Professional Documents
Culture Documents
FRP Reinforcement:
Case Studies
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Editors:
Sami Rizkalla international
®
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Editors
Sami Rizkalla
Antonio Nanni
®
international
SP-215
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The Institute is not responsible for the statements or opinions expressed in its publications.
Institute publications are not able to, nor intended to, supplant individual training,
responsibility, or judgment of the user, or the supplier, of the information presented.
The papers in this volume have been reviewed under Institute publication procedures by
individuals expert in the subject areas of the papers.
Copyright © 2003
AMERICAN CONCRETE INSTITUTE
P.O. Box 9094
Farmington Hills, Michigan 48333-9094
All rights reserved, including rights of reproduction and use in any form or by any means,
including the making of copies by any photo process, or by any electronic or mechanical
device, printed or written or oral, or recording for sound or visual reproduction or for use in
any knowledge or retrieval system or device, unless permission in writing is obtained from
the copyright proprietors.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
In 2001, ACI Committee 440 published ACI 440.1R-Ol, "Guide for the Design and
Construction of Concrete Reinforced with FRP Bars," as the first document in the Emerging
Technology Series (ETS) adopted by ACI. ETS documents provide recommendations on
the use of new technologies based on available test data, technical reports, and limited
experience with field application. Based on the intense research that has been conducted
worldwide, the committee revised and updated this report in 2003 (ACI 440.1R-03).
FRP systems have been used extensively for strengthening concrete structures as an alternative
to traditional strengthening techniques. In 2002, ACI Committee 440 published ACI 440.2R-02,
"Guide for the Design and Construction of Externally Bonded FRP Systems for Strengthening
Concrete Structures." This was the second document in the ETS series published by ACI.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
As a result of the extensive use of FRP as internal and external reinforcement for new
structures and strengthening concrete structures, respectively, the committee organized three
technical sessions on "Field Application ofFRP Reinforcement-Case Studies" with a total
of 21 papers presented at the ACI Fall 2003 Convention, in Boston, Massachusetts. All
papers dealt with case studies and demonstration projects to provide clear evidence of the
practicality, credibility, and maturity attained by this technology.
This volume includes the papers presented during the Fall 2003 Convention, plus five
additional papers that augment the range of field applications and case studies. The goal of
this document is to help practitioners implement FRP technology, while providing testimony
that design and construction with FRP material systems is rapidly moving from emerging to
mainstream technology.
iii
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
ACKNOWLEDGMENTS
The editors extend their sincere gratitude to the numerous volunteers and ACI staff who
assisted in producing this volume. Special thanks are given to the members of ACI
Committee 440 for this effort in reviewing the papers published in this volume.
R. Aboutaha J. Gergely C. Paolo
T. Alkhrdaji W. Gold R. Parretti
C. Bakis N. Grace M. Porter
P. N. Balaguru M. Green M. Postma
L. Bank M. Greenwood H. Rasheed
A. Belarbi D. Gremel T. Rizk
B. Benmokrane H. R. Hamilton M. Schupack
G. Blaszak I. Harik D. Scott
T. Bradberry K. Harries R. Sen
S. Brena M. Henderson M. Shahawy
G. Brown B. Horeczko C. Shield
V. Brown T.lbell K. Soudki
J. Busel V. Karbhari S. Steere
I. Campbell J. Korff R. Steffen
P. Casadei M.Lee G. Tadros
C. Dolan J. Lees J. Thomas
D. Duthinh M. Levar H. Toutanji
R. El-Hacha I. Mahfouz M. Vadovic
G. Fallis H. Marsh D. Vanderpool
A.Fam 0. Marshall M. Vatovec
E. Fyfe A. Mirmiran L. Walden
A. Ganjehlou A. Mosallam D. White
D. Gee K. Neale P. Ziehl
T. R. Gentry J. Newhook
Special thanks are given to the Session Chairs and staff at ACI Headquarters, especially
Todd Watson and Bonnie Schmick, who have worked tirelessly to complete this publication
in time for the 2003 Fall ACI Convention in Boston.
Editors
Sami Rizkalla
Chair, ACI Committee 440, 1997-2003
Distinguished Professor of Civil Engineering and Construction
North Carolina State University, Raleigh, NC
Antonio Nanni
Chair, ACI Committee 440, 1991-1997
V &M Jones Professor of Civil Engineering
University of Missouri-Rolla, Rolla, MO
iv
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
TABLE OF CONTENTS
Acknowledgments ............................................................................................................. iv
SP-215-10: New Bridge Piers Using Load Bearing Concrete-Filled GFRP Tubes ........ 181
by M. Pando, A. Fam, J. Lesko, and G. Filz
v
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
SP-215-12: Strengthening of a Bridge Using Two FRP Technologies ........................ 219
by P. Casadei, N. Galati, R. Parretti, and A. Nanni
SP-215-14: Flexural Strengtheing of Impacted PC Girder with FRP Composites ......... 249
by R. Parretti, A. Nanni, J. Cox, C. Jones, and R. Mayo
SP-215-20: Field Retrofit of Prestressed Concrete T-Beam Using CFRP ................... 335
by I. N. Robertson, A. A. Agapay, and L. M. Nakashima
SP-215-23: Shear Strengthening with Bonded CFRP L-Shaped Plates ...................... 373
by M. Basler, D. White, and M. Desroches
vi
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
vii
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
viii
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CHAPTER!
3
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
INTRODUCTION
With the recent publication of design guidelines for the application of FRP internal and
external reinforcements to reinforced concrete, design engineers now have tools with
which to meet challenges, which steel reinforcement has not satisfactorily met in the past.
Not the least of these challenges has been that presented by the aging and deterioration
transportation infrastructure of the United States. To a significant degree these same
challenges face engineers at the Texas Department of Transportation (TxDOT). Fiber-
reinforced-polymer (FRP) reinforcement has been used in Texas to meet such challenges
on a number of projects in recent years. Some applications, such as using FRP as internal
reinforcement for the purpose of eliminating electrochemical corrosion of reinforcement
and subsequent corrosion product induced deterioration of surrounding concrete, have
limited application in Texas. This is because, other than in marine environments,
concentrations of chlorides in concrete exceeding corrosion threshold levels are not as
prevalent in Texas as they are in jurisdictions to the north. Other applications, such as
externally bonded FRP strengthening of bridge elements, whose original design strength
is now substandard making otherwise healthy looking concrete bridges load rate below
acceptable levels for widening or other rehabilitation, are likely to be more widely
employed. Externally bonded FRP fabrics used in the repair of impact damaged bridge
girders are fast becoming an accepted practice by TxDOT bridge maintenance engineers
and thus shows promise for more widespread use in Texas. Yet another emerging
application for FRP internal reinforcement exploits the non-magnetic property of GFRP
reinforcement to provide reinforced concrete that does not interfere with vehicle imaging
loops requiring magnetic/electrical isolation near turnpike toll plazas. Examples of these
projects plus an application where FRP reinforcement is used to correct a design flaw are
presented in this paper. Finally, the future of FRP reinforcement in Texas transportation
infrastructure is briefly considered.
RESEARCH SIGNIFICANCE
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
FRP REINFORCEMENT
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Three completed projects are reported to illustrate the use of externally bonded FRP
fabrics and laminates in the repair and strengthening of Texas transportation
infrastructure projects:
• The repair of an impact-damaged bridge superstructure using a carbon FRP
(CFRP) system.
• The strengthening of a pan girder bridge using CFRP systems to increase its load
rating.
• The repair and strengthening of the ends of inverted-tee bent caps having cracked
reentrant comers, using a CFRP system.
In the Texas panhandle, frequent use of chloride-containing deicing agents has pushed
the chloride concentration at the level of reinforcement in many concrete bridges past the
steel corrosion threshold, resulting in corrosion-induced deterioration of the
reinforcement and deleterious cracking of the surrounding concrete. One completed
Texas project illustrates use of glass FRP (GFRP) internal bar reinforcement as an
alternative to epoxy-coated steel reinforcement for preventing corrosion-induced concrete
deterioration in a bridge deck.
Project Overview-In the year 2000, TxDOT used GFRP bars as top mat
reinforcement in the concrete deck of the Sierrita de la Cruz Creek Bridge constructed
near Amarillo, Texas. Placed in service in January 2001, the phase-constructed bridge
carries an estimated 1,650 vehicles per day (including approximately 300 trucks, many of
which are fully loaded gravel trucks) along a portion of Ranch-to-Market Road 1061
(RM 1061) situated in Potter County, Texas. The bridge is the first Texas transportation
application of internal FRP reinforcement.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
required testing. Two interim research reports have been published (8, 9).
The Texas funded TxDOT Research Project No. 0-4138, "Full-Scale Crash Tests of FRP
Bar Reinforced Bridge Rails," conducted by TTl grew directly out of the IBRCP funded
project in order to investigate the crash worthiness of fully FRP reinforced concrete
bridge rails. This project addresses the behavior of such concrete bridge rails on concrete
deck slab structures where both rail and slab are reinforced with GFRP bars when the
rail/slab system is subjected to full-scale vehicle crash tests in accordance with the
requirements of NCHRP 350, Test 3-ll. This project has been completed and the final
and summary reports approved for publication (1 0, 11).
TxDOT has also sponsored UTA's participation in the bond test portion of the
"ConFibreCrete" European research network, Federation of Concrete (FIB) Task Group
9.3, and ISIS Canada jointly organized "Round Robin Tests for FRP Reinforcement" (12,
13).
To demonstrate and generate interest in the IBRCP project, the Federal Highway
Administration (FHWA) sponsored a half-day innovation showcase on July 25, 2000, in
conjunction with the July 26, 2000 placing of the concrete for the first phase of the bridge
deck. Also, the bridge design engineer has received and filled requests for technical
information about this project. The requests came from several interested state DOTs
who were considering, planning or designing their own FRP bars in bridge deck projects
as well from Dr. Brahim Benmokrane was designing such a project to be build in Canada.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Patching of the concrete and the epoxy injection took four days; CFRP composite
installation took one day. The total cost was $47,000, which included $22,000 for
concrete repair and $25,000 for CFRP composite installation. The area of CFRP
installation was about 13.94-sq m (150-sq ft). This small area of CFRP and the cost of
mobilization may be the primary reasons this job was so costly on a unit basis. After
completion of all repair work, all lanes under the bridge were opened to traffic in April of
2002 (see Figure 5).
Past Project: Externally Bonded CFRP Systems for Increasing the Load
Rating of a Bridge
Background-Brena, et al (15, 16) conducted TxDOT sponsored research to
investigate the feasibility of using externally bonded FRP composite material to increase
the load rating of bridges in order to make them eligible for widening 1• The research
testing program involved two phases-first, static and dynamic tests on a total of thirty
1
A bridge widening project requires that the portion of the bridge to remain have a load
capacity high enough to justifY widening over replacement. 'This capacity is quantified
through various load ratings that are normally based on the original bridge plans and
material specifications but sometimes include such items as mill test reports to get a
better idea on the "real" material properties. However, assessment of this load rating
involves other information on the bridge along with engineering judgment. As a rule of
thumb, a proposed bridge widening will not be approved unless the load rating of the
remaining portion of the bridge after widening is at least H-20 (operating level) and no
less than about HS-17 (inventory level). The number shown for the H rating is the
allowable weight of the vehicle in tons, and the number shown for the HS rating is the
rating factor multiplied by 20. Thus, a rating factor of 1.0 would result in a rating of HS
20, a rating factor of 1.25 would result in an inventory rating ofHS 25, and a rating factor
of0.75 would result in an inventory rating ofHS 15.
Figures 6 through 8 are the pan-girder bridge specimens during testing and after
de bonding failure of the external CFRP laminate system.
Overview of the TxDOT Research Implementation Program-TxDOT has a
policy encouraging the implementing of all "field ready" results from its research
programs as well as from that of other agencies or even private entities such as
transportation product manufacturers. In this policy, implementation is defined as the
adoption of a product for use, including technology transfer activities that promote
adoption, such as the following:
• Information Dissemination-includes the development, packaging and distribution
of brochures, manuals, articles, reports, videos, and other materials which provide
product descriptions and instructions to enable and promote use.
• Training-includes training course development and conduct necessary to enable
and promote use.
• Demonstration-the placing of a product into TxDOT's operational environment
to demonstrate its use, which includes the following:
• Deployment-the initial procurement and dissemination of a product to users,
and
• Implementation Field Testing-the demonstration and/or verification of
product performance in TxDOT's operational environment, including
District/Division/Office (DIDIO) or National Experimental and Evaluation
Program (NEEP) projects.
Overview of the Implementation Project A demonstration type project was
approved for implementation ofthe research results ofBrena, et al (15, 16). The research
results were implemented in the widening of a bridge carrying Farm-to-Market Road
1632 (FM 1362) traffic over Sue Creek in Burleson County. The original bridge is a two-
span reinforced concrete pan-girder structure-a very common bridge type in Texas that
was issued as a standard for various pre HS-20 design loading bridges built between 1948
and 1964. The two simply supported spans are each 9.14-m (30-ft) long, and the overall
width of the bridge is 6,553-mm (21.5-ft), with proposed widening out to 9,754-mm (32-
ft).
first span was strengthened with longitudinally oriented CFRP fabric produced by Master
Builders as the primary strengthening reinforcement and transverse CFRP fabric straps
produced by Master Builders as the secondary reinforcement used to control debonding
of the longitudinal reinforcement (see Figure 9). The second span was strengthened with
longitudinally oriented CFRP pultruded laminates produced by Sika Corporation as the
primary strengthening reinforcement and transverse CFRP fabric straps produced by Sika
Corporation as the secondary reinforcement used to control debonding of the longitudinal
reinforcement (see Figure 10). The CFRP application took less than a week per span to
complete. The CFRP strengthening systems were designed to bring the bridge load rating
up to at least H-20, which is the same load level as HS-20 for this short span structure.
In accordance with the recommendations of the two CFRP systems manufacturers the
contractor used a combination of sand blasting and grinding to prepare the surface to
receive the CFRP materials. Grinding was mostly used to smooth out the bottom of the
girders. A scaffold was used to provide easy access to the beam stems where the FRP
materials were applied.
The construction costs for the project are detailed as follows:
Sub-cost Cost
Roadway ............................. .. $207,907
Bridge ................................ .. $162,092 ($84.42/f
• Strengthening (FRP appl.)...... $63,500 ($33.07/ft2)
• Repairs (epoxy injection, etc.) $15,000 ($7.81/fr)
• Widening......................... $83,592 ($43.54/ft2)
Project Total. .......................................................... $369,999
TxDOT widened the bridge at a cost of $908.69/m2 ($84.42/ft2) rather than replacing it
for between $377/m2 and $484/m2 ($35/ft2 and$45/ft2). However, the project did not have
a high enough priority to compete for federal bridge replacement funds, and thus was not
eligible for widening without first being strengthened. TxDOT's solution allowed the use
of federal funds to increase the capacity of the structure and funded the implementation
of research findings in an area of great potential value to TxDOT and other transportation
agencies. Replacement, if the bridge load rating had warranted it, would have required
either acquisition of the additional right-of-way to build a detour during bridge
demolition and reconstruction or use of phasing to handle traffic during construction.
This would have complicated the project design and construction, lengthening the project
development and construction time, adding both construction cost and highway user cost
to the total cost of the bridge replacement project.
Work on this bridge was completed on August 8, 2002. TxDOT engineers have not
categorically recommended or dismissed the use of CFRP for strengthening of such
bridges. The strengthening alternative must compete against all the other alternatives
regarding managing an existing bridge that is in some way deficient. Each such project is
evaluated on a case-by-case basis with regard to hydraulic issues, project location, right-
Past Project: Externally Bonded CFRP System for Strengthening the Shear
Capacity of the Ends of Inverted Tee Bent Caps
Background-The design procedure for inverted tee bent caps in Texas is based on
the research of Furlong and Mirza (17-20). That procedure, in part, involves detailing
web stirrups that act as hangers that effectively deliver the concentrated loads applied to
the bent cap flanges (inverted-tee ledges) into the body of the web (inverted-tee stem).
The effect of these "hangers" is easily visualized by considering a section through an
inverted tee bent cap (Figure 11) where the hangers function to prevent the flanges from
separating (cracking) excessively from the web and pealing open from the reentrant
comer. This reentrant comer is often cracked in structures that are in service, especially
when the exterior beam is near the end of the cap. If sufficient hanger steel is not
provided in this region the aesthetically displeasing reentrant comer crack could be
transformed into a crack which compromises the safety of the structure.
In such a scenario, one method of fixing the structural deficiency is to use an externally
bonded FRP system to compensate for the insufficiency of the size and/or number hanger
bars provided near the end of the cap. This was the choice made by the Houston District
bridge design engineer who was faced with this problem after such an inverted tee bent
cap had already been constructed. The engineer's design methodology was loosely based
on guidance from the April 1999 working draft of the now published ACI Committee 440
document Guide for the Design and Construction of Externally Bonded FRP Systems for
Strengthening Concrete Structures (5) and the ASCE document Structural Plastics
Design Manual (21). The design engineer employed a strength-based approach to
determine the amount and direction of carbon fiber reinforcement, relying on strength
reduction factors to limit strain across the expected crack and included superstructure
dead load in the analysis because the beams had not yet been erected. An ongoing
University of Houston conducted research project, TxDOT Research Project 0-1854,
"New Method for Serviceability Design of Inverted 'T' Bent Cap Ledges", to develop
serviceability criteria for inverted tee bent caps, provided cracked specimens for use in
testing the repair methodology. A specimen was wrapped using the same CFRP materials
as planned for the strengthening and then loaded to failure. This test showed that the
CFRP materials are forgiving in terms of workmanship and that the design approach
employed to remedy the hanger bar deficiency was conservative (Figures 12 through 14).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The IH 10 Overpass carrying IH 10 traffic over SH 304 in Gonzales County was hit by an
overheight vehicle in October 2002 causing significant damage to two girders and slight
damage to a third. Engineers proposed repair procedures similar to those used for the FM
1927 structure. For the exterior beam, which sustained the most damage, CFRP wrapping
around the entire bottom flange and beam webs has been detailed; the adjacent less-
damaged beam will only have CFRP fabric wrapped around the bottom flange and
halfway up the bottom beam chamfers. The repair work for these two beams is Category I
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
in accordance with the Special Specification Item 4421, "Repair Impact Damaged
Prestressed Concrete Bridge Beams" (Appendix "A"). The classification of the repair for
the third damaged beam drops to a Category II because the damage is less extensive.
Figures 18 and 19 show some of the details that have been developed for the repair of this
impact damage. As of April2003, this maintenance project was pending.
surface of the concrete. To meet these requirements, TT A requested that the designer
provide a top mat of GFRP bars in all required non-magnetic zones because GFRP bars
are non-conductive and non-magnetic. This creates unique requirements for the bridge at
the southbound entrance:
• It needs to accommodate a toll plaza.
• It must not use traditional steel reinforcing in the top mat of the concrete deck.
• It must not have reinforcing bars of any kind in the top 76-mm (3-in) of the deck.
TTA contracted with Carter & Burgess to change the original design to accommodate the
loop detectors and to provide full-width GFRP-reinforced sections of the deck extending
along the end 25% of each span in the plaza area. The typical section will be a 254-mm
(10-in) thick slab with a top mat of GFRP reinforcing with 76-mm (3 in) of clear cover.
The bottom mat will have traditional steel reinforcing. The deck will transition from the
Future Projects
Future Texas transportation infrastructure projects involving FRP reinforced concrete are
a certainty although none beyond one or two over-height vehicle impact damaged
prestressed concrete bridge beam repairs are currently scheduled for design and
construction. Such projects await an attempt to eliminate corrosion-induced deterioration
of a concrete bridge deck in a severe corrosive environment, the next over-height vehicle
impact on a prestressed concrete girder bridge, the opportunity to keep an existing bridge
in service by increasing its load capacity, the need to strengthen or repair bridge elements
where shear strength is in question, a decision to provided magnetic transparency in
bridge deck or pavement sections of toll plazas requiring magnetic transparency, or
possibly an as yet unconceived application to solve some sort of transportation
infrastructure problem.
Design and performance issues need to be addressed before wholesale use of FRP
reinforcement in Texas transportation infrastructure can be approved. Design guidelines
must address questions about the structural implications of the reduction in ductility that
use of FRP reinforcement creates, giving designers mitigation options for dealing with
design-limiting ramifications. Questions about long-term residual material properties,
especially with regard to internal reinforcement, are a significant barrier to general
acceptance of FRP by practicing engineers.
This paper shows that engineers responsible for design, rehabilitation and repair of Texas
transportation infrastructure have so far had a positive experience with FRP reinforced
concrete, with FRP reinforcement offering performance or application that exceeds the
performance or applicability of traditional steel reinforcement. Although TxDOT has not
sanctioned the wholesale use of FRP reinforcement, the TxDOT projects described in this
paper demonstrate that FRP reinforcement is emerging as a viable option for the repair
and strengthening of existing concrete and for reinforcement of new concrete
transportation infrastructure in Texas. The use of externally bonded CFRP systems for
repair of over-height vehicle impact damaged prestressed concrete bridge beams is stated
as having already becoming common practice in Texas. Using externally bonded CFRP
systems for increasing the load ratings of reinforced concrete bridges has been
implemented on one bridge in Texas and thereby has become a rehabilitation alternative
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
ACKNOWLEDGEMENTS
The authors acknowledge the contributions. of Dingyi Yang, P.E., and Mark Steves, P.E.,
of the TxDOT Bridge Division, and John Vogel, P.E., of the TxDOT Houston District,
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Kayvon Jahedkar, P.E., of the TxDOT Bryan District, and Doug Woodall, P.E. of the
Texas Turnpike Authority for each of their contributions of material for this paper.
REFERENCES
1. Composites for Infrastructure - A Guide for Civil Engineers. Ray Publishing, Inc.,
Wheat Ridge, Colorado, 1998, I 00 pp.
2. Thippeswamy, H. K., Franco, J. M., and GangaRao, H. V. S. FRP Reinforcement in
Bridge Deck. Concrete International, FRP Reinforcing Issue, Vol. 20, No. 6, June
1998.
3. Sourcebook 2003. High-Performance Composites. Ray Publishing, Inc., Wheat
Ridge, CO. http://www.hpcomposites.com/hpcsb/sboverview.html. Accessed March
4, 2003.
4. ACI Committee 440. Guide for the Design and Construction of Concrete Reinforced
with FRP Bars. ACI 440.1R-Ol. American Concrete Institute, Detroit, Michigan,
May 2001, 41 pp.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
SPECIAL SPECIFICATION
4421
Repair of Impact Damaged Prestressed Concrete Bridge Beams
1. Description. This Item shall govern for the repair of impact damaged prestressed
concrete beams as specified herein and/or as directed by the Engineer.
2. Materials. All materials that become part of the final product will be furnished by
the Contractor. All tools, equipment, labor, scaffolding, incidentals and safety
equipment shall be furnished by the Contractor.
Materials to be used in the repair of the structural concrete members shall be submitted to
the Engineer for approval. For proprietary materials, this submission shall include all
manufacturer's recommendations and material application instructions. The Contractor's
work shall be done in accordance with these instructions and recommendations unless
otherwise authorized by the Engineer. Concrete for repair shall contain Grade 1 fine
aggregate, Grade 7 coarse aggregate and a minimum of seven (7) sacks of cement per
cubic yard. All concrete repair material, or proprietary repair material submitted for
approval, shall have a minimum 28-day Compressive Strength of 5000 psi or as shown
on the plans.
All materials shall conform to the pertinent requirements of the following Items:
Item 420, "Concrete Structures"
Item 421, "Portland Cement Concrete"
Item 431, "Pneumatically Placed Concrete"
Item 440, "Reinforcing Steel"
Special Specification "Epoxy and Adhesives"
If requested by the Engineer, the Contractor shall be required to prepare trial mixes of
patch and/or crack injection material to ensure their ability to perform the work as
required.
3. Scope of Work. The damaged prestressed concrete beams shall be returned to their
original geometric section. The beams to be repaired may require one or more of
the basic repair methods listed below and will be so categorized elsewhere on the
plans as follows:
Category 1: Significant spalling, section loss and cracking
Category II: Surface spalling only
Category III: Crack injection only
(1) General.
(a) If directed by the Engineer, the Contractor shall furnish and place, on the
bridge deck, a vehicle such as a loaded 12 yard dump truck or other
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 2: Concrete placement in the GFRP reinforced concrete deck of the Sierrita de la
Cruz Creek Bridge.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 3: Exterior Girder Bottom Flange after Impact with Overheight Vehicle/Payload.
Figure 4: Interior Face of External Girder after Impact with Overheight Vehicle/Payload.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 7: Externally Bonded CFRP Laminate Strengthened Pan Girder Specimen After
Partial Debonding of CFRP Laminate.
Figure 9: Pan Girder Span Strengthened with Externally Bonded Longitudinal CFRP
Fabric Anchored by Transverse Straps of CFRP Fabric.
Figure 10: Pan Girder Span Strengthened with Externally Bonded Longitudinal CFRP
Extruded Laminate Anchored by Transverse Straps of CFRP Fabric.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
.,. • ~ ~. • • •• •
,
~ ~ ~ -.·~Shea
•
• •
r Stirrup
on d Hanger
~ • • • • Rei nforceing
. . .
. ~ I
.~ . . . . . 2" CLR. I
"'0
Q) . . . .• • .
~
.
.~ . .
0
()_
"'0 . . •.
'CXl
... .~ . . 9 '12"!
:J
0"
w
.~ . . . . .
'
..
t
I
BRG •
. .; i
I
o- ~[
.
• . •. . . .
. .
. .
. •
.
. .• p
.
w c:i
"' . . . . .. .
•. .•-=~ ~
N
• . •. . • . • . ••.
v /'--'----
Figure 11: Section Through Inverted Tee Showing Basic Dimensions And Reinforcement
Pattern and Identifying Shear Stirrup and Hanger Reinforcement.
Figure 12: Inverted Cracked Inverted Tee Test Specimen Prior to Application of the
Externally Bonded CFRP Repair System Being Tested.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 13: Application of Externally Bonded CFRP Repair System to the Inverted
Cracked Inverted Tee Test Specimen.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 14: CFRP Fabric Strengthened Inverted Tee Test Specimen Loaded to Failure.
01)e.r lo..p
t €>l' ~~...
j2.Y 2;-,J
C." ~
Figure 15: Design Sketch Showing the Order the Application of the CFRP Fabric Strips
to the End of the Inverted Tee Bent Cap.
Figure 16: CFRP Fabric Strengthened Adjacent Inverted Tee Bent Caps
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 17: Girders Flush Framed With An Inverted Tee Bent Cap
Squared Edges
"'-~-Polyethylene Tubing
lfor injection of
internal cracksl
Patch material
CFRP SECTION
FOR BEAM •S
Squared edge (for
all patchmg areasl
** Place the Long1tudmal
direction of the CFRP
Sheet perpendicular SECTION A-A
to the Beam. TYPICAL SECTION OF REPAIR
!TYPE C BEAMI
Figure 19: Prestressed Concrete Beam Repair Details IH 10 Overpass at SH 304 Showing
Approximation of Extent of Damage to Exterior Beam Cross Section along With Repair
Instructions, Including CFRP Wrap
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: A new highway bridge was constructed using FRP bars as reinforcement for
the concrete deck slab. The bridge, located on highway 55 North over the Magog River
(Quebec, Canada), is a girder type with a total length of 83.7 m and five main steel
girders continuously supported over three spans. The two end spans are 26.2 m each and
the middle one is 31.3 m. The deck is a 220-mm thickness concrete slab continuous over
four spans of 2.845 m each with an overhang of 1.352 m on each side. One full end span
(26.2 m) was totally reinforced with FRP bars in top and bottom mats. The other two
spans of the bridge were reinforced with galvanized steel. The steel reinforced concrete
deck slab was designed according to Section 8 of the New Canadian Highway Bridge
Design Code, CHBDC (CAN/CSA-S6-00 2000). The steel reinforcement was then
replaced with FRP bars according to Section 16 ofCHBDC (CAN/CSA-S6-00 2000).
This design resulted in using glass FRP bars (No.l6- 15.9 mm-diameter) in all directions
except in the transverse direction on the bottom mat where carbon FRP bars (No.IO - 9.5 mm-
diameter) were used. The construction of the bridge was completed on September 2002
and opened for traffic during the second week of October 2002. Before opening to the
traffic, the bridge was tested for service performance using standard truckloads as
specified in the new CHBDC (CAN/CSA-S6-00 2000). During all load tests, strains in
FRP and steel reinforcements and deflections ofFRP and steel reinforced spans (slabs
and girders) were recorded. Design, construction details, and the results of this first
series of field tests are discussed in this paper.
37
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Brahim Benmokrane, is an ACI member and NSERC Research Chair Professor in FRP
reinforcement for concrete structures in the Department of Civil Engineering at the
Universite de Sherbrooke, Sherbrooke, Quebec, Canada. He is a project leader in ISIS
Canada Network of Centres of Excellence. His research interests include the application
and durability of FRP composite materials in civil structures and structural health
monitoring with fiber optic sensors. He has been involved in the design, construction,
and monitoring of the first two bridges (Joffre and Wotton Bridges) constructed in
Quebec using FRP reinforcement.
INTRODUCTION
Bridge deck deterioration is one of the most common deficiencies in a bridge system.
Concrete bridge decks deteriorate faster than any other bridge component because of
direct exposure to the environment, de-icing chemicals, and ever-increasing traffic loads.
The magnitude of deck cracking and delamination due to corrosion is a major problem
when measured in terms of rehabilitation costs and traffic disruption. To overcome
corrosion-related problems, the steel reinforcement should be protected from corrosion,
or be replaced with alternative non-corroding materials. One of these alternatives is fiber
reinforced polymer (FRP) composite reinforcement, which has been used successfully in
many industrial applications and more recently has been introduced as concrete
reinforcement in bridge decks and other structural elements. The use of the non-
corrosive FRP composite bars as reinforcement for concrete bridge decks provides a
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Through the NSERC research chair in FRP reinforcements for concrete structures, which
started in 2000 at the Department of Civil Engineering, Universite de Sherbrooke
(Quebec, Canada), a collaboration with the Ministry of Transportation of Quebec (MTQ)
was established to develop and implement FRP reinforcement in concrete bridges. After
Magog Bridge, presented in this paper, is the second application, which is different from
Wotton Bridge. Magog Bridge is larger in size, has different structural system, is
exposed to heavier traffic and harsher environment (located on a highway). Upon the
successful construction of this highway bridge during summer 2002, a field test was
conducted to evaluate the service performance under wheel truckloads before the opening
of the bridge to traffic. This paper presents the design, construction details, and field test
results of Magog Bridge.
RESEARCH SIGNIFICANCE
The paper presents a new and innovative field application of FRP bars as reinforcement
for the concrete deck slab of highway bridges. The new Canadian Highway Bridge
Design Code, which includes a new section on using FRP composite bars, has been used
to design the concrete deck slab. This field application contributes to the
validation/improvement of the existing design codes and guidelines, establishes
construction details, and evaluates the performance of FRP reinforcement under real
service loading and environmental conditions.
The bridge presented in this paper is located over Magog River on Highway 55 North
(Quebec, Canada) in the vicinity of Magog City near the US/Canadian border. The
bridge is a girder type with a total length of 83.7 m and five main steel girders
continuously supported over three spans (Figure la). The two end spans are 26.2 m each
and the middle one is 31.3 m. The deck is a 220-mm thickness concrete slab continuous
over four spans of 2.845 m each with an overhang of 1.352 m on each side (Figure 1b).
Clear concrete covers of 35 and 60 mm at bottom and top, respectively were used. The
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Material Properties
The bridge deck slab was constructed using normal-weight ready-mixed concrete (Type
V MTQ). The concrete compressive strength was 52 MPa, which was obtained based on
the average values from the tests performed on three 150 x 300-mm cylinders. The
tensile strength of the concrete was 3.5 MPa as determined by performing the split
cylinder tests. Two hundred and sixty cubic meter of concrete were cast into the bridge
deck slab through a 12-hour period (started at 7:00 p.m. and finished 7:00 a.m. the
following day). The properties of the newly developed sand-coated carbon and glass
FRP bars used in reinforcing the deck slab are listed in Table l.
The design of the bridge deck slab reinforced with FRP composite bars was originally
made with steel bars using the flexural design method according to the new Canadian
Highway Bridge Design Code (Clause 8.18 - CAN/CSA-S6-00). The steel
reinforcement was then replaced by FRP reinforcement according to Clause 16.8.7
(CAN/CSA-S6-00). According to this Clause of the Code, the steel replacement is based
on equivalent stiffness for bottom reinforcement layer ( p sB) and based on equivalent
strength for top reinforcement ( Psr ). The FRP reinforcement ratios, p JB and p fT,
were determined using equations 1 and 2.
IJ>sEs
PjB = -:h£Pss for bottom reinforcement (1)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
'I'J f
IJ>sfy
for top reinforcement (2)
p jT = lj> f ff p sT
where /y, E., and lj> s are the yield strength, the modulus of elasticity, and the resistance
factor of steel reinforcement, jj, E1, and lj> f are the tensile strength, the modulus of
elasticity, and the resistance factor of FRP reinforcement. The resistance factor lj> is
taken as 0.75, 0.85, and 0.9 for glass FRP, carbon FRP, and steel bars respectively
(Clause 16.5.3 and 8.4.6 - CAN/CSA-S6-00). The design moments given in Table 2
were based on a wheel load of 87.0 kN with the associated load factor of 1.7 (Clause
3.5.1 - CAN/CSA-S6-00), a dynamic load allowance of 0.4 (Table 3.5.la- CAN/CSA-
S6-00) and a load combination factor of0.9 (Clause 3.8.4.5.3- CAN/CSA-S6-00).
This design approach led to the use of glass FRP bars (No.16- 15.9 mm) in all directions
except in the bottom transverse direction where carbon FRP bars (No.10- 9.5 mm) were
This FRP reinforcement covered the deck slab of the first span (26.52 m long). The deck
slab of the other two spans was reinforced with galvanized steel reinforcement (No. 15M
@ 160 mm in the transverse direction and No.l5M @ 240 mm in the longitudinal
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
direction). Table 3 gives the reinforcement configurations and ratios, which were similar
at the top (0.82 and 1.64 % for steel corresponding to 0.87 and 1.87 % for GFRP).
However, the carbon FRP reinforcement ratio at the bottom (1.34%) was approximately
twice that of steel (0.70 %). The total amount ofGFRP (No.l6) and CFRP bars (No.IO)
used in this project was 8855 and 13911 m, respectively.
Instrumentation
The bridge was well instrumented at critical locations for internal temperature and strain
data collection. Instrumentation was distributed on 4 sections along the bridge. Sections
1-F and 2-F at quarter and middle of the FRP reinforced end span, respectively and
identically sections 1-S and 2-S at quarter and middle of the steel reinforced end span,
respectively (F: FRP; S: Steel). A total of 35 Fabry-Perot fiber optic sensors (FOS) were
used. Twenty-three FOS were glued on reinforcing FRP and steel bars at all 4 sections.
Two FOS were embedded in concrete at section 2-F at the level of top and bottom
reinforcements. Ten FOS were glued on the surface of steel girders at sections 2-F and 2-
S. Furthermore, four wire thermocouples were embedded in concrete on top and bottom
of sections 2-F and 2-S to measure temperature changes (Figure 2). In addition, during
load testing, deflections of concrete slabs and girders were measured using a system of
rulers (located on sections 2-F and 2-S) and theodolites. For the deck slab, these rulers
were installed at mid-span between girders. This instrumentation enables the acquisition
of data and long term monitoring of the bridge behaviour under service loading and
environmental conditions. The selected instrumentation will allow a direct comparison
between the behaviour of the composite material reinforcement and the steel
reinforcement under identical traffic and environmental conditions.
Bridge Construction
Figure 3 shows photos of the bridge during construction. The construction crew reacted
positively saying that more FRP bars could be handled and placed in less time due to
their lightweight. Continuous plastic supports were placed underneath the bottom
reinforcement parallel to the longitudinal direction at spacing of 1.0 m to maintain a clear
The bridge was tested for service performance as specified by the new CHBDC
(CAN/CSA-86-00) on October 81h, 2002. The bridge was tested under static loads using
three truckloads (two trucks and one trailer) with three axles each as shown in Figures 4
and 5. During testing, 32-channel FOS data acquisition systems (DMI-32) were used to
collect readings from FOS at a rate of 10 readings/sec. After testing, this system was
permanently installed underneath the bridge for long-term remote monitoring. During
testing, deflections of concrete slabs and girders were measured using a system of rulers
and theodolites (Figure 3). The temperature of the concrete slab was measured twice
during testing, at 12:15 p.m. and 3:30 p.m. The measured temperatures are listed in
Table 4.
Five paths in the direction of traffic (A, B, C, D, and E) were marked on the bridge as
shown in Figure 6. Fifteen stations (truck stops) were also marked along the longitudinal
direction of the bridge at selected positions to give critical loading cases at the
instrumented sections of the bridge. The test was carried out using a single truck (No. 1)
pulling the trailer over the 5 paths and the two truck (No. 1 and 2) simultaneously along 2
paths such that total of 105 (15 stations x 7 paths) readings were recorded for each gauge
and deflection ruler. At each truck stop (station), readings collected from instrumentation
were recorded when the midpoint of the second axle (first rear axle) of the truck was
directly over such station.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
TEST RESULTS
Figures 7 and 8 show maximum measured tensile strains in FRP reinforcing bars against
truck position along the bridge. In these figures, the zero value on the horizontal axes
represents the location on the bridge that contains the gauge under consideration. The
strain values depend on the case of loading, namely truck position and path. Therefore,
for each graph, the truck path that gives the maximum strain readings is considered. Due
to the chosen truck paths, it is predicted to have the maximum strains in bottom FRP bars
in the case of one truck and the maximum strains in top FRP bars for two trucks. Actual
results in case of a single truck with a trailer showed maximum measured strains in
bottom and top FRP bars (path C) of 19 and 4 micro-strains, respectively (Figure 7). It
should be noted that there are two peak values of the measured strain due to the influence
of the trailer loads when the first rear axle is not directly over the location of the gauge.
In case of two trucks (path C-E), as predicted, the maximum measured strains in bottom
Figures 9 and 10 show comparisons of the measured strains in FRP and steel
reinforcement at identical positions in the top and bottom reinforcement mats. The
measured strains in both types of reinforcement were similar. The maximum measured
stain in top steel and FRP bars were 10 and 9 micro-strain respectively, while these
values were 14 and 18 micro-strain, respectively, for bottom reinforcement. These
maximum measured strain values in FRP bars are less than 0.16% of the ultimate strain
of the material.
Strains in Concrete
For the same truckload path as in Figure 8, the corresponding values for strains in
concrete (measured by the two FOS embedded in concrete) were -7 and +4 micro-strain
at top and bottom, respectively as shown in Figure 11. Considering a concrete cover at
top and bottom of 60 and 35 mm respectively, and using simple bending theory, it can be
shown that the tensile strain at the top and bottom surfaces of concrete reached a
maximum of 16 and 27 micro-strains respectively at these gauge locations. These strain
values at the concrete surface of the deck slab are well below the cracking strain of
concrete, Ecr = 153 micro-strain (for fc' =52 MPa and Ec = 34 GPa).
Figure 12 shows the maximum measured tensile strains in the bottom flange of steel
girder C at locations 2-F and 2-S (approximately 14 m from each end of the bridge)
against truck position along the bridge. It should be noted here that the steel girders are
continuous over three spans and the strains are measured at the middle of the two end
spans. Therefore, when the truckloads moved along the bridge and came on the middle
span, compressive strains were measured on the steel girders. The patterns and
maximum measured values of strain, for the same case of loading, at both FRP and steel
sides were very similar. The measured peak strains were -16 to +59 and -32 to +115
micro-strain on path C (single truck with the trailer) and Path C-E (two trucks),
respectively.
Deflection Measurements
During static tests, deflections of concrete slabs and steel girders were measured by a
theodolite using a system of rulers installed at the middle section of the two end spans (2-
F and 2-S). The maximum measured deflections at both sections (2-F and 2-S) under the
same loading condition (same truckload and path) were very similar. Using the single
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
truckload with the trailer on Path C, the measured peak deflection of the steel girder C
was -1 to +4 mm as shown in Figure 13. This peak deflection was -2 to +6 mm for the
two-truck case travelling simultaneously on Path C-E.
The deflection of the concrete deck slab was calculated by subtracting the measured
value at the considered position on the slab from the average of the measured values at
CONCLUSIONS
This paper presents the design, construction details, and test results for Magog Bridge on
highway 55 North, Quebec, Canada. This bridge is the second in a series of field
application research projects using FRP composite bars as reinforcement for concrete
structures, especially bridge decks, conducted through the NSERC industrial research
chair at the Civil Engineering Department, Universite de Sherbrooke. Based on the
construction details and the results of the first field-loading test, the following
conclusions can be drawn:
1. No obstacles to construction were encountered due to the use of the FRP bars. The
FRP bars withstood normal on-site handling and placement with no problems. In
addition, due to its light weight, the bars were easy to carry and could be placed
with less effort.
2. The measured strains on steel and FRP bars were very similar.
3. Maximum tensile strain values in concrete were very small, 16 to 27 micro-strain, as
the truckload moves over the gauge. These strains are well below cracking strain
for concrete, which is in the range of 150 micro-strain for normal weight concrete
with compressive strength of 52 MPa (Ec = 34 GPa).
4. As tested, the maximum tensile strain in glass and carbon FRP bars was 9 and 19
micro-strain, respectively. These values are less than 0.07/0.16% of the ultimate
strain of the glass and carbon FRP material and suggest that the use of the flexural
design method of CHBDC overestimates the calculated design moments (service
and ultimate).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
5. Deflections of the bridge deck and slab were well below AASHTO and CHBDC
allowable limits. During all stages of testing, the maximum measured deflection for
steel girders and concrete slabs never exceeded 6 and 2 mm, respectively.
RECOMMENDATIONS
Based on the field test results of the Magog and Wotton Bridges, and the empirical design
method (Clause 8.18.4 - CAN/CSA-S6-00 2000), concrete bridge deck slabs supported
on girders with span-to-depth ratio less than 15, should have FRP reinforcement ratios of:
0.6% of carbon FRP reinforcement in the bottom transverse direction, and 0.6 %of glass
FRP reinforcement in all other directions. This is valid for carbon and glass FRP
composite bars with a minimum modulus of elasticity of at least 1I 0 and 40 GPa,
respectively. These recommendations are in good agreement with test results of other
researchers (Hassan et al. I 999).
The authors wish to extend special thanks to the following organisations and individuals
for their contribution to this project:
The Ministry of Transportation of Quebec, Gerard Desgagne, Structural Division
Head, bridge owner, Quebec City, Quebec,
Le Groupe Teknika Inc., Dominique Nadeau, consultant of the project, Sherbrooke,
Quebec,
Pultrall, Division of ADS Composites Group, supplier of the composite material
reinforcement ISOROD™, Thetford Mines, Quebec,
Roctest Ltee, Marco Quirion, supplier of the fiber optic sensors and its data
acquisition system, St. Lambert, Quebec,
The Natural Science and Engineering Research Council of Canada (NSERC), partial
finance, Ottawa, Ontario,
The Network of Centres of Excellence ISIS-Canada, partial finance, Winnipeg,
Manitoba,
Department of Civil Engineering, Universite de Sherbrooke, Franyois Ntacorigira
and Simon Sindayiagaya, technicians.
REFERENCES
1. ACI 440.1R-Ol, (2001) Guide for the Design and Construction of Concrete
Reinforced with FRP Bars, American Concrete Institute, Farmington Hills,
Michigan, 41 p.
2. Benmokrane B., and Rahman, H., editors, (1998). "Durability of Fiber Reinforced
Polymer (FRP) Composites for Construction," Proceeding of the First International
Conference, Sherbrooke, Quebec, Canada, 692p.
3. Benmokrane, B., Rahman, H., Mukhopadhyaya, R., Masmoudi, R., Zhang, B., Lord,
I., and Tadros, G., (2001) Fiber-Optic Sensors Monitor FR-Reinforced Bridge, ACI
International, American Concrete Institute, Vol. 23, No.6, Detroit, USA, pp." 33-38
4. Benmokrane, B., Zhang, B., Laoubi, K., Tighiouart, B., and Lord, I., (2002).
"Mechanical and Bond Properties of New Generation of Carbon Fibre Reinforced
Polymer Reinforcing Bars for Concrete Structures," Canadian Journal of Civil
Engineering, Vol. 29, No.2, April, pp. 338-343.
5. Benmokrane B., and El-Salakawy, E., editors, (2002). "Durability of Fiber
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
tifli
<t>Movable <Traffic Direction <fJFixed Movable
r-------l---·~~--- l
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
® @
I
© ® I
®
14083
(b) Cross-section
Figure 1. Layout of the bridge: (a) Plan view and (b) Cross-section
Figure 2. Fiber optic sensors glued on the surface of the FRP bars and steel girders
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 3. Bridge deck construction
Rear axis of Truck 1
2050
2700
11~mm
II~I
11~1m
II~I
1850 II~ II
Figure 4. Truckloads
~
1r· . o.·.··"·······,···r< ·····r·~u Path A oo oo
I·~-® =· I =· I
©
,~. ® =· @
J
11···r Pathi., . . l
l-, - 6 - • - • - ·-
11·············-I. 8
Path C
ld. ···d····I:·--·-···1
oo oo
l-, - 6 - • - • - · -
~
oo oo Path D
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
-e
---- FRP BOTTOM
I! 4125 -+-FRPTOP
f l)
I 4120
u 4115
-E
r::
·-....
4110
-ca
en
4105
4100
-7 0 7 14 21 28 35 42 49 56 63 70 77
Distance along the bridge (m)
Figure 7. Maximum tensile strains in FRP bars (One truck with a trailer- Path C)
4130
-I!
r:: ~ -FRPBOTTOM
-e
·- 4125 1-----
'
--+-FRPTOP
cp 4120
j ~\
.~
E
~ 4110
·-~
4115
v ~ --`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
4105
en
4100
-7 0 7 14 21 28 35 42 49 56 63 70 77
Distance along the bridge (m)
Figure 8. Maximum tensile strains in FRP bars (Two trucks- Path C-E)
-l! 18
en
e
-~
I
16
14
-FRPTOP
--+-STEEL TOP
.§. 12
c
"'-
10
"i! 1\
.5
8
6
"'
71 71
Cl)
4
7 1 I
0)
c
co 2 7 \ I
..c 1/ _,_. J
......
0
0 -- 7
0
-7 14 21 28 35 42 49 56 63 70 77
Distance along the bridge (m)
Figure 9. Comparison between strains in FRP and steel bars
(Top reinforcement- Two trucks- Path C-E)
-c
-e
"i! 20
18
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
'f -FRPBOTTOM
16
-~ 14
-E
c
12
-"'
"i! 10
.5 6
Cl)
8
0) 4
c
co 2
..c
0 0
-7 0 7 14 21 28 35 42 49 56 63 70 77
Distance along the bridge (m)
Figure 10. Comparison between strains in FRP and steel bars
(Bottom reinforcement - One truck with a trailer - Path C)
-
·-
l! 4114
cp 4112
e 411o ~
,/
;--.....
'-
_r.....::::.
-Concrete TOP t-
\
- ~
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
r
'
.~
E 4108 I
~ 4106
-
·-
l! 4104
t n 4102
4100
)J
-14 -7 0 7 14 21 28 35 42 49 56 63 70
Distance along the bridge (m)
Figure 11. Maximum strains in concrete (Two trucks - Path C-E)
4300
-'i!
r::
4275 +-------trJ~rr-------1- FRP Side f------1;E----1
-e
If
.2
4250
4225
4200
+--~t-------J-+- Steel Side r-------c!i!'fW-----1
-'i!
E
r:: 4175
-
tn
4150
4125
4100
0 14 28 42 56 70 84
Distance along the bridge (m)
Figure 12. Peak strains in the bottom flange of steel girder C (Two trucks- Path C-E)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: Aspects of the design and installation of a novel carbon fibre reinforced
polymer (CFRP) prestressed high strength concrete lighting column (Carbolith®) are
presented. The tapered cylindrical columns have a nominal height of 8 m and contain an
opening above the foundation to allow for the insertion of the lamp fuse box. The
bending/torsion behaviour of a total of five full-scale prototype columns was tested in
accordance with the relevant European standards (EN). In the experimental programme,
the location of the fuse box opening relative to the loading direction was varied. All five
poles fulfilled the EN serviceability and ultimate limit state requirements for lighting
columns in pedestrian and/or low speed lightly trafficked areas. This successful outcome
has lead to the first field application of the CFRP prestressed concrete lighting columns.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
55
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
56 Terrasi and Lees
Giovanni P. Terrasi received his PhD from ETH Zurich in 1997. He is currently the
head of the R&D/Engineering Department at SACAC AG in Lenzburg, Switzerland and
has a particular interest in the behaviour of high performance concrete reinforced or
prestressed with fibre reinforced polymer tendons.
Janet M. Lees received her PhD from the University of Cambridge in 1997. Since 1998,
she has been working as a University Lecturer in Structural Engineering at the University
of Cambridge, UK. Her research interests include the use of advanced composites in
concrete applications and joining techniques for all-composite structures.
INTRODUCTION
In September 2000, the world's first carbon fibre reinforced polymer (CFRP) prestressed
high strength concrete electricity pylon was installed in Switzerland0 l. The 27 m high
pole was manufactured using a centrifugally-cast high strength concrete containing
blended silica-fume cement (see Figure 1).
Building on this successful application of CFRP tendons, the potential of transferring the
technology to lighting columns was identified. In particular, road environments are
known to be very aggressive, causing corrosion at the fixture of plain steel, steel-
reinforced or steel-prestressed concrete lighting posts. It is thus a promising application
for a structure that combines concrete with a non-corrodible, lightweight and high
strength tendon material such as CFRP.
The durability of the CFRP prestressing material is a key advantage in the construction of
slender prestressed concrete cylindrical poles. Steel corrodes and, in many applications,
a significant concrete cover (40-50 mm) is required to protect the prestressing steel from
aggressive internal and/or external environments. In contrast, when durable CFRP
tendons are used, only a relatively small concrete cover (15-20 mm) is required.
Therefore, CFRP-prestressed poles are significantly lighter than equivalent steel-
prestressed structures resulting in lower transportation and installation costs. An
additional advantage is that the non-metallic tendons are expected to have low
maintenance requirements.
The focus of this project was the design and installation of an 8 m nominal height CFRP
prestressed concrete lighting column fulfilling the requirements of the relevant European
Standard EN40 Series<2l( 3)(4 )(Sl. A particular design consideration was the requirement of
an opening above the pole's foundation for the electrical fuse box of the lamp.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
RESEARCH SIGNIFICANCE
The research investigates the performance of CFRP prestressed high strength concrete
lighting columns. The work gives insight into the behaviour of this novel combination of
materials when used in a practical application where conventional steel and steel-concrete
solutions are known to have limitations in terms of their long-term durability
performance.
FIELD APPLICATION
In March 2003, an 8 m high prototype Carbolith® lighting column was erected in the
SACAC AG car park in Lenzburg, Switzerland (see Figure 2).
The tapered high strength concrete pole was prestressed with six CFRP tendons
(prestressed to 60% of the ultimate tendon stress, fu) and contained no shear
reinforcement. It was thus directly comparable to two of the prototype poles tested in the
experimental programme that will be described later in this paper. A 1.2 m deep hole was
dug and the pole was lifted into place using a light crane. A 0.4 m x 0.4 m x 1.2 m long
concrete block was then cast at the foot of the 9.2 m pole (resulting in a final height of 8
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
m) and acted as the foundation. The pole weight was only 350 kg which represents 70%
of the weight of a conventional steel reinforced concrete lighting pole with the same
nominal height. The ground was then compacted around the foundation block and
additional mortar used to secure the pole in position.
The total material, production and installation costs were equal to that of an equivalent
steel-prestressed concrete pole. Thus the poles are expected to be a competitive
alternative for lighting in pedestrian and lightly trafficked areas. Further work on the
performance under vehicle impact will investigate the suitability for use adjacent to
higher speed traffic.
The basic configuration and dimensions of the prototype 9.2 m (8 m nominal height)
tapered cylindrical lighting columns (henceforth referred to as 'poles') are shown in
Figure 3. A Quadralux-G lantern type was assumed since this is one of the most common
luminaries for street lighting columns in Europe. The position of the fuse box opening, at
1500 mm from the end of the fixture, and the opening geometry (height = 300 mm, width
= 75 mm) were chosen to reflect typical fuse boxes used in central Europe. The pole taper
of 10 mm/m results in a variation of the outer diameter from 120 mm at the tip to 212
mm at the foot, the average wall thickness of the pole being 40 mm ± 10 mm (measured
at the tip and foot surfaces).
The poles were centrally prestressed with six CFRP tendons of diameter 4 mm (the
tendon material properties are shown in Table 1). The tendons were pultruded rods and
had a ceramic coating on the outer surface to improve the bond properties. The initial
prestressing force of each CFRP tendon was 16.6 kN (crpo = 1'320 N/mm2) corresponding
to a total central prestressing force in the pole of 100 kN (at prestress transfer). It is of
note that due to the requirement of a fuse box, the tendons were deviated in this region.
The opening was defined and reinforced by a stainless steel casing consisting of two
deflector-beams to deviate the two tendons running through the fuse box region (see
Figure 3).
Three different types of shear reinforcement were adopted for pole numbers 3, 4 and 5
(there was no shear reinforcement in poles 1 and 2). In pole 3, a PVA geogrid (180
kN/m-transverse, 180 kN/m-longitudinal, 90/0°) was used apd in pole 4, an aramid-fibre
geogrid ( 150 kN/m-transverse, 30 kN/m-longitudinal, 90/0°) was incorporated. In both
cases, the geogrids (used typically for slope stabilisation) had a mesh size of 30 mm, and
were wound and fixed over the prestressing reinforcement over a length of either 3.8 m
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
SPECIMEN DESIGN
Applied Loads
The most relevant loading condition for the lighting columns was the wind. A terrain
category II (which is for flat areas with occasional small structures) and a basic wind
velocity at 10 m above sea level (30 m/s) were assumed. Using a mean shape factor of
0.52, the resulting average unfactored wind pressure on the pole was 600 N/m2• Based on
the manufacturer's data, the side-wind pressure surface area of the lantern was 0.14 m2
and the front wind pressure surface area was 0.1 m2 • Using a shape coefficient of 1.0, the
resulting unfactored wind pressure on the lantern was calculated in accordance to EN 40-
3-1 <4l to be 1'325 N/m2 •
A partial wind load factor of Yr = 1.4 (for transient wind gust action) was used<4l and the
partial load factor for the self-weight was taken to be y0 = I .2.
Wind pressure acting on a cantilevered pole will result in flexural moments relative to
any given longitudinal axes. Furthermore, because of the lantern eccentricity, when the
wind blows on the side of the lantern, the pole will be subjected to an additional torsional
moment. Therefore, the experimental programme included an investigation of both the
pure torsion and the bending/torsion response of the poles.
The pole was designed to be fully prestressed at maximum service load in order to limit
deflections: The formation of bending cracks is to be avoided so that the moment of
inertia of the entire cross section is available to sustain the service moment. The
horizontal service deflections of the pole were determined by considering the unfactored
(y = 1.0) characteristic design loads (self-weight and wind<4l). In the pre-cracked stage
the centrally prestressed pole behaved as an elastic tapered cantilever beam (of 8 m free
length). The tip deflection under wind load (service load) was limited to a calculated
value of 20 mm.
The ultimate bending moments of resistance were predicted using an inelastic analysis
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
based on the short-term stress/strain curves of CFRP and high strength spun concrete.
The characteristic material strengths were reduced by appropriate partial material safety
factors ( 1.15 for the CFRP tendons and 1.3 for the dense spun concrete). As in classical
prestressed concrete analysis, the equilibrium of internal forces and moments was
considered assuming perfect bond and the principle of plane sections remaining plane.
Since the section geometry and concrete prestress varied along the length, each pole
cross-section was considered separately. The combined interaction of X-X flexural
moments (where the fuse box was on either the compressive or tensile face), Y -Y flexural
moments (fuse box on the neutral axis) and torsional moments was taken into account.
STATIC TESTS
General Arrangement
The five prototype specimens (poles 1-5) were tested in Cambridge, UK. Each pole was
fixed to the strong floor by prestressing the end foundation block (over the fixture length
of 1200 mm) using three spreader beams and threaded tensioning bars (see Figure 4). The
total pressure on the foundation block was 45 tonnes.
In order to test in uni-axial horizontal bending (combined with torsion), all the pole
specimens were centrally supported to 'compensate' for the self-weight moment in the
vertical plane. The vertical support was located at 3.2 m from the pole tip and consisted
of a trolley with wheels that was capable of following the pole deflection. A serial load
cell was integrated into the purpose-built trolley support to monitor the reaction force.
The off-axis tensile load was introduced at the tip of the pole by means of a specially
manufactured steel clamping rig.
To verify the integrity of the torsion clamping system and to investigate the torsion-only
behaviour of the pole, a test was carried out on pole 1 where a tubular steel bar (of mass
4.2 kg) was welded to the clamping device fixed at the pole tip, giving an effective
torsion lever arm of 2.0 m. The torsion load was introduced incrementally by adding
standard 2-kg weights on a pan hanging on the end of the steel bar (to a maximum total
torque of 280 Nm). For every load step, the corresponding torsion angular displacement
was determined by reading an inclinometer bonded on the pole at 65 mm from the tip.
The resulting vertical bending moment was negligible because of the limited vertical
testing load (with a maximum value of 12 kg) acting on the pole.
Bending/Torsion Tests
In the main bending/torsion cantilever tests, the fuse box opening was positioned on the
compression edge for poles 1 and 3 (M+x), along the neutral axis for pole 2 (My) and on
the tension edge for poles 4 and 5 (M-x) (see Table 2).
The poles had an effective bending lever arm of7.99 m and a torsion lever arm of0.14 m
(the load eccentricity was imposed using the clamping rig). The clamping rig at the pole
tip was pulled horizontally using a threaded rod (with a pitch of 1.5 mm/rotation)
connected to a 125 W electric motor, that transferred its rotation onto a MlO nut via a
chain-gear (see Figure 4). The electric motor was driven by a potentiometer, from where
the rotation (loading) speed and the loading direction could be controlled by the testing
engineer. The pulling load was measured and controlled by a tension load cell which was
connected via a logger to a data monitoring PC.
The torsion angular displacement was the only data measured by hand using a spirit level
inclinometer bonded over the tip of the pole (mid-point at x = 7.935 m from the fixture).
The measurement of the crack pattern (crack widths, lengths and positions) was carried
out at a reference test load of 1.2 kN by means of a crack magnifying glass and a crack
meter.
The serviceability and structural test loads for each verification test were determined by
considering the characteristic lantern dead load and wind loads<4l. The serviceability test
loads were calculated using the unfactored characteristic wind pressures (Yr = 1.0) for the
pole shaft and lantern under consideration. By multiplying the serviceability test loads by
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Torsion
No torsion cracks were visible over the load range considered in the torsion-only tests
and, accordingly, the experimental torsion moment vs. angle relationship was linear
within the measurement precision (0.005°) of the inclinometer used.
In Figure 6, the measured response of pole 1 is compared with the predicted linear elastic
pure torsion behaviour of a vertical, 10 mm/m tapered cantilever pure concrete beam
(uncracked high strength concrete) with round cross section, outer diameter of 200 rnm at
fixture and constant wall thickness of 40 mm loaded with an eccentric point load. The
fuse box opening and stainless steel casing were not taken into account in the calculation
model.
The measured torsion angle of 0.135° at the maximum test load (T = 276.6 Nm) is 12.7%
lower than the calculated value of 0.155°. The slightly higher experimental stiffness can
be attributed in part to a higher effective shear modulus for the pole than the assumed
average value of G = 16.4xl09 N/m2 computed using a Poisson's ratio of v = 0.175 and
an average Young's modulus of concrete Eco = 38'550 N/mm 2 (at low stresses Eco is
higher than the average value). Differences in wall thickness t (the taper of the production
mould results in t > 40 mm in the pole foot region and t < 40 mm in the pole tip region)
may also play a role.
No slippage was observed in the clamping rig designed to apply the torsion load to the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
pole, which confirmed that the rig was suitable for use in the subsequent cantilever
bending/torsion tests.
Static Bending/Torsion
The main results of the bending/torsion cantilever tests of poles 1 - 5 are shown in tables
2 and 3.
The experimental cracking loads (between 0.80 and 1.0 kN) were above the serviceability
test loads calculated for the poles according to EN 40-3-1 (4) (where F,e,x,-x = 0.56 kN, F,e/
= 0.506 kN). The poles could therefore be confidently considered to be fully prestressed
under serviceability conditions. Furthermore, there was ample reserve load capacity after
The un-loading of the poles from the service test load gave rise to very low residual
deflections, in the range of 1-3 mm. These values were significantly lower than the EN-
40-3-2 allowable limits which correspond to residual deflections of between 8.5 and
10.36 mm depending on the particular pole considered (the code states that the residual
deflection after removal of the service test load should be less than 20% of the defection
caused by the test load).
The deflections started to increase rapidly after cracking of the pole occurred and the
rotations concentrated in many thin and well distributed cracks near the fixture and the
opening region of the pole. A crack mapping in the tensile region of each pole was
carried out at a reference load of 1.2 kN. The fairly good bond between the ceramic
coated tendons and the concrete matrix led to the formation of between 4 and 10 thin
vertical bending cracks (of widths 0.01 - 0.20 mm) at regular spacings near the fixture
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
and opening regions of the pole (see Figure 7). All the observed cracks were vertical
bending cracks and had a maximum width of about 0.5 mm.
When un-loading the specimens from the EN minimum ultimate test load (cracked stage),
the bending cracks closed (most of them were no longer visible) and consequently the
deflections recovered almost completely (almost perfect elastic behaviour). Therefore
only a modest amount of energy dissipation took place during the loading/un-loading
cycle.
The deflected bending shapes of pole 3 at various stages of loading are shown in Figure
8. Large deflections were noted in the later stages of testing and the tip deflection was
around 925 mm immediately prior to failure. The high rotation capacity came from the
The pole failure loads (between 1.56 and 1.89 kN) were considerably higher than the EN
minimum required ultimate loads of Fumin,x,·x = 0.95 kN, Fumin,y = 0.86 kN( 4) which
indicates a factor of safety in excess of the minimum requirement. The experimental
failure moments were reasonably close to the predicted design moment capacities,
computed under the assumption of pure bending along the plane considered and including
material partial safety factors. In Table 3, both the predicted and experimental ultimate
moment capacities at selected cross-sections, and the corresponding tip loads (shown in
parentheses) are indicated. The fuse box opening location for each of the tests may be
found in Table 2.
Several failure modes were observed, depending on the testing position (bending plane
considered) and on the reinforcement configuration (see Figure 10). Poles l and 2 failed
by the sudden snapping of the CFRP tendons on the tensile edge at cross sections near the
fixture (the corresponding moments are Mu+x (pole l, x = 0.28 m) = 12.6 kNm, Muy(pole
2, x = 0.05 m) = 14.2 kNm). However, while the failure for pole 2 was expected to be
near the fixture (Figure lO(a)), pole 1 was predicted to fail in the concrete at the lower
edge of the fuse box opening, which was not the case, possibly because of the reinforcing
effect of the stainless steel casing that was neglected when determining the pole
capacities and the inherent material partial safety factors. Pole 3 (with the opening on the
compression edge) contained additional longitudinal passive reinforcement in the form of
the PV A fibre bundles. This reinforcement acted to increase the tensile capacity of the
cracked pole in the area of the opening and at the fixture and shifted the failure from the
tensile to the compressive edge. The result was a combined failure with buckling of the
steel casing and concrete crushing at a height 1.5 m from fixture (Figure lO(b)). The
experimental failure bending moment was 12.3 kNm, which is 23% higher than that
calculated for the HSSC crushing at the lower edge of the opening (Table 3). Again, the
difference is attributed to the reinforcing effect of the steel casing and material safety
factors. Poles 4 and 5 were tested with the box opening at the tensile flange, therefore the
bending resistance of the area around the lower opening edge was reduced since the two
CFRP tendons running next to the tensile edge were deflected inwards. Consequently,
poles 4 and 5 failed by the sudden snapping of the CFRP tendons on the tensile edge at
cross sections at x = 1.23 m and x = 1.5 m from the fixture respectively. The failure loads
and failure locations were reasonably close to the predicted values taking into account the
offset of the tendons.
In general, the design computation models gave fairly good predictions of the ultimate
load capacities.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
to the experimental values. However, the design tip deflection at the ultimate load of
1 '304 mm, was considerably higher that the measured values which ranged from Outip =
474.8 mm (pole 4) to 920.8 mm (pole 3). The reasons for this overestimate of the
deflection at failure were most likely due to the underestimation of the tensile strength of
the HSSC and the neglection of the tension stiffening effect of the concrete between the
bending cracks in the original design model.
In these equations, E1 and Es are the modulus of elasticity of the FRP tendon and steel
respectively, Mer is the cracking moment and Ma is the maximum moment at the section
where the deflection is considered. The pole cross-section tapers along its length, so to
simplify the calculation, the ratio of the cracked moment of inertia fer to the gross
moment of inertia Ig was taken to be constant throughout the cracked region. The total tip
deflection was then determined by integrating the cracked curvatures Kc = M/EJe (which
takes into account tension stiffening) from the fixture to the end of the cracked region
(xcr) and then integrating the uncracked curvatures (Ke = MIEJg) from Xcr to the pole tip.
The resulting curve has been plotted in Figure 11 and the analysis shows very good
agreement with the experimental deflections.
For the five pole specimens, no slippage (draw-in) of the tendons, even at the maximum
load, was recorded by the LRDTs monitoring the tendons' end-surfaces at the end of
fixture (Figure 5). Transfer tests on a plate specimen with similar 0 4 mm CFRP
tendons and a concrete cover of 18 mm suggested that for a prestressing level of 1000
N/mm2 the transfer length was only 70 mm (17x 0cFRP). Therefore the foundation length
of 1.2 m was more than sufficient to anchor the tendons.
CONCLUSIONS
(l) The full prestressed design of the CFRP prestressed high strength spun concrete
lighting column was verified by testing five specimens in accordance with EN-40-3-2<4l.
The serviceability, structural and minimum ultimate load requirements were fulfilled by
all the pole specimens tested in cantilever bending/torsion.
(2) The poles showed sufficient bending rotation capacity to make up for the lack of
plasticity of the brittle CFRP prestressing tendons. This high rotation capacity in the
cracked state (in the load range above the maximum service loads) led to the attainment
of a high deflection at failure, with a medium amount of well distributed and thin bending
cracks, which can be interpreted as a warning of the pending failure of a pole when
loaded to an unforeseen high load level.
(3) The static calculation procedures developed for planning and predicting the
behaviour of the prototypes were in accordance with the relevant standard EN40 series
and proved to be adequate and mostly conservative mathematical tools for the design of
the novel lighting columns. The deflections calculated using the draft ACI 440 I
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(4) The three different types of shear reinforcement had no practical effect on the
static behaviour of the cantilever specimens in the bending/torsion moment ranges
achieved in the tests. Nevertheless the use of a 0-90°-biaxial geogrid reinforcement
(made of PVA fibres) proved to be advantageous giving a higher failure safety by
increasing the ultimate load of the pole and allowing for a certain rest load carrying
capacity after reaching the peak testing load.
(5) The ultimate capacity was roughly the same regardless of whether the fuse box
was on the compressive or tensile flange. A slightly higher capacity was obtained when
the fuse box opening was located on the neutral axis. Except where the PV A geogrid was
used, all the poles failed due to tendon rupture.
(6) It is expected that the poles will provide a durable and cost-effective alternative
to steel-prestressed poles for use in pedestrian and low speed traffic areas.
The authors would like to thank Dr. C.J. Burgoyne for his valuable advice regarding
design and testing matters. The help of the director of SACAC AG,
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Mr. G. Biittig, of several SACAC engineers, and of the company's laboratory and
production teams involved in the manufacture of the specimens, is greatly appreciated.
We are also grateful to the technical staff of the Department of Engineering, University
of Cambridge for their important contribution in the design of the experiments and
carrying out the testing programme.
REFERENCES
[I] Terrasi G. P., Biittig G., Bronnimann R., 2001: "High-strength spun concrete
poles prestressed with CFRP", FRPRCS-5 "Non-Metallic Reinforcement for
Concrete Structures", ISBN 07277-3009-6, University of Cambridge, England
16-18 Jul2001, Ed. C. Burgoyne, pp. 1103-1112
[2] CEN/TC 50 "Lighting columns and spigots": European Standard EN 40-1: 2000:
Definitions and terms, European Committee for Standardization CEN, Brussels
[3] CEN/TC 50 "Lighting columns and spigots": European Standard EN 40-2: 2000:
General requirements and dimensions, European Committee for Standardization
CEN, Brussels
[4] CEN/TC 50 "Lighting columns and spigots": European Standards EN 40-3-1:
2000, EN40-3-2: 2000, prEN 40-3-3: 1999 (final draft): Design and verification,
European Committee for Standardization CEN, Brussels
[5] CEN/TC 50 "Lighting columns and spigots": European Standard EN 40-4: 2000:
Specification for reinforced and prestressed concrete lighting columns, European
Committee for Standardization CEN, Brussels
[6] ACI 440 I "Guidelines for prestressing concrete structures with FRP tendons"
Draft document, Feb 2003.
D
~
2 My 0.50 0.86 42.6 0.86 1.79 CFRP
'
D
~
4 M_, 0.56 0.95 47.7 0.90 1.56 CFRP
~
5 M_, 0.56 0.95 51.2 1.0 1.64 CFRP
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
"'1"1
Table 3: Predicted versus experimental load capacities (shaded boxes denote critical sections)
!!:
CL.
Height Pole Po Pole 1 (M+x) Pole 2 (My) Pole 3 (M+x) Poles 4 & 5 (M_x) ):II
from dia (MPa) ultimate moment ultimate moment ultimate moment (applied ultimate moment "CS
"CS
fixture (mm) (failure load) (failure load) failure load) (applied failure load) I
n·
(m)
=
:::!'.
Cl
=
-
(2.1 kN) (2.0 kN) (2.3 kN) (2.1 kN)
en
1.8 182 5.57 9.2kNm 11.9 kNm 9.6kNm lOJJkNID Cl
(1.5 kN) (1.9 kN) {MkN) (1.8kN) I
HSSC HSSC CFRP "'1"1
1.5 185 5.43 9.5kNm 12.1 kNm W.OkNm 113kNm n.2kNm 10.6kNm :a
(1.5 kN) (1.9 kN) (l.5 kN) (1.89kN) (L7kN)
""C'
(1.64kN} I
.CFRP :a
-...=
HSSC HSSC HSSC CFRP
1.23 10.6kNm !!!.
Not for Resale, 08/10/2015 01:23:41 MDT
Licensee=UNI OF NEW SOUTH WALES/9996758001
(1.56kN)
Cl
CPRP I
1 190 4.43 12.6kNm 12.1 kNm 14.5 kNm 12.6kNm I n
CD
(1.8 kN) (1.7 kN) (2.1 kN) (1.8 kN)
0.5 195 4.29 12.9kNm 12.4 kNm 15.0kNm 12.9kNm
3
CD
=
I
(1.7 kN) (1.7 kN) (2.0 kN) (1.7 kN)
0.28 12.6kNm r.-
(1.64kN)
I n
CFRP =
en
0 200 4.15 13.3 kNm
(1.7 kN)
l2.8kNm
(1.6kN)
<:Fro>
14.2kNm
(1.79kN}
CFRP
x=0.05..!!!._
15.5 kNm
(1.9 kN)
13.3 kNm
(1.7 kN)
L . _ __
-
CD
en
c
e:
Daneshlink.com
-
CD
en
en
=
70 Terrasi and Lees
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
I 37.l
300
37.5
A
+-- tapered pole y
(10mm/m) J.
9200 185
14 ~I
75
CFRP tendons
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
E' 300
-
~
c:::
Q)
E
0
250
200
E
c::: 150
0
-~
100
.9
"C
.~ 50 ......... measured angle (inclinometer}
c. - linear elastic torsional response
a. 0
m
0.00 0.05 0.10 0.15 0.20
0
torsional angle at tip (x = 7.965 m} [ ]
-E
c:
0
:;:::;
(.)
600
500 9
F=1.4 kN
F=1.0 kN
400 ~E F=0.6 kN
~
"0 300 • F=0.2 kN
200
100
0
0 2 3 4 5 6 7 8
position from pole fixture (m)
Figure 8: Deflection profiles of pole 3 plotted at selected loads.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 10: Pole failure modes (a) tendon failure- pole 2 (b) steel buckling/concrete
crushing- pole 3.
2
1.8
.. . .. .
1.6
~ 1.4
z
..lO:
~ 1.2
113
..Q
....
Q)
1;:~1§~::r'::...._----_.:::__.._ ____l • • pole 4
> - - pole3
~
:;:::> 0.8 +--11~-'-------"----.---------1 - - - - pole 2
c::
113
0
0.6 i+;-=:::::-r=:;-----..----~ - pole 1
design
0.4 • • ACI 440 I- eq. (4-1)
-,.I
0.2 "' ""
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0
0 200 400 600 800 1000 1200
tip deflection [mm]
Figure 11: Load vs. tip deflection behaviour of the poles 1-5 tested in cantilever bending/
torsion.
Synopsis: This paper describes the use of Carbon Fiber Reinforced Polymer, CFRP,
tendons and rods for prestressing concrete highway bridges completed in 1993 and 1997.
Due to the lack of design codes, the paper presents briefly the research work undertaken
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
before the final design of the two bridges. The first bridge is a 75ft. (23.85 m) span
skew bridge, which consists of bulb-tee pre-tensioned girders made continuous with post-
tensioned steel strands. Four girders were pre-tensioned by two types ofCFRP. The
second bridge is 541 ft. (165 m) long and consists of five simply supported span girders,
110 ft. (33 m) long. Four girders were prestressed by two different types of CFRP using
straight and draped tendons. The AASHTO girders were also reinforced with CFRP
stirrups. A portion of the deck slab is reinforced by CFRP reinforcement. Design
considerations, safety features and construction of each bridge are described briefly. The
paper summarizes also the results of monitoring the behavior of each bridge. The
experience gained from these two bridges provides valuable information for the
development of the design guidelines, currently under consideration by the ACI
Committee 440, Fiber Reinforced Polymer.
75
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
76 Rizkalla and Tadros
Sami Rizkalla: FACI, is Distinguished Professor of Civil Engineering and Construction,
Director of the Constructed Facilities Laboratory, North Carolina State University. He is
Chairman of ACI Committee 440, Fiber Reinforced Polymer, member of the ACI-ASCE
joint Committee 550, Precast Concrete Structures and a member of ACI Committee 423,
Prestressed Concrete.
INTRODUCTION
During the last two decades, fiber reinforced polymer (FRP) materials have been used in
a variety of configurations as an alternative reinforcement for new and strengthening civil
engineering structures and bridges. The attractiveness of the material lies mainly in their
high corrosion resistance, high strength and fatigue resistance. In some cases