussets spaced 30
in. (762 mm) on center as our standard on most forms.
However, we have some customers who ask for 1/4 in. (6.35
mm) gussets on 20 in. (508 mm) centers and 3/s in. (9.53
mm) gussets where form sections bolt together. These producers want their forms to last a long time! Other customers
will tell us that they purchase a form to cast one job only. In
these cases, gusset spacings of 40 in. (1016 mm) centers
will probably be adequate.
Frequently, we get requests concerning modifications required to cast a slightly different product using an existing
form. This generally requires additional stressing capabilities. Often, this happens with stadium riser forms.
As an example, some producers have used the same forms
in casting several different stadiums. They have requested
that we design their stadium riser forms with the intent of
using them for multiple projects. This requires the use of
design concepts that may make the form a little more expensive up front, but much more economical over the years.
In general, thicker form skin and closer gusset and stiffener spacings will result in a more durable and longer-lasting form.
When ordering forms, it is important to specify the exact
form needs for each particular project. Are the forms intended to last 20 years, or will they be used for only six
months with the intention of discarding them after the job
is complete? Table 1 provides a checklist of questions that
should be asked before ordering forms.
In summary, some form designs may vary from one
manufacturer to the next. It is important to know what
kind of form is needed, how long the form is intended to
last, and how much experience the form designer possesses. The old adage, "you get what you pay for," also
applies to custom steel forms. Therefore, in planning the
purchase of forms, it is prudent to deal with an experienced form manufacturer.
Does the form have drafted or straight sides?

Are there any keyways required and/or included?
[Contributed by John E. Dobbs, Executive Vice President,

and B. A. (Skip) Plotnicki, Vice President, Engineering,
Hamilton Form Company, Inc., Fort Worth, Texas]
DISCUSSION NOTE
The Editors welcome discussion of reports, articles, and
problems and solutions published in the PCI JOURNAL
The comments must be confined to the scope of the article
being discussed. Please note that discussion of papers appearing in this issue must be received at PCI Headquarters
by February 1, 2001.
129
REVIEWS OF TECHNICAL PUBLICATIONS

Durability Survey of Segmental
Concrete Bridges
Maurice D. Miller
The first American Segmental Bridge
Institute (ASBI) inventory of concrete
segmental bridges built in the United
States and Canada was published in
1994. The 1994 survey identified 194
structures by type and age. The ASBI
survey evaluated 96 of these bridges
based on bridge inspection reports that
rated the bridge conditions using the
(FHWA) guidelines. The 1994 report
rated all bridges as fair or better with 98
percent of the bridges rating as satis
factory or better. Five years later, ASBI
again solicited inspection reports from
the owners to continue the evaluation of
concrete segmental construction.
Responses were obtained for 164
bridges, with inspection reports and
condition ratings for 131 bridges. Again
in 1999, all bridges were rated as fair
or better. Of the 131 bridges, 99 per
cent had superstructure ratings of sat
isfactory or better, 79 percent had
superstructure ratings of good or bet
ter, and 31 percent had superstructure
ratings of very good or better. The
findings of the survey again indicate
that concrete segmental bridges are
performing well with time. Three Appen
dices provide a list of segmental
bridges, bridge condition ratings (1994
and 1999 surveys), and bridge wearing
surface.
American Segmental Bridge Institute,
9201 N. 25th Avenue, Suite 150B,
Phoenix, Arizona 85021-2721, 2000,
30 pp.
Excessive Strand End Slip in

Prestressed Piles
Michael F. Petrou, Baolin Wan, Walter
S. Joiner, Constantin G. Trezos, and
Kent A. Harries
Presents the results of a research pro
ject that investigated excessive strand
end slip observed recently in some pre
stressed concrete piles. From mea
surements taken in the field, it is appar
ent that the problem of excessive initial
strand slip is independent of pile shape
and size. Strand end slip is evident in
piles of different states in the southeast
130
United States. Excessive strand end

slip was found in both the top and bot
tom of the cross section of the piles,
although the top portion of the section
generally exhibited much higher initial
slip. Several preventative measures
can be adopted to reduce the exces
sive strand end slip: (a) proper con
crete mixture proportioning to reduce
top bar effect; (b) use of higher
strength concrete with the lowest pos
sible slump and setting time; (c)
assessment of the condition of the
strands prior to installation to ensure
excellent bond characteristics; (d)
gradual release of prestress, with an
optimal release sequence; and (e) use
of adequate vibration to ensure consol
idation.
posium were presented at the ACI Con

vention in Chicago, Illinois, March 1999.
The papers were reviewed in accor
dance with ACI policy. The proceedings
deal with different aspects of high-per
formance fiber reinforced thin products:
the extrusion process; the pultrusion
technique for producing high-perfor
mance thin products; long-term durabili
ty of glass fibers; PVA fibers; long-term
durability of nonwoven and woven
glass fibers; integral formwork panels
made of glass roving; permanent formwork with carbon fibers; hybrid laminate
composites with various types of mesh
es and discontinued fibers; and the
effect of low contractile admixture in
glass fiber concrete.
AC! Structural Journal, V. 97, No. 5,

September-October 2000, pp. 774-782.
American Concrete Institute, P.O. Box

9094, Farmington Hills, MI, 483339094, 2000, 216 pp.
Super-T Bridges of the Southern

Expressway
Harry Turner
The Southern Expressway in Adelaide,
Australia, is one of the largest infras
tructure projects ever undertaken in this
country. A major ingredient of the
superstructure of this bridge system is
the so-called precast, prestressed
Super-T girder. The principal advantage
of Super-Ts is the immediate availability
of soffit formwork, which allows a con
tractor to pour the composite deck slab
far quicker than would be possible with
a steel beam bridge. In addition, the
thin-walled box girder shape of the
Super-T provides stability in handling,
good load distribution, low lifted weight
and clean appearance.
Engineering Essentials, Cement Con
crete Association of Australia, Sydney,
Australia, V. 3, No. 1, September 2000,
pp. 4-5.
High-Performance FiberReinforced Concrete Thin Sheet

Products
Alva Peled, Surendra P. Shah, and
Nemkumar Banthia (Editors)
These proceedings are the results of

the effort of ACI Committee 579, Ferro
cement and Other Thin Reinforced
Products. Original papers for this sym
Factors Affecting the Use of

Precast Concrete Systems
David Arditi, Uluc Ergin and Suat Gunhan
The reasons precast concrete systems
are not used extensively in the United
States were explored through four mail
surveys that were sent to 100 contrac
tors, 100 design firms, 100 precast con
crete manufacturers and 100 local labor
unions. The findings indicate that most
of the prerequisite conditions for the
extensive use of industrialized building
systems already exist. The findings also
indicate, however, that there is a major
shortage of expert personnel capable of
designing and organizing projects involv
ing precast concrete systems and that
almost half the contractors surveyed did
not observe significant cost savings
when they used precast concrete sys
tems. Architects and engineers aware
ness of precast concrete systems and
the number of competent precast con
crete designers and erectors can be
increased by providing adequate expo
sure of engineering and architecture stu
dents to these systems and by providing
frequent and numerous continuous edu
cation programs for the benefit of active
professionals.
Journal of Architectural Engineering,
American Society of Civil Engineers, V.
6, No. 3, September 2000, pp. 79-86.
PCI JOURNAL
REFERENCE CARDS
KEYWORDS: architectural precast concrete; buildings; commercial
buildings; columns; construction; design (structural); faade; parking
structures; precast concrete; wall panels.
KEYWORDS: beams; development length; design (structural); pre

cast concrete; prestressed concrete; prestressing strands; reinforce
ment; research; semi-lightweight concrete; transfer length.
ABSTRACT: Precast concrete played a prominent role in building

the new multi-million dollar Lazarus Department Store in downtown
Pittsburgh, Pennsylvania. The glitzy department store is tour stories
high with an underground three-level parking structure. Highly articu
lated precast concrete panels with deep reveals give the building
scale and character, and also complement the nearby historic build
ings. To speed construction, structural precast concrete columns are
used in the parking structure. This article presents the conceptual
and architectural design features of the building, design considera
tions, and erection highlights of the project.
ABSTRACT: In this study, 18 development length tests were carried

out on single strand rectangular and multiple strand T-shaped semilightweight beams having design compressive strengths of 7000 psi
(48 MPa). In the rectangular beam tests, the design moment capacity
was exceeded in all specimens. However, in the T-beam tests, bond
failures at loads below the design capacity occurred in some speci
mens immediately after the formation of a flexure-shear crack near
the loading point. Additional T-beam tests showed that the bond fail
ure at loads below the design capacity occurred in some specimens
immediately after the formation of a flexure-shear crack near the
loading point. The authors recommend that the current AASHTO and
ACI requirements for strand development should be enforced at a
critical section that is located a distance d from the point of maxi
mum moment towards the free end of the strand, where d is the dis
tance from the extreme compression fiber to the centroidof the pre
stressed reinforcement.
REFERENCE: Muse, Gar, and Di Giacomo, Anthony, Precast Con

crete Panels Give Scale and Grandeur to Lazarus Department Store,
PCI JOURNAL, V. 45, No. 5, September-October 2000, pp. 20-29.
KEYWORDS: bridges; construction; marine construction; piers; pre
cast concrete; prestressed concrete.
ABSTRACT: Precast, prestressed concrete is used at many loca
tions for a multitude of functions throughout the multi-billion dollar
Central Artery/Tunnel Project in Boston, Massachusetts. This article
describes two bridges and a marine pier where precast, prestressed
concrete was selected as the ideal material to span these structures.
REFERENCE: Donington, Keith, Towell, Paul, and Chandra, Vijay, Cen
tral Artery/Tunnel Project: Precast/Prestressed Structures Span the Big
Dig, PCI JOURNAL, V. 45, No.5, September-October 2000, pp. 30-33.
KEYWORDS: anchorages; design (structural); edge effects; headed
studs; precast concrete; research; shear strength; steel strength.
ABSTRACT: The Precast/Prestressed Concrete Institute sponsored a
comprehensive research program to assess the shear capacity of headed
stud group anchorages. This program was initiated in response to new
provisions introduced into the 2002 ACI 318 Building Code. The test pro
gram examined headed stud connections loaded toward a free edge, a
free edge near a corner, parallel to one free edge, parallel to two free
edges, away from a free edge, and in-the-field of a member, such that
edge distance was not a factor. The information reported herein address
es the steel capacity failure mode. Test data were obtained when the
shear force was directed away from a free edge, in-the-field testing, and
from other edge distance tests where steel failure governed the capacity.
REFERENCE: Anderson, Neal S., and Meinheit, Donald F., Design
Criteria for Headed Stud Groups in Shear: Part 1
Steel Capacity
and Back Edge Effects, PCI JOURNAL, V. 45, No. 5, SeptemberOctober 2000, pp. 46-75.
REFERENCE: Peterman, Robert J., Ramirez, Julio A., and Olek,

Jan, Influence of Flexure-Shear Cracking on Strand Development
Length in Prestressed Concrete Members, PCI JOURNAL, V. 45,
No. 5, September-October 2000, pp. 76-94.
KEYWORDS: box girders; beams; camber; development length;
design (structural); flexural strength; high strength concrete; high
performance concrete; precast concrete; prestressed concrete; pre
stress losses; prestressing steel; research; transfer length.
ABSTRACT: Presents the results of an investigation on the transfer
and development lengths required for Grade 270, 0.6 in. (15.2 mm)
diameter prestressing strand spaced at 2 in. (51 mm) on center in high
performance concrete prestressed box girders. Three 411 in. (10400
mm) long box girders with composite topping slabs were tested. The
specified compressive strength of the girder concrete is 10,000 psi (69
MPa). The AASHTO/ACI formulas overestimate the transfer length of
the girders by 18 percent and the development length by 53 percent.
The average pullout strength attained in strand pullout tests is 48 kips
(215 kN). The average strand slip measure right after stress transfer is
0.06 in. (1.52 mm).
REFERENCE: Shing, P. Benson, Cooke, Daniel E., Frangopol, Dan
M., Leonard, Mark A., McMullen, Michael L., and Hutter, Werner,
Strand Development and Transfer Length Tests on High Perfor
mance Concrete Box Girders, PCI JOURNAL, V. 45, No. 5,
KEYWORDS: building codes; codes; design (structural); Internation

al Building Code; model codes; NEHRP Provisions; precast con
crete; prestressed concrete; seismic-force-resisting system.
ABSTRACT: This is the second in a series of three papers dis
cussing significant modifications expected to be included in the 2000
NEHRP Provisions, dealing with the design of precast, prestressed
concrete seismic-force-resisting systems. These modifications are
expected to be part of the 2003 International Building Code.
REFERENCE: Hawkins, Neil M., and Ghosh, S. K., Proposed Revisions
to 1997 NEHRP Recommended Provisions for Seismic Regulations for
Precast Concrete Structures: Part 2 Seismic-Force-Resisting Systems,
PCI JOURNAL, V. 45, No.5, September-October 2000, pp. 34-44.
KEYWORDS: beams; carbon fiber strands; CFCC strands; CFRP

strands; development length; design (structural); double-T beams;
precast concrete; prestressed concrete; reinforcement; research;
transfer length.
ABSTRACT: Presents the results of an experimental investigation
addressing the variations in transfer lengths in two types of carbon fiber
reinforced (CFRP) prestressing strands. In all, eight double-T (DT) pre
stressed girders were prestressed with carbon fiber composite cable
(CFCC) strands. It is shown that the calculated and measured transfer
lengths are in close agreement with those obtained by others in the
case of Leadline tendons only. It is also noted that the level of release
of prestress has no significant effect on transfer length.
REFERENCE: Grace, Nabil F., Transfer Length of CFRP/CFCC
Strands for Double-T Girders, PCI JOURNAL, V. 45, No. 5,
131
MAIL TO:
PROFESSIONAL
PRECAST/PRESTRESSED CONCRETE INSTITUTE
MEMBERSHIP
209 WEST JACKSON BOULEVARD
APPLICATION
CHICAGO, ILLINOIS 60606
IPRECAST/PRESTRESSED CONCRETE INSTITUTE
PHONE: 312/786-0300
FAX: 312/786-0353
Membership Defined: Professional Member: Any person who is a registered architect or engineer, or any person who is deemed
by the Board of Directors to have abilities or credentials equal to or above the registration qualifications.
NAME OF INDIVIDUAL APPLYING FOR MEMBERSHIP
ADDRESS FOR RECEIPT OF MAIL
CITY, TOWN OR POST OFFICE, STATE AND ZIP CODE
PHONE
FAX
E-MAIL
istered
jineer
WEB PAGE
I HOLD REGISTRATION NUMBER
OF THE STATE OF_________________________
COMPANY AFFILIATION
POSITION
istered
hitect
COMPANY ADDRESS (IF OTHER THAN ABOVE) NUMBER
CITY, TOWN OR POST OFFICE, RO. NUMBER IF APPLICABLE
NATURE OF COMPANYS BUSINESS AND/OR PRODUCTS
IMPORTANT:
Please read definitions of membership classifications and eligibility on reverse side.
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 (SEE OVER)
PLEASE
INITIAL:__________
NOTE: PROFESSIONAL FEES ARE $105 ANNUALLY, ($125 FOREIGN). If this application is submitted April through August, enclose $105
($125 Foreign). If submitted September through March, enclose $75, ($95 Foreign). PAYMENT IN U.S. DOLLARS DRAWN ON A U.S. BANK
MUST ACCOMPANY THIS APPLICATION.
Contributions or gifts to PCI are not deductible as charitable contributions for federal income tax purposes.
However, dues payments are deductible by members as an ordinary and necessary business expense.
VISA
[[)]
Card #________________________________
AMOUNT ENCLOSED
Exp. Date_________________________________
SIGNATURE___________________________________________________ DATE___________________________
MEMBER #
Note application for Professional Intern Membership on other side.

SHAPING THE FUTURE WITH THE SOLUTION OF CHOICE
D APPLICATION FOR
PROFESSIONAL INTERN
MEMBERSHIP
PRESTRESSEO
CONCRETE INSTITUTE
BYLAWS EXCERPT
Membership Defined: Professional Intern Member: Any person who has

graduated in architecture or civil engineering with a Bachelors Degree, and has not
yet attained professional registration.
NAME OF INDIVIDUAL APPLYING FOR MEMBERSHIP
ADDRESS FOR RECEIPT OF MAIL
CITY, TOWN OR POST OFFICE, STATE AND ZIP CODE
PHONE
FAX
E-MAIL
WEB PAGE
I DO NOT HOLD PROFESSIONAL REGISTRATION

UNIVERSITY
UNDERGRADUATE DEGREE
YEAR
COMPANY AFFILIATION
POSITION
COMPANY ADDRESS (IF OTHER THAN ABOVE) NUMBER AND STREET
CITY, TOWN OR POST OFFICE, RD. NUMBER IF APPLICABLE
NATURE OF COMPANYS BUSINESS AND/OR PRODUCTS
IMPORTANT: Please read definitions of membership classifications and eligibility.

THIS CERTIFIES THAT TIlE 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. I FURTHER UNDERSTAND THAT IF PROFESSIONAL
REGISTRATION IS NOT ATTAINED WITHIN SIX (6) YEARS OF GRADUATION, THIS
MEMBERSHIP WILL BE TERMINATED AND APPLICANT WILL BE ELIGIBLE FOR AFFILIATE
MEMBERSHIP.
PLEASE
INITIAL:
NOTE: PROFESSIONAL INTERN MEMBER FEES ARE $45 ANNUALLY, ($70 FOREIGN). If this application is sub
mitted April through August, enclose $45, ($70 Foreign). If submitted September through March, enclose $30,
($55 Foreign). PAYMENT IN U.S. DOLLARS DRAWN ON A U.S. BANK MUST ACCOMPANY THIS APPLICATION.
Contributions or gifts to PCI are not deductible as charitable contributions for federal income tax purposes.
However, fee payments are deductible by members as an ordinary and necessary business expense.
Card
AMOUNT ENCLOSED
#_
Exp. Date
SIGNATURE
MAIL TO:
PRECAST/PRESTRESSED CONCRETE INSTITUTE
209 WEST JACKSON BOULEVARD
CHICAGO, ILLINOIS 60606
PHONE: 312/786-0300 FAX: 312/786-0353
ARTICLE Ill
MEMBERSHIP
Section 1. GENERAL QUALIFICATIONS. Any reputable
individual or firm engaged or interested in the Plant-Produced
Precast and Prestressed Concrete Industry as a producer, sup
plier of materials and/or services, architect, engineer, contractor,
developer, technician, educator or student, shall be eligible to
apply for Membership in the Institute.
Any person or firm desiring to become a Member of this Institute
shall make application for Membership in writing, on the
appropriate form and addressed to the Board of Directors,
setting forth the qualifications of the applicant for Membership
in this Institute, signed by the individual applicant or an
executive officer of the organization so applying, and shall agree
that, as a condition to becoming a Member of this Institute,
applicant shall be bound by these Bylaws and by any and all
policies and procedures adopted by the Board of Directors.
Any person who is employed on a Continuous basis by, or has
a financial interest in a non-member firm which is eligible for
Producer or Associate Membership shall not be eligible for
Membership in the Institute as an individual Professional,
Professional Intern, Affiliate or Student Member.
Any changes in an individual Members employment or pro
fessional status incompatible with the Membership
requirements of the Institute shall invalidate the Members
Membership unless otherwise determined by the Board of
Directors.
Section 2. CLASSES. Membership in this Institute shall be in
one of the following classes:
a. Producer Member: A firm or a division of a firm with a U.S.
manufacturing facility regularly engaged in the production of
Plant-Produced Precast and Prestressed Concrete.
b. Foreign Producer Member: A firm or a division of a firm with
a manufacturing facility outside of the U.S. regularly engaged
in the production of Plant-Produced Precast and Prestressed
Concrete.
c. Associate Member: Any individual or firm engaged in sup
plying goods, services, and information that are used by the
industry in the development or production of precast and
prestressed concrete products.
d. Professional Member: Any person who is a registered
architect or engineer, or any person who is deemed by the
Board of Directors to have abilities or credentials equal to or
above the registration qualifications.
e. Professional Intern Member: Any person who has
graduated in architecture or civil engineering with a Bachelors
Degree, and has not yet attained professional registration.
f. Affiliate Member: Any person, who has an interest in the
Plant-Produced Precast and Prestressed Concrete Industry and
does not conform to the requirements for other classes of
Membership.
g. Student Member: Any person who is enrolled in any
recognized institution of higher learning.
h. Lifetime and Honorary Membership may be conferred upon
individual Members of the Institute at such times and under
such terms as determined by the Board of Directors. Lifetime
Membership shall accompany the PCI Medal of Honor Award.
i. Allied Organization Member: City, state, regional or any other
organization in the Plant-Produced Precast and Prestressed
Concrete Industry having and maintaining purposes and ob
jectives similar to those of this Institute may become Allied
Organization Members upon meeting and maintaining such
standards as set forth in the PCI Policies and Procedures
Manual. Producer, Associate or Professional Members of such
organizations who are not Members of the Institute shall not
be extended the privileges of Membership in the Institute.
j. Other categories of Membership may be established by the
Board of Directors.
four submission of thIs applIcation is greatly appreciated.
The PCI Board of Directors and entire membership look for
ward to welcoming you as a particIpant In the PCI.
flTP
6?. /ztL
MEMBER #
Note application for Professional Membership on other side.
Thomas B. Battles
President
do1-tk-
&t
GFRC
RECOMMENDED
PRACTICE
THIRD EDITION
A Recommended
Practice Publication
at Your Fingertips,
the Knowledge and
Experience of an Industry!
his 8i1 x 11 softcover publication is
filled with 99 pages of information
concerning the planning, preparation
of specifications, design, execution and
supervision of the manufacture and installa
tion of glass fiber reinforced concrete (GFRC)
panels.
Part I
In this section there are 10 chapters and 12
appendices covering the following topics: (1)
Introduction, (2) Project Responsibilities, (3)
Materials, (4) Physical Properties, (5) De
sign, (6) Tolerances, (7) Manufacturing, (8)
Quality Control, (9) Loading and Delivery,
and (10) Installation.
Part II
This section of the publication contains a 17

page Guide Specification for Glass Fiber
Reinforced Concrete Panels which may be
used for development of an office master spec
ification or the specifications for a particular
project using glass fiber reinforced concrete
panels.
Data presented covers portland cement based
composites manufactured by the spray-up pro
cess using special alkali resistant glass fibers
that are chopped and sprayed onto a form with
an appropriate cement/aggregate slurry.
The primary emphasis of this report is on
thin-walled architectural panels made of glass
fiber reinforced concrete by the spray-up pro
cess under controlled factory conditions. These
cladding panels are capable of accepting and
transferring wind and self-weight and their
own inertial seismic loads to the buildings
load resisting system, but are not considered
as vertical loadbearing components or as part
of the lateral system.
PCI members
ORDER TODAY
Nonmembers
$15
$30
lease send me
copies of the Recommended
Practice for Glass Fiber Reinforced Concrete
(MNL 128-93). Payment must accompany order.
NAME
TITPIs,c
A nnnncc
Crrv_________________ STATE
E
CHECK
LI
ZIP
VISA/MASTERCARD
ExPiuTIoN DATE
Payment must be in U.S. dollars and drawn on a U.S. bank.

Illinois residents add 8.75% sales tax.
PrecastlPrestressed Concrete Institute

209 W. Jackson Blvd., Chicago, IL 60606
Phone: (312) 786-0300 Fax: (312) 786-0353
INDUSTRY NEWS
LETTERS TO THE EDITOR

Kudos for july-August
JOURNAL Issue
The July-August Special Convention
Issue of the PCI JOURNAL was par
ticularly outstanding because it con
tained something for everybody. A few
examples:
Precast modular construction
This efficient, versatile and cost ef
fective method of building struc
tures should become the dominant
method in the construction of hous
ing, prisons, schools and many other
types of structures. (See article on
Precast Concrete Modules Speed
School Construction.)
As
Rehabilitation of buildings
our urban infrastructure ages, the fa
cades and interiors of buildings will
need renovation. There is no better
material than architectural precast
concrete (in its many variations) to
use in renovating these old struc
tures. (See article on Brick-Faced
Precast Concrete Panels Help Reha
bilitate Allen Theatre)
It is
High performance concrete
becoming increasingly apparent that
a solution is needed to lower the
high cost of maintenance and re
placement of our nations bridges
and support structures. High perfor
mance concrete, in the form of pre
cast and prestressed concrete, is an
obvious means to decrease costs by
increasing the durability and
longevity of bridges, while at the
same time producing a more effi
cient bridge structure. (See article
on High Performance Concrete Ex
tends Life of Charenton Canal
Bridge and Analysis and Struc
tural Benefits of High Performance
Concrete for Pretensioned Bridge
Girders.)
Precast and
Urban transportation
PRECAST, INC.
COMPLETE DESIGN & DETAILING
ARCHITECTURAL
STRUCTURAL
CAST STONE
(307) 733-2872
P.O. BOX 1185
140 E. BROADWAY
JACKSON, WY 83001
(817)426-4411
201 W. ELLISON
SUITE 202
BURLESON, TX 76028
prestressed concrete plays a signifi

cant role in congested cities which
need a major overhaul in their urban
transportation system as is evi
denced in Bostons Big Dig pro
ject. (See article on Central
Artery/Tunnel Project: Innovative
Use of Precast Segmental Technol
ogy.)
Architectural precast concrete
Stone veneer-faced precast concrete
is yet another effective variation of
architectural precast concrete. The
article, Stone Veneer-Faced Pre
cast Concrete Panels provides
state-of-the-art information on the
many applications of this versatile
product.
Fire resis
Problems & Solutions
tance and sound attenuation of pre
cast/prestressed concrete products
are important issues to architects,
engineers and producers (p. 108).
John Baxter
Baltimore, Maryland
(505) 260-1823
5206 B CONSTITUTION NE
ALBUQUERQUE, NM 87110
High Performance Concrete

What I liked the most about the arti
cle (High Performance Concrete Ex
tends Life of Charenton Canal Bridge,
July-August PCI JOURNAL) is that it
not only presented the structural bene
fits in using high performance concrete
but it also discussed the details and
precautions needed in producing this
material.
Henry Smith
New Orleans, Louisiana
Modular Construction
The article on the design-construc
tion of the Josiah Quincy Upper
School (Precast Concrete Modules
Speed School Construction, July-Au
gust PCI JOURNAL) provided valu
able information on the structural de
tails and economic advantages in using
precast modular construction. This
construction method has great potential
not only for schools but many other
135
CALL FOR ENTRIES

ENGINEERING STUDENT DESIGN COMPETITION
The Student Education Committee is inviting entries from engineering

students to participate in a new Engineering Student Design Competition
for the year 2001. With the help of local PCI Producer Members, students
will construct and test 6 x 12 in. x 15 ft (152 x 305 mm x 4.6 m) precast
concrete beams. The awards program, sponsored by Sika Corporation, will
include cash prizes for the most efficient design, highest load capacity,
best report and other categories. Applications are due by December 31,
2000 and results by April 1, 2001 at PCI headquarters. PCI Producer
Members are urged to encourage their local engineering schools to partici
pate in this program. For additional information and an application, con
tact PCI Research Director, Paul Johal, at (312) 786-0300.
CALL FOR NOMINATIONS

DISTINGUISHED EDUCATOR AWARD
The PCI Student Education Committee (Alvin C. Ericson, chairman) in

vites nominations from PCI members for PCIs annual Distinguished Edu
cator 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 10, 2001. For
nomination forms and additional information, contact PCIs Research Di
rector, Paul Johal, at (312) 786-0300.
types of applications. Congratulations

to the authors for a very fine article.
John Alexander
TECHNICAL ACTIVITIES
COMMITTEE NEWS
The Technical Activities Committee
(TAC) (C. Douglas Sutton, chairman)
met Tuesday afternoon, September 26,
at the PCI Convention in Orlando,
Florida. The following is a summary
of the topics discussed and actions
taken:
Walter Korkosz of The Consulting
Engineers Group (Texas) has been
appointed chairman of the Parking
Structures Committee. He replaces
Ted Wolfsthal, who has been chair
man of this committee for the last
several years.
Don Theobald of Gulf Coast Pre
stress Inc., has been appointed
chairman of the Prestressed Con
crete Piling Committee. He replaces
Bill Wieners who has chaired this
committee during the last few years.
The first part of the PCI-sponsored
investigation on Headed Studs,
being carried out by Wiss, Janney,
Elstner Associates, has been ap
proved for publication and appears
in this current issue of the PCI
JOURNAL. The report is titled De
sign Criteria for Headed Stud
Groups in Shear: Part 1 Steel Ca
pacity and Back Edge Effects.
The Bridges Committee (Roy
Eriksson, chairman) held a full-day
meeting. Among the topics dis
cussed were:
Extending span ranges of precast,
prestressed concrete bridges.
Further collaboration with the
regarding high performance con
crete, strand bond development
length, increasing speed of erec
tion and other advanced technolo
gies.
Completing Volume 2 of the PCI
Bridge Design Manual.
The PCI Connections Committee
(Jagdish C. Nijhawan, chairman)
reviewed the editorial subcommit
tees comments on the Standard
Connections Manual from both a
technical and format standpoint. A
Precast concrete beams are a key feature in the rehabilitation of a historic

viaduct in Canton, Massachusetts. Northeast Concrete ProdLicts, Plainville,
Massachusetts, provided the precast concrete beams for the historic viaduct. The
structure can now carry high-speed rail traffic it has been unable to do since it
was originally built in 1835. Project engineers chose precast concrete for its
structural soundness and also because precast concrete allowed the structure to
be built in phases so that rail traffic was not interrupted. Recently, Northeast
Concrete won national awards from the Associated Building Contractors and the
Portland Cement Association for its work on the viaduct.
136
PCI JOURNAL
revised text is to be completed this

winter with plans to formally submit
the Connections Manual to TAC in
the middle of 2001.
The High Performance Concrete
Committee (Richard A. Miller,
chairman) reviewed TACs com
ments on its report, High Perfor
Showcase
mance
Concrete
Bridges. An updated version of the
report is currently being worked
upon by Henry Russell. The final
submission of the report is expected
early next year for publication in the
PCI JOURNAL.
The Prestressed Concrete Piling
Committee (Don Theobald, chair
man) is developing a chapter on
piles for the PCI Bridge Design
Manual as well as planning the de
velopment of a Pile Design Manual.
The Prestressed Concrete Poles
Committee (Fouad H. Fouad,
chairman) had a productive meeting
discussing a wide range of topics.
Currently, the committee is review
ing the report, Guide for Handling
and Erection of Prestressed Con
crete Poles, which it expects to
submit to TAC next year. The com
mittee is also working on two other
documents:
Guide Specification for Pre
stressed Concrete Poles for Streetlighting Applications
Specification Guide for Pre
stressed Concrete Poles
The Seismic Committee (Ned M.
Cleland, chairman) is working on
the development of a chapter on
seismic design considerations and
connections for the next sixth edi
tion of the PCI Design Handbook.
The Hollow Core Slab Producers
Committee (Robert McCormack
and Jeffery Butler, co-chairmen)
had a joint meeting with the Interna
tional Prestressed Hollow Core As
sociation. A major topic of discus
sion centered on fire ratings for
restrained and unrestrained hollowcore slab floors.
Kenneth Baur reported on the ini
tiative being made by CPCI (Cana
dian Precast/Prestressed Concrete
Institute) to develop a standard pro
gram to integrate software for the
precast concrete industry in North
America. An initial meeting, titled
-
Creating a Digital Building Product Model for the Precast, Pre

stressed Concrete Industry, was
held September 8 and 9 in Toronto,
Ontario. The purpose of the meet
ing was to outline the steps and
cost needed to undertake such a
program.
The manual, Precast Prestressed
Concrete Parking Structures: Rec
ommended Practice for Design and
Construction is now available on a
CD ROM.
The Professional Member Commit
tee (Donald C. Raths, chairman)
has developed an attractive promo
tional brochure on the benefits of
being a PCI Professional Member.
R&D COMMITTEE NEWS

The R&D Committee (Thomas J.
DArcy, chairman) met during the
PCI Annual Convention/HPC Sympo
sium in Orlando, Florida, to discuss
the status of various research projects.
These projects included many fellow
ships and research projects on com
posite behavior of insulated precast
concrete wall panels, design criteria
for headed stud groups, strand devel
opment length, and discussion of a
high priority project on development
of design factors for precast concrete
elements connected with dry joints ex
pected to be initiated next year.
Additional research funded and
monitored by the R&D Committee in
cludes projects on the behavior of
spandrel beams under torsion, shear
key connections for adjacent box
girder bridges, precast, prestressed
shear walls with large openings, and
studies of parking structures associ
ated with their performance during re
cent earthquakes.
The Ad Hoc Committee on ATLSS
and PRESSS (Mario J. Bertolini,
chairman) reviewed the status of vari
ous major projects being carried out
under the ATLSS (Advanced Technol
ogy for Large Structural Systems),
PRESSS (Precast Seismic Structural
Systems) and NIST (National Institute
for Standards and Technology) re
search programs.
Several research programs are
four research fel
planned for 2001
lowships; a new project on develop-
New Appointments to
PCI Committees
The following individuals
have recently accepted appoint
ments to PCI committees. We
appreciate their interest and vol
untary participation.
Ad Hoc Committee on
ATLSS and PRESSS
Neil M. Hawkins
University of Illinois
Urbana, Illinois
PCI Student Education

Committee
Mark Fusani
Springboro, Ohio
Plant Safety Committee
William Tomlin
Finfrock Industries, Inc.
Orlando, Florida
Precast Sandwich Wall Panels
Committee
Robert T. Elliott
H. Wilden & Associates, Inc.
Allentown, Pennsylvania
J. Scott Heuvel
Iowa Prestressed Concrete, Inc.
Des Moines, Iowa
ment of design factors for dry precast

concrete connections; a discretionary
fund to investigate catastrophic events
(such as an earthquake or bridge col
lapse) or other projects of extreme
emergency in which precast or pre
stressed concrete is involved; and sup
port and involvement as industry advi
sors in the ATLSS and PRESSS
programs. Also, a budget of $100,000
has been proposed to initiate research
to support industry codes of practice.
Phase II of the project Design Crite
ria for Headed Stud Groups, includ
ing combined shear and tension, is ex
pected to be initiated this year and
completed by the end of next year.
The committee also approved an in
crease in the Daniel P. Jenny Research
137
Spancrete Wraps Up Work on Milwaukee Marriott West Hotel

Spancrete Industries, Inc., Waukesha, Wisconsin, is currently
completing work on the Milwaukee Marriott West Hotel.
Overall, Spancrete has provided 180,000 sq ft (16722 m) of
hollow-core floor slabs and approximately 30,000 sq ft (2787
) of additional products for this prominent project.
2
m
Spancretes precast hollow-core was the best material for this
project becaLise it met design requirements. Hollow-core
slabs provide cohesion and stability within the entire
structure. Hollow-core slabs were used for the five above
Fellowship award amount to $15,000

each for a total of $60,000 in 2001.
STUDENT EDUCATION SESSION
ORGANIZED AT PCI CONVENTION
The Student Education Committee
(Alvin C. Ericson, chairman) offered
a very exciting session at the PCI Con
vention. The session was attended by
several civil engineering and architec
tural students and professors from
local universities.
The session included presentations
on various structural and architectural
applications of precast concrete, career
opportunities at PCI, student design
competition programs, and the impor
tance of student education in the fu
ture growth of the industry. As usual,
past committee chairman Charles E.
(Budd) Hilgeman played a major role
in the success of this program. The
students had the opportunity to visit
the Exhibit Hall for hands-on experi
ence with materials and technology
used in the precast, prestressed con
crete industry.
The Student Education Committee
has proposed several programs for the
next year. Among these programs are
the distribution of PCI Architectural
138
ground levels as well as the roof of the hotel. While hollowcore slabs provide structural elements to this new Marriott,
Spancretes architectural wall panels and spandrels add to
the exterior aesthetics, blending well with the glass and
exterior insulated finish system faade. In addition to the
structural and visual functions provided by precast concrete,
the material also exceeds fire codes ensuring the safety of the
hotel guests and employees. The Milwaukee Marriott West is
scheduled to open in March 2001.
and Design Handbooks, PCI Hollowcore Manuals, distinguished educator

award, PCIs continued support of the
Association of Collegiate Schools of
Architecture program, student design
competitions in architecture and engi
neering, and development of an audio
and visual library for students and oth
ers.
PRESSS SESSIONS AT PCI AND
ACI CONVENTIONS
A special PRESSS session entitled
Developing Innovative Solutions for
Precast, Frame and Wall Systems
was conducted September 24 at the
PCI Convention. It included seven
presentations on various PRESSS re
search topics. Two additional sessions,
prepared and organized by the PCI
staff and sponsored by the ACI Com
mittee 550 on Precast Concrete, were
included in the ACI Convention in
Toronto, Canada, on October 15-20.
These sessions entitled, Innovative
Precast Concrete Systems
New
Concepts for the New Century, Part I
and Part II, included several PRESSS
presentations on design, construction,
testing, evaluation and code related is
sues to discuss some of the significant
changes in the new IBC 2000 that are

likely to affect precast concrete pro
ducers in most seismic zones.
Willis Construction Precast

Concrete Components Used
in Sheraton Grand Hotel
GFRC panels produced by Willis
Construction, San Juan Bautista, Cali
fornia, are being used as the exterior
facade for the new Sheraton Grand
Hotel in Sacramento, California. The
architects of the project chose precast
concrete as the exterior material be
cause it will provide an attractive
faade that fits in with the existing
buildings. The first four floors of the
building will feature rusticated GFRC
panels in a darker tone than the tower.
About 80,000 sq ft (7432 m
) of pre
2
cast concrete components will be used
for the hotel which is scheduled for
completion in 2001.
PCIs Freedman to Speak at

John Earley Symposium
Sidney Freedman, architectural di
rector of PCI, will speak at the Fourth
Biennial Symposium on the Historic
Development of Metropolitan WashPCI JOURNAL
CONCRETE
REINFORCING STRAND
A World Of Differpnce
ecists between
I
Start
Finish...
Lull
For Prestressed and

Post-Tensioned Applications
In stress relieved and low relaxation grades.
FOR COMPLETE INFORMATION CALL:
(800)683-9473 or Fax (216) 292-4444.
AMERICAN SPRING WIRE CORR
LEAP brir s a World of
iowledge
to its cust mers in the E tions
of Precast Prestressed .oncrete
--
--
--
ington, D.C., at the University of Maryland, College Park,

Maryland. Also speaking at the symposium will be many
talented architects, engineers and researchers who work in
the field. The symposium will be held March 31 to April 1,
2001. For more information, call Jere Gibber at (703) 7686987 or send e-mail to jnibber@aol.com.
Metromont Supplies Precast Products for

Office Building
Metromont Prestress Company, Greenville, South Car
olina, has produced 3120 precast concrete components for
the Hearst Tower and Parking Deck in Charlotte, North
Carolina. At 50 stories high, the Hearst Tower is Charlottes
second tallest building. The building consists of 345,000 sq
ft (32000 m
) of architectural precast concrete and is de
2
signed to complement surrounding buildings.
--
--
Parking Struc es
:jlitjes
Correctional
Stadiums anc renas
Bridges
Food Process and
Industrial Fa ties
Put knc viedge, exp rience,

technc gy and inn )vation
to work n your ne). project.
Call LEAR
SOCIATES
ematlonal, Inc.
Iverson Joins Pomeroy

James (Jim) K. Iverson has joined the Pomeroy Corpo
ration, Petaluma, California, as chief engineer. Pomeroy is a
major precast producer in Northern California involved in
many large-scale marine and inland structures. Prior to this
position, he was project manager for Carter & Burgess, Inc.,
where he was responsible for the seismic retrofitting of the
San Mateo-Hayward Bridge in San Francisco Bay. He has
had nearly 40 years of experience in teaching, design, and
CONS(
ING STRUCTURAL E
11602 N. 5lth S
te. 100
Tampa, FL
517
Phone 813-9 i870
Fax 81 3-980 100
7475 D
Den
Phon
Fax
;INEERS
in St., Ste 205
r, CO 80221
03-426-0222
3-426-0299
139
construction of precast and prestressed

concrete structures throughout the
United States. Earlier he obtained his
BS, MS and Ph.D. degrees from
Michigan State University. Active in
PCI Committee work, he has been
chairman and a member of the Seis
mic, Bridges and Building Code com
mittees. He also served as director on
the PCI Board of Directors and is the
author of several technical articles in
the PCI JOURNAL.
1=
Quarries Group, Inc.
Tin
Dynamic Color
Solutions, Inc.
offers a wide
selection of
pigments in both dry and liquid
slurry form for the addition to any
concrete mix. Custom blending, color
matching and packaging is available.
Web Site: www.dynamiccolorsolutions.com
ast
Fister Quarries Group
High Concrete Provides
of aggregates of varying
color, size & texture. Design
flexibility is possible by varying aggregate
color & size, matrix color, finishing processes,
& depth of exposure. Complete new aggregate
color guide available.
WIN System wet
abrasive sand blasthig
production rates and superior dust

suppression versus ordinary water
ring systems. The technologically
advanced, patented design works with
your existing equipment. NoMes operate
at low water pressure levels and use
only 3/4 to 6 quarts of water per minute.
Precast Concrete for 250th

Parking Structure
The first of four parking structures
being built on the Merrill Lynch cor
porate campus in Hopewell, New Jer
sey, marks the 250th parking structure
for which High Concrete Structures,
Inc., Denver, Pennsylvania, will pro
vide precast concrete components.
Work on the four parking structures is
expected to continue through spring
2001. Incidentally, the first parking
structure project that High Concrete
was involved with was the Duke

Street parking structure constructed in
1969 in Lancaster, Pennsylvania.
Flexicore Building System

Shortens Residence Hall
Construction Time
Flexicore Systems, Inc., Dayton,
Ohio, provided an innovative precast
concrete system for quality fast-track
construction of a new residence hall
on the campus of Rose-Hulman Insti
tute of Technology, Terre Haute, Indi
ana. The Flexicore structural system
provides components more durable
and quicker to erect than steel, cast-inplace or masonry construction, while
providing a fire-resistant, sound and
thermally insulated structure. The sys
tem also eliminates disruption to the
construction area and minimizes space
requirements by factory-casting offsite, while eliminating production
downtime due to cold weather. In only
11 weeks, Flexicore was able to erect
a 56,000 sq ft (5200 m
), four-story,
2
stately campus home for more than
800.542.7393 (USA) 800.626.4099 (canada)

630.424.6209 (fax)
Website: www.fisterquarries.com
QUALITY IS IN THE DETAILS

ROSS BRYAN ASSOCIATES, INC.
1025 16TH AVE. SOUTH, NASHVILLE, TENNESSEE 37212
Phone (615) 329-1300
Fax (615) 327-4446
GREGORY DEVELOPMENT SERVICES, INC.
Marketing and Management Consulting for the Precast Concrete mdustry

6500 W. 166th Street, Tinley Park, Illinois 60477
Office: (708) 429-1444
Fax: (708) 429-9222
140
PCI JOURNAL
200 students on time and well within

budget.
Shockey Precast Receives

Two Contracts
Shockey Precast Group, Winch
ester, Virginia, has been awarded two
contracts to supply precast concrete
components for two Maryland-area
sports facilities. Shockey will supply
precast concrete for the new 420,000
sq ft (39000 m
) Comcast Center at
2
the University of Maryland, College
Park, Maryland, that is to be com
pleted in early 2002. Shockey will
also provide precast products for the
Towson University Stadium & Sports
Complex Phase II in Towson, Mary
land. Shockey will deliver more than
3000 double and triple seating units
for this project scheduled for comple
tion in 2001.
Precast Concrete Lends

Strength to New A&P Market
Strescon Industries, Inc., Balti
more, Maryland, provided over
) of precast
2
23,000 sq ft (2140 m
concrete wall panels for the new
A&P Supermarket which recently
opened in Mansfield, New Jersey.

The precast concrete exterior wall is
not only aesthetically pleasing, but
also provides necessary structural
support. The architect also chose pre
cast concrete because using precast
concrete wall panels does not pro
duce the block and mortar residue
that masonry does, which helps pro
vide for a clean and efficient work
site.
Spancrete of Illinois
Expands Manufacturing
Capacity
Spancrete of Illinois, Inc., Crys
tal Lake, Illinois, is expanding and
upgrading its Crystal Lake facility.
This process includes the renova
tion of six 4 ft (1.2 m) outdoor
manufacturing beds as well as the
installation of a new GT 240 Flare
Out hollow-core extruder produced
by Spancrete Machinery Corpora
tion in Waukesha, Wisconsin. The
new machine will manufacture
Spancrete 4 and 8 ft (1.2 and 2.4
m) hollow-core floor slabs and wall
panels on the newly renovated pre
stressing beds.
NewcreteTM Supplies Precast

Components for Parking
Structure
NewcreteTM Products, a division
of New Enterprise Stone and Lime
Co., Inc., New Enterprise, Pennsyl
vania, provided precast concrete
components for the Newark Interna
tional Airports new parking struc
ture. Over 2000 precast/prestressed
concrete components were used in
the structure, including double tees,
inverted tees, L-beams, grade span
drels, helix spandrels, bridge span
drels and deck slabs, and other
products.
Coreslab Acquires Structural

Precast Products Company
Coreslab Structures (Ark), Inc.,
Conway, Arkansas, has acquired the
production facility of Structural Pre
cast Products Company LLC from
the Rogers Group, Inc., Nashville,
Tennessee. Coreslab will operate
this manufacturing facility as Coreslab Structures (Ark), Inc. Howell P.
Anderson will continue to be vice
president and general manager at
Coreslab Structures.
INTERESTED IN ADVERTISING IN THE PCI JOURNAL?
Request a media kit

To receive an advertising rate card and the latest issue of the PCI JOURNAL, please complete
the form below and send to: PCI, Attn: Susan Bowden, Editorial Assistant, 209 W. Jackson
Boulevard, Chicago Illinois 60606. You may also contact Ms. Bowden at 312 786-0300
(phone) or 312 786-0353 (fax) to request a rate card.
Please send me a PCI JOURNAL media kit.
Name
Company name
Address
City
Ad agencies In the space below, provide the name of your client and the city/state/country
where that client is located.
-
141
High Concrete Announces

New Contracts, Completes
Other Projects
High Concrete Structures, Inc., was
recently awarded several new con
tracts to produce precast concrete
components:
RCN Corporate Headquarters,
Five
Lawrenceville, New Jersey
parking structures totaling 395,000
).
2
sq ft (36700 m
State College Municipal Building,
High
State College, Pennsylvania
Concrete will produce 13,000 sq ft
) of architectural precast
2
(1200 m
concrete wall panels for this build
ing.
Ephrata Hospital, Ephrata, Pennsyl
),
2
vania
A 70,000 sq ft (6500 m
two-elevated level parking structure.
The Borgata Casino, Atlantic City,
A 1,200,000 sq ft
New Jersey
(111500 m
), six-elevated level
2
parking structure.
Metro Top Parking Garage, Woodbridge, New Jersey
A 180,000 sq
), four-level parking
2
ft (16700 m
structure.
Harrahs Casino Parking Garage,
A
Atlantic City, New Jersey
310,000 sq ft (28800 m
) parking
2
structure.
Wyoming Valley Health Care Sys
tem, Willces-Barre General Hospital
Parkade, Wilkes-Barre, Pennsylva
),
2
nia
A 65,000 sq ft (6040 m
three-elevated level parking struc
ture.
High Concrete recently completed
)
2
work on an 80,000 sq ft (7430 m
AIDS research facility for the National
Institute of Health, Bethesda, Maryland.
For the project, High Concrete provided
precast concrete architectural cladding.
The company also provided precast
concrete components for the expansion
of an existing parking structure and the
construction of two new parking struc
tures at the Fair Oaks Mall, Fairfax,
Virginia, as well as the construction of
Princeton Universitys new 186,000 sq
) parking structure.
2
ft (17280 m
PCI President Tom Battles (second from left) attended the annual meeting of
the Georgia/Carolinas PCI in August where he presented a certificate and
check for $10,000 to Clemson University for the Daniel P. Jenny Research
Fellowship entitled Preliminary Study of Precast Concrete in the Construction
of Tornado Shelters. Shown receiving the award is Dr. Scott D. Schiff,
Associate Professor of Civil Engineering at Clemson University. Also pictured
is David lmpson (left), the graduate student who will be working on the
research program, and Mike McConochie (right), Director of the
Georgia/Carolinas PCI.
PACS
PRESTRESSED &
ARCHITECTURAL
CONCRETE
SERVICES
Quality Drafting &

Engineering Services
PACS
3666 N. Miller Road #101

Scottsdale, AZ 85251
Phone: (480) 949-0005
Fax: (480) 949-0634
LPILE Plus 30
for WIndows
Response of piles and drilled shafts under lateral loads
GROUP 4.0
for Windows
Response of group of piles under combined loads
SHAFT 4.0 for Windows
Response of drilled shafts under axial loads
NEW!
APILE Plus ndows

Response of driven piles under axial loads
EnBeamC 3,rxiows I NEW!
Analysis and design of Prestressed Beams

in simple and continuous BRIDGE p.gj
CAD Autodrafting non-std. live load
Automatic comp. of restraining moments
AASHTO 97 + AREA Coopers E Railroad :ii:
A
Database
rders. box bean
ENSOFT INC
Ph. 512-458-1128 Fax 512-467-13194
Your best help for your next engineering projects
Download now yourfree software demos!
a.ensoftinc .com
R
R
CONSULTANTS
STRUCTURAL ENGINEERING AND MANAGEMENT
Offering Full Service to the Ever-Challenging Precast Concrete Industry

Contact:
142
Ram Ramadev, S.E.

2 Blackburn, Dove Canyon, Calif. 92679
949-858-4420 714-393-4499 FAX 949-888-1660
Chinese Academy of
Engineering Elects Tang
The Chinese Academy of Engineer
ing has elected Man-Chung Tang,
PCI JOURNAL
chairman of the board and technical

director of T.Y. Lin International,
San Francisco, California, as a for
eign member. Dr. Tang has world
wide experience in the design
and/or construction of more than
100 major bridges throughout the
world.
Currently, he is working on six
cable-stayed bridges worldwide
and is overseeing the design of the
San Francisco-Oakland Bay Bridge
East Span Replacement. The Chi
nese Academy of Engineering
elects foreign members biannually;
Dr. Tang is one of six foreign
members elected in the last two
years.
The Chinese Academy of Engi
neering is the highest honorary
and advisory organization for en
gineering and technology in China.
Members are elected for life and
are selected for their achievements
and contributions in engineering
fields.
Shockey Precast Fills Large

Order for Developer
More than 700 precast concrete
components, supplied by The
Shockey Precast Group, Fredericks
burg, Virginia, will provide important
support and design elements to the
McKinley/Piece and Rappahannock
Office Buildings in Tysons Center,
Virginia. Two parking structures, in
cluded in the project, will feature pre
cast concrete spandrels, columns and
column caps, wall panels, double tees,
beams and stairs. Construction is
scheduled to be completed in mid2001.
DAVID SEAGREN
(1959-2000)
David Seagren, quality assurance manager and chief
engineer for Charles Pankow Builders, Ltd., died August
6 in Long Beach, California, at the age of 40. During his
17 years of dedicated service with Pankow, he con
tributed enormously to the company and the engineering
profession. Mr. Seagren obtained his BS and MS degrees
in civil engineering from Purdue University in 1982 and
1983, respectively. At Pankow, Mr. Seagren was respon
sible for corporate research and development as well as quality assurance pro
grams. As research director, he was particularly involved in the development
of the precast hybrid moment resisting frame system which has now resulted
in the erection of a 39-story building in San Francisco (the tallest precast, pre
stressed concrete building in the United States). A long-time PCI Professional
Member, he participated in both ACT and PCI committee work. Dave was
tireless in his pursuit of excellence, said Tom Verti, president of Charles
Pankow Builders, Ltd. Always the consummate professional, he applied his
midwest work ethic to everything he did. With his mild manner and humble
nature, he purposefully camouflaged his many contributions and achieve
ments. But those who knew him and worked with him know well the impact
of the contributions of this talented, dedicated, and wonderful individual.
JOHN H. BASS, JR.

(1919-2000)
John H. Bass, Jr., who worked for Hamilton Form
Company from 1968 to 1991, died July 21 in Clifton,
Texas. He was 81. Born in Rosewood, North Carolina,
Mr. Bass served as a consultant to Hamilton Form after
serving 19 years as vice president and sales manager for
the company. Prior to his work with Hamilton Form,
Mr. Bass worked with North Carolina Concrete Prod
ucts (1955 1957), John A. Roebling & Sons Co. (19571962), Southern Prestressed Concrete Co. (1962 1964), and Valmont Man
ufacturing Co. (1964 1968). A World War II veteran who served under
General George Patton, Mr. Bass was known for his good sense of humor
and loved to tell stories. He never missed a PCI Convention where he made
countless friends while managing the Hamilton Form booth. John Bass will
be remembered fondly by many PCI members.
-
Nitterhouse Produces
Hollow-core Slabs for Hotel
Nitterhouse Concrete Products,
Inc., Chambersburg, Pennsylvania,
along with the Hampton Inn hotel
chain, recently celebrated the grand
opening of an 80-room facility lo
cated on Rt. 81 in Maugansville,
Maryland. Nitterhouse produced its
SpanDeck hollow-core slabs for the
structure. The design engineers chose
the material for its efficiency in con
struction and cost.
PSBearnH
AASHTO Standard & LRFD Specifications
US Customary and metric units
Bulb tees, I-beams, boxes, channels & slabs
MS Windows interface
Graphical plots: moment, stress & shear
On-line LRFD knowledge base
UI
(813) 989-3317
FREE
DEMO
www.eriktech.com
143
CNL Center
Architectural precast concrete panels, with a distinctive
salmon/beige granite color, provided a beautiful faade for this
14-story office building in Orlando, Florida
major challenge facing the architect was to design
the exterior faade of this high rise building,
which had to complement a neighboring struc
ture s architectural precast concrete skin and copper dome
while at the same time create a building with its own dis
tinctive character. The $24 million CNL Center is a 14story reinforced concrete framed building situated in Or
lando, Florida. The facility covers a floor area of 363,100 sq
), Founded in 1973, CNL is an industry leader
2
ft (33768 m
in real estate, finance and development.
The architects solution was to adopt a contemporary de
sign by using architectural precast concrete panels with a
salmon/beige granite color. Floor-to-ceiling glass windows
placed in precast concrete frames emphasize the buildings
long-term endurance and stability.
A dominating exterior feature of the building is the semi
circular tower within a tower that rises from the plaza
level along the north face of the building, rising above the
144
roof and culminating into an open, circular cupola. Visu

ally, this element is an effective focal point, especially at
dusk with the buildings dramatic night lighting.
The facility is a blend of traditional architecture and stateof-the-art technology. The center itself offers many ameni
ties including state-of-the-art, 24-hour security, on-site
banking, online concierge service, a full-service restaurant,
covered parking and a prime location.
Both the first and second floor common areas of the
tower offer access to and from the second floor air-condi
tioned skybridge which connects to a newly constructed
seven-level parking structure with approximately 1700 ve
hicular spaces.
The rotunda is capped with a sculptured ceiling. Escala
tors provide easy access from the lobby rotunda to the sec
ond floor, which features open balcony views into the ro
tunda and the plaza area.
A total of 1574 precast concrete panels, comprising a wall
PCI JOURNAL
area of 125,320 sq ft (11655 m

), were
2
used to clad the building. There were
two concrete mixes for the building, the
majority of which was the rose color
comprised of pink and salt and pepper
granites, white cement and dyes.
The precast concrete panels were
manufactured by Gate Precast Com
pany at their plant in Monroeville, Al
abama. In addition to the precast pan
els for the building, the precaster
furnished architectural precast covers
for the column bases of the parking
structure. The panels were shipped by
truck trailer to the project site, a dis
tance of about 500 miles (800 km).
Transportation and erection of the pan
els were the responsibility of the pre
caster. Gate used Precon Construction
of Tampa, Florida, to erect the panels.
Construction of the project began in
late 1998 with the foundations and
structural frame. After about seven
stories had been constructed, the pre
cast panels were erected during a fourmonth period. The entire building was
completed in November 1999.
When the project was bid, the archi
tect stated that Architectural precast
concrete provides a consistent manu
factured product that can include
forms, reveals and is quick to erect.
The product coloration and consis
tency had to be compatible with a
neighboring buildings palette. In ret
rospect, the successful completion of
this building demonstrated that the
above statement was spoken correctly.
The new CNL Center stands proudly
as a magnificent addition to the Or
lando skyline.
II liii
111111
111111
IIIII
llilimiti
Iii
111111
111111
(1I ii
IiI
I!I II. Ii!.I

1IIpII
j iiI
I
_-.
CREDITS
Owner: CNL Plaza Ltd., Orlando,
Florida (Vice President: Jim
Kersey)
Architect: HKS Inc., Dallas, Texas
(Project Manager: Jeff Vandersall,
AlA)
Structural Engineer: Brockette, Davis
& Drake, Dallas, Texas (Project
Manager: Robert Hill)
General Contractor: Brasfield & Gor
ne, Maitland, Florida
Precaster: Gate Precast Company,
Monroeville, Alabama
Photographer: Ben Tanner, Orlando,
Florida
145
50 YEARS OF PRECAST ENGINEERING & GROWING!

:
\
ASS OC;ATS
LEAP Associates International, Inc., specializing in large scale, high

profile, precast projects is seeking experienced Project Engineers,
Project Managers, Detailers, and Checkers due to expansion and rapid
growth in our Denver, CO and Tampa, FL offices. Candidates should
have a minimum of 2 years experience in precast concrete. AutoCAD
proficiency is required for project detailers, checkers and project
managers. Excellent benefit and compensation package including 401(k)
and long term disability. If you are seeking a challenging and rewarding
career working with the nationwide innovative leader providing High
Performance Engineering Services for High Performance Concrete, apply
in confidence to Craig Barrett, P0 Box 21298, Denver, CO 80221 or e
mail C.Barrett(LEAPDenver.com or Bryan Trimbath, P0 Box 16007,
Tampa, FL 33687 or e-mail B.Trimbath@LEAPTampa.com or visit our
website at www.LEAPTampa.com. EOE
IMMEDIATE CAREER
OPPORTUNITIES
inpIoyment
a professional employment service
serving the concrete building mate
rials and construction industry since
1973, has been retained by numer
ous North American prestress/pre
cast concrete producers. Engineer
ing managers, design engineers,
CAD operators, project managers,
estimators, sales reps, production
managers and supervisors, and
quality control positions are being
sought.
If you are interested, please reply in
confidence to:
United Employment Associates
Michael S. Stauffer
2030 Tilghman Street, Suite 201
Allentown, PA 18104
Tel.: (610) 437-5040
Fax: (610) 437-9650
All Fees Company Paid
STRUCTURAL ENGINEER
PROGRESSIVE FAMILY OF
PRECAST PLANTS
Located in Iowa, Illinois and
Michigan
Hiring for the following positions:
Minimum 2 years
Engineers
experience in design of precast
prestressed concrete
Minimum 2 years ex
Drafters
perience in detailing precast pre
stressed concrete
Minimum 5
Chief/Lead Drafters
years experience in detailing pre
cast prestressed concrete
We offer:
Competitive salary
Health insurance
401(K) benefits
Profit sharing
Paid vacation
Relocation assistance
Send or fax resume to:

Dr. Larbi Sennour
VP Sales and Engineering
205 S. Gear Ave.
West Burlington, IA 52655
Phone: (319) 752-8266
Fax: (319) 754-0027
146
Tpac
A Division of Kiewit Western Co.
Phoenix, Arizona
A leading producer of quality pre
castlprestressed concrete products
has the following positions available:
Construction Estimator mini
mum 5 yrs exp.
require
Design Engineer
BSCE or Structural Engineering,
3+ yrs precast/prestressed and/or
structural design exp. Prefer PE
or EIT.
Sales Representative mini
mum 5 yrs exp. in the precast/pre
stressed industry.
We offer: competitive salary, medi
cal, dental, life, L.T.D., vacation, sick
leave, 401(k), profit sharing and re
location assistance.
Substance abuse testing repre
sents a condition of employment.
Ohio based architectural and struc

tural precast concrete manufacture is
seeking a professional engineer with
expenence in designing architectural
and structural precast concrete mate
nals. 4 years experience, good com
munication and computer skills re
quired. Competitive salary and
benefits provided.
Please respond to:
Director Human Resources
Marietta Structures
P.O. Box 653
Marietta, OH 45750
or: gld @ mariettastructures.com

Personnel Coordinator
Tpac
P.O. Box 20128
Phoenix, AZ 85036-0128
602-262-1351
fax: 602-262-1375
EOE M/FN/H
PCI JOURNAL
Executive Director
DRAFTING
The Precast Prestressed Producers

of Illinois and Wisconsin (PPPI & W)
is searching for a full time Executive
Director located in the greater
Chicago area to develop and coordi
nate promotional and marketing pro
grams for the precast industry in Illi
nois and Wisconsin. Marketing
activities will be directed to archi
tects, engineers, educational institu
tions, builders and owners. The
ideal candidate will have an archi
tectural, engineering or related de
gree, some construction or related
industry experience, outstanding
communication and presentation
skills, PC proficiency and the ability
to represent the precast industry in
a professional manner. The suc
cessful candidate will be able to
work out of a home office and will be
responsible to a steering committee
comprised of producer members.
The position provides a competitive
compensation and benefit package,
including pension. Interested candi
dates should send resume and
salary requirements to:
Precast Prestressed Producers of
Illinois and Wisconsin
424 N. Fourth Street
Springfield, IL 62702
Attention Search Committee
EOE
Metromont Prestress Company is

seeking experienced Senior Project
Drafters and Detailers. Ideal candi
date possesses three to five years of
experience in structural drafting using
Auto-CAD (Release 12, 13, or 14).
Structural steel detailing experience is
transferable. Metromont Prestress
Company manufactures commercial
prestressed and precast concrete
products. We have openings in
Greenville, SC; Charlotte, NC; and
Dalton, GA. We offer excellent salary
and benefits, including comprehen
sive health plan, tuition reimburse
ment, and 401K with company match.
Metromont Prestress Company
P.O. 8ox2486
Greenville, SC 29602
Attn: Kevin McSloy
Fax: 864-295-5515
PCI SEEKS MANAGING EDITOR

The Precast/Prestressed Concrete
Institute is seeking a technical edi
tor to assume managing editor po
sition of PCI JOURNAL. A mini
mum of a BS degree in civil
engineering with emphasis on
concrete structures is required in
addition to writing and computer
Experience in the
skills.
precast/prestressed concrete in
dustry would be an asset. Excel
lent salary, based on experience,
plus attractive fringe benefits.
Respond in writing only with pro
fessional resume to PCI, 209 West
Jackson Boulevard, Suite 500,
Chicago, Illinois 60606.
e-r k1@rrrcTIcDm
Visit our website:

www.metromont-prestress.com
EOE
PROJECT ENGINEERS
CADD PROJECT MANAGERS
CADD DETAILERS
The Gate Construction Materials
Group is one of the leading produc
ers of architectural precast concrete
in the United States. Architectural
precast plants are located in Mon
roeville, AL; Oxford NC; Winchester,
KY; and Nashville, TN. Gate is cur
rently seeking experienced Project
Engineers, CADD Project Man
agers, and CADD Detailers for a
new precast engineering office to be
opened in Raleigh, North Carolina. If
you are seeking a challenging new
career opportunity with a progres
sive industry leader, please forward
a resume in confidence to:
Gate Concrete Products Company
PC Box 1604
Oxford, North Carolina 27565
Attention: Laurie Meador
Phone: 919-603-1633
Fax: 919-603-0580
E-mail: mquinlan@gatepetro.com
PROJECT MANAGERS
DRAFTING COORDINATORS
Metromont Prestress Company, a
leading commercial prestressed/pre
cast products manufacturer, has
openings for experienced Drafting
Coordinators. Responsibilities in
clude checking erection drawings
and shop cards, generating drawings
to meet manufacturing needs, and
coordinating with project team to as
sure schedule compliance. Positions
available in Greenville, SC; Charlotte,
NC; and Dalton, GA. Excellent salary
and benefits, including comprehen
sive health plan, tuition reimburse
ment, and 401K with match.
Well-established Midwestern pre

cast concrete plant has immediate
openings for project managers
preferably with experience in the
precast concrete industry. Good
salaries with a generous bonus and
profit sharing plans. Excellent med
ical insurance package that is fully
paid for by the company. Respond:
Mr. Ned Piccinini, at National Pre
cast, Inc., 30066 Little Mack, Ro
seville, Ml 48066. Phone: 81 0-2946430. Fax: 810-294-6527. email:
fibrahim @ national precast.com.

Metromont Prestress Company
P.O. Box 2486
Greenville, SC 29602
Attn: Kevin McSIoy
Fax: 864-295-5515
Eal:
mebol-presbss.cxjm
Visit our website:
www.metromont-prestress.com
EOE
PCI is an Equal Opportunity Employer
147
Consulting
Services.
PCI Professional Members
-.
A-1,Ig-ff ptint
i Concrete
Analysis
Hard Working
Software
at your
Fmger Tips
-.
148
451-5327
psoftcom
r
Now you can have full

access to LEAPs new
release of Pte,to, the
best tool for design and
analysis of prestressed or
mildly reinforced
concrete beams in accordance with the
ACI code. Add Prb to your
design libraiy and have the power
-l
of hard worldng software at your
nger tips.
free trial pathige

Call LEAP Software to request a bmchure.
PCI JOURNAL
LG
DE$IGNI
INC
Professional fngineering & Drafting Services

6422 Bellingham Ave., Suite 220
No. Hollywood, CA 91606
Tel. (818) 761-6276 :
Fax. (818) 761-6277
http://wvw.ldesins.com
Experienced in ArchieduraI & S1rucfural Precasf, FRC & Sfone

PROUDLY SERVING THE PRECAST CONCRETE
INDUSTRY
Losch Engineering Corporation

Structural Design Consultants
Architectural and Structural Precast Engineering and Shop Drawings
i/Commercial i/Industrial i/Parking i/Stadium i/Correctional
Loech Software, LLC

Precast/Prestressed Concrete Design Software
LECWALL
for Wall and Column Design
LECPRES
for Beam and Slab Design
Call (847) 705-1396 or visit loschsoft.com for a free 30 day trial

19 S. Bothwell, Suite 200, Palatine, IL 60067
www.loscheng.com
www.loschsoft.com
info@loscheng.com
IATES, INC
(303)42
7000 Broadway #1-1 01, Denver, Corado 80221.
CONSULTING ENGINEERS
Complete Engineering Design and Shop Drawing Services
%COMMI1TED EXCLUSIVELY TO THE PRECAST CONCRETE BUILDING_INDUf
H.
LZIWiIden &
Associates, Inc.
Consulting Engineers
Structural Design
Development of Erection Drawings
Preparation of Piece Details
Coordination of the Review Process
.
Production and Erection Assistance
7310 Tilghman Street, Suite 600, Allentown, PA 18106-9295

610-871-3935 FAX 610-871-4515 E-mail UWA@hwilden.com
149
CALENDAR OF COMING EVENTS
PCI Meetings and Events (2001)

Technical Activities Committee Winter
Meeting, The Westin Regina Resort, Puerto Vallarta, Mexico
January 20-22, 2001
Board of Directors Winter Meeting, The

February 22-24, 2001
Westin Regina Resort, Puerto Vallarta, Mexico
Other Meetings and Events (2001-2002)

11th Annual Construction Safety Conference
February 6-8, 2001
& Expo, Rosemont, Illinois. Contact: Marie Fogo, Exhibits Coor
dinator. Tel.: 1-800-552-7744; e-mail: mfogo@buildsafe.org.
PCI Zone Meeting, Zone 1 (North), Renaissance Madison Hotel, Seattle, Washington
March 7, 2001
PCI Zone Meeting, Zone 2 (South), Park Plaza

Airport, Oaldand, California
March 8, 2001
Hotel
PCI Zone Meeting Zone 3, Minneapolis Airport

Marnott, Minneapolis, Minnesota
April 5, 2001
PCI Zone M..itrng, Zone 2 Meeting, Little

America Hotel, SaltLake City, Utah
April 12, 2001
April 19-22 2001
Committee Days, Holiday Inn Mart Plaza,
Chicago, Illinois
PCI Zone Meeting, Zone 5, Hyatt Regency BaltiMay 3, 2001
more, Baltimore, Maryland
PCI Zone Meeting, Zone 4, Grand Geneva Hotel,

Lake Geneva, Wisconsin
May 10,2001
May 13-19, 2001
World of Concrete Annual Cornmercial Construction Show, Las Vegas, Nevada. Contact: World of
Concrete, 426 South Westgate Street, Addison, flhinois 60101. Tel.: 1800-837-0870 ext. 2653 (in United States), (630) 705-2653 (outside
United States) fax: (630) 543-3112; e-mail woc@hanley
wood.com.
February 27-March 2, 2001
International Conference on Safety and

Risk in Engineering, Malta 2001, St. Julians, Malta. Contact:
Safety and Risk in Engineering do International Association
for Bridge and Structural Engineering (IABSE), ETH Hongger
berg, CH-8093 Zurich, Switzerland. Tel.: 41-1-633 2647; fax:
41-1-633 1241; e-mail: malta.2001 @iabse.ethz.ch or secre
tariat@iabse.ethz.ch; internet: www.iabse.ethz.ch/confer
ences/malta!.
March 21-23, 2001
PC-21 Tour, Crowne Plaza Madrid City Cen
tre, Madrid Spain

PCI Zone Meeting, Zone 6, Savannah Marriott
May 26, 2001
Riverfront,Savannah,Georgia
Board of Directors Summer Meeting, Hyatt

Fishermans Wharf, San Francisco, California
June 7-9, 2001
6th World Conference: Cities

February 26-March 2, 2001
and the Third Millennium, Melbourne, Australia. Contact:
Council on Tall Buildings and Urban Habitat, Lehigh Universi
ty, 11 East Packer Avenue, Bethlehem, Pennsylvania 18015.
Tel.: (610) 758-3515; fax: (610) 758-4522; internet:
www.icms.com.au/tbuh.
ACI Convention, Philadelphia, Penn

March 25-30, 2001
sylvania. Contact: Conventions and Meetings, American
Concrete Institute, P.O. Box 9094, Farmington Hills, Michi
gan 48333-9094. Tel.: (810) 848-3700; fax: (810)
848-3701.
4th Biennial John Earley Symposium

on the Historic Development of Metropolitan Washington, D.C.,
College Park, Maryland. Contact: Jere Gibber, Conference Cor
rdinator. Tel.: (703) 768-6987; e-mail: jgibber@aol.com.
March 31-April 1 2001
12th International GRCA Congress, Dublin,

May 14-16, 2001
Ireland. Contact: GRCA, 26 Gorsey Brow, Billinge, Wigan,
WN5 7NX, Lancs, United Kingdom. Tel.: 44 (0)1744 893423;
fax: +44 (0)1744 892359; e-mail: rferry.grca@dial.pipex.com.
PCI Conventions
2001 October 21-24
John Ascuagas Nugget
Reno Nevada
2002 October 6-9
Opryland Hotel
Nashville, Tennessee
2003 October 19-22
Renaissance Orlando Resort

O r lando Flo iid a
150
The American Institute of Architects 2001

May 17-19, 2001
National Convention and Expo, Denver, Colorado. Contact: The
American Institute of Architects, 1735 New York Avenue, NW,
Washington, D.C. 20006. Tel.: (202) 626-7395.
10th International Congress on Polymers in

May 21-24, 2001
Concrete, Honolulu, Hawaii. Contact: Association Concepts. Tel.:
(248) 848-3820; fax: (248) 848-3801; e-mail: info@aci-int.org.
2nd Worldwide ECCE Symposium on Infor

mation and Communication Technology in the Practice of
Building and Civil Engineering, Espoo, Finland. Contact: Sym
posium Secretariat, Association of Finnish Civil Engineers RIL,
June 6-8, 2001

.
PCI JOURNAL
Meritullinkatu 16 A 5, FIN 00170, Helsinki, Finland. Tel.:

358-9-6840 7818 (Contact: Siv Forsten); fax: 358-9-1357
670; e-mail: siv.forsten@ril.fi; internet: www.ril.filecce.htm.
International Association for Bridge and
Structural Engineering Conference, Cable-Supported
Bridges
Challenging Technical Limits, Seoul, Korea.
Contact: IASBE/AIPC/IVBH, ETH Honggerberg, CH
8093, Zurich, Switzerland. Tel.: +41-1-633 2647, fax: 411-633 1241; e-mail: secretariat@iabse.ethz.ch; internet:
www.iabse.ethz.ch/conferences/seoul/seoul_f.html.
June 12-14, 2001
INDEX TO ADVERTISERS
139
American Spring Wire Corp
BMH Systems
3rd International Conference on Concrete Under

Severe Conditions of Environment and Loading, Vancouver, British
Columbia, Canada Contact: Professor N. Banthia, Depatiment of Civil
Engineering, University of British Columbia, 2324 Main Ma]], Vancou
ver, British Columbia, Canada V6T 1Z4. Tel.: (604) 822-9541; fax:
(604) 822-6901; e-mail: banthia@civil.ubc.ca.
June 18-20, 2001
International Conference on Building En

June 26-29, 2001
velope Systems and Technologies (ICBEST), Ottawa, Cana
da. Contact: Monique Myre, National Research Council of
Canada (NRC), Institute for Research in Construction (IRC),
Bldg. M-20, Montreal Road, Ottawa, Ontario, Canada KIA
0R6. Tel.: (613) 993-0435; fax: (613) 952-7673; e-mail:
icbest@nrc.ca.
9th International Conference and Exhibition

July 4-6, 2001
on Structural Faults & Repair, London, England. Contact:
Conference Secretariat, ECS Publications, 46 Cluny Gardens,
Edinburgh EH1O 6BN, United Kingdom. Tel.: +44-(0)13l447 0447; fax: +44-(0)131-452 8596; e-mail: conference@ec
spublications.com.
Bonnybrook Custom Steel

Forms, Ltd
Inside Front Cover
The Consulting Engineers Group, Inc
148
Ensoft, Inc
142
Eriksson Technologies, Inc
143
Fister Quarries Group, Inc
140
3
Florida Wire and Cable, Inc

Gregory Development Services, Inc
Hamilton Form Company, Inc
140
Back Cover
High Concrete Accessories
JVI, Inc
7
139, 146
LEAPAssociates
International Conference on Tall

September 5-7, 2001
Building Structures and Architecture, Frankfurt-on-Main,
Germany. Contact: Universitt Leipzig, Institut fr Massivbau
und Baustofftechnologie, Marschnerstrabe 31, D-04109
Leipzig, Germany. Tel.: 49 (0) 341/97 33 800; fax: 49 (0)
341/97 33 809; e-mail: hoepfner@wifa.uni-leipzig.de OR
laubach@wifa.uni-leipzig.de.
55th Rilem Annual Week Symposium

on Connections Between Steel and Concrete, Stuttgart, Ger
many. Contact: Symposium Secretariat, do IWB University of
Stuttgart, Pfaffenwaldring 4, 70550 Stuttgart, Germany. Tel.:
+49 (0) 711-685-3320; fax: 49 (0) 711-685-3349; e-amil: con
tact@iwb.uni-stuttgart.de; internet: www.iwb.uni-stuttgart.de.
September 9-12, 2001
LEAP Software, Inc
148
LGDesign, Inc
149
Losch Engineering Corp
149
Master Builders, Inc
Mi-Jack Products
Mixer Systems, Inc
PACS
142
Precast, Inc
135
5th World Congress on Joints, Bearings

October 7-11, 2001
and Seismic Systems for Concrete Structures, Rome, Italy. Con
tact: Studio Ega s.r.], Viale Tiziano, 19-00196, Rome, Italy.
Tel.: +39-06-328121, fax: 39-06-3240143, e-mail: ega@ega.it.
Prestress Supply, Inc
142
R R Consultants
RATEC GmbH
10-11
Co
Ross Bryan Associates, Inc
140
Sirko Associates, Inc
149
ACT Convention, Dallas,

Meetings, American
and
Conventions
Texas. Contact:
Concrete Institute, P.O. Box 9094, Farmington Hills,
Michigan 48333-9094.Tel.:(810)848-3700; fax: (810)
848-3701.
October 28-November 2, 2001
March 15-24 2002
Inside Back Cover
ACT Convention, Boston, Mas

sachusetts. Contact: Conventions and Meetings, American
Concrete Institute, P.O. Box 9094, Farmington Hills,
Michigan 48333-9094. Tel.: (810) 848-3700; fax: (810)
848-3701.
Sumiden Wire Products

Corporation
Tuckers Machine & Steel
H. Wilden & Associates, Inc
2
149
Interested in advertising in the PCI

JOURNAL? Request a media kit by con
tacting PCI, 209 W. Jackson Boulevard,
Ste. 500 Chicago, Illinois, 60606. Tel.:
(312) 786-0300. Fax: (312) 786-0353.
151
PCI
qiona(7v1arktii Offices
Regional assistance is available in the areas listed below. If there is not a representative in your
region, please contact PCI Headquarters in Chicago, Illinois.
Colorado Prestressers Association
WalterJ. Prebis
phone: (303) 238-5674; fax: (303) 238-8734
e-mail: wallycpa@aol.com
Georgia/Carolinas PCI
W. Michael McConochie
phone and fax: (704) 873-3071
e-mail: pcigc@i-america.net
Central Region PCI
Edward P. Tumulty
phone: (614) 436-3322; fax: (614) 436-4260
e-mail: tumulty@netwalk.com
MAPA Mid-Atlantic Precast Association
Monica Schultes
phone: 800-453-4447; fax: (302) 235-1 1 39
e-mail: monica.schultes@gte.net
New England Region PCI
Rita L. Seraderian
phone and fax: (61 7) 489-5670
e-mail: rlsera@aol.com
PRECASTIPRESTRESSED CONCRETE INSTITUTE

175 W. Jackson Boulevard
PCI Headquarters
phone: (312) 786-0300; fax: (312) 786-0353
PCMAC Precast/Prestressed Concrete
Manufacturers Association of California, Inc.
Doug Mooradian
phone: (818) 247-61 77; fax: (818) 240-3041
Precast/Prestressed Concrete Producers of Illinois
Edward J. Gregory
phone: (708) 429-1444; fax: (708) 429-9222
e-mail: gregoryed@msn.com
Prestressed Concrete Association of Pennsylvania
Heinrich 0. Bonstedt
phone: (610) 395-2338; fax: (610) 395-8478
e-mail: bonstedt@internetmci.com
www.pcap.org
STRESSED CONCRETE INSTITUTE
Subscription Rates:
United States
Other Countries
$38 one year
$53 one year
$90 three years
$131 three years
PLEASE SEND ME the bimonthly PCI JOURNAL. (Payment in U.S. dollars must accompany your order.)
NAME
FIRM
ADDRESS
CITY____
STATE_
71P
COUNTRY
152
PCI JOURNAL
YOU CAN FIND STRAND N

IN THE NICEST PLACES.
Freeway overpasses, parking garages, and yes, even the most prestigious of high-rises
and hotels require the strength that only Sumiden strand can provide.
Our customers can always count on the high quality and precision of our American-made
products to stand for superior performance and on-time delivery.
Call your StrandMan
Bill, Jeff or Ken, today!
Sumiden Wire Sales

Nashville, TN (615) 463-9750
San Clemente, CA (949) 366-9577
Las Vegas, NV (702) 450-3990
Sumiden Wire Products Corporation

RD. Box 8719-95208
The Strand That
Stands On its Own
1412 El Pinal Drive Stockton, CA 95205

(209) 466-8924 Fax (209) 941-2990 Telex 359424
710 Marshall Stuart Drive Dickson, TN 37055
(615) 446-3199 Fax (615) 446-3197
13290 Sabre Boulevard Victorville, CA 92394
(760) 246-6091 Fax (760) 246-3023
I
I
70091k
WAY R ID
FOi
ii
Ham On 4
m C
Wbw
3
WORT TEXAS6118. thone (8
w.haml nformjm sale1
@hamllt
Sne.
9 590 11
ftrm
Pa (817k) 5 5.1110

PCI Journal Sept-Oct 2000 VOL. 45 NO. 5 PDF

Uploaded by

Document Information

Original Description:

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 Sept-Oct 2000 VOL. 45 NO. 5 PDF

Uploaded by

Copyright:

Available Formats

I JI

This state of the art concrete transporter features full 6-yard

Long, high reach

Completely sealed hydraulics

Massive frame, axle and tires

14 auger for faster delivery less auger wear

MACHINE & STEEL SERVICE, INC

Leesburg, Florida 34749-2810

Visit our website: vvww.tuckerbilt.com

For a Strong Start...

components that offer high performance and keep you competitive.

long-term structural integrity and enhanced surface finish.

performance. A wide variety of repair products and

epoxy adhesives offer low-cost solutions for all

formliners, architectural stains, and protective coatings and sealers.

vol. 45, no- 5

STREST SED CONCRETE INSTITUTE

14 2000 PCI Design Awards

Central ArterylTunnel Project: PrecastlPrestressed

130 Reviews of Technical

Proposed Revisions to 1997 NEHRP Recommended

Nell Hawkins and S. K. Ghosh

131 Reference Cards

In Part 2 of this series of articles, the authors discuss specific

135 Industry News

150 Calendar of Coming Events

Design Criteria for Headed Stud Groups in Shear:

Nell S. Anderson and Donald F. Meinheit

Influence of Flexure-Shear Cracking on Strand

Based on a comprehensive test program, the authors recommend specific

Strand Development and Transfer Length Tests on High

Pennsylvania. For more details on this

project, see p. 20.

Transfer Length of CFRP/CFCC Strands for Double-T Girders

Presents the results of a study on variations in transfer length in two types of

PCI DIRECTORS, TAC & STAFF

Reliability of I8 in. Anchor Stud

Evaluation of a High Performance

Continuous CFRP Prestressed

PRECAST/PRESThESSED CONCRETE PNSTTUTE

PCI JOURNAL STAFF

Plant Certification Director

7550 North Linder Avenue

Skokie, Illinois 60077 USA 847/675-1560

BMH SYSTEMS Custoifi Concrete

Concrete Products Plants

r,i. (450) 3494770/f- (450) 449-4898

How to Save Money When You

Cost Comparison Per Piece

900 Double Tee Stem Blockouts

HIGH CONCRETE ACCESSORIES

lnclodcs Coot of Field Palchrng

A AotOn olhi6, cotuem.

and components that effectively rationa

our Ratec catalog today or arrange a per

Europ3e: RATEC GmbH + Co. Karisruher Strasse 32 68766 Hockenheim/Germany

u!ir@j1i YOU OAN

PUT CONCRETE WHERE YOU WANT IT! WHEN YOU WANT IT

PRODUCTS YOU CAN DEPEND ON

1804 W. Lake Parker Drive Lakeland, Florida 33805 863-683-4492

The PSI Hercules Single Strand Tensioning unit is over

NEW FEATURES INCLUDE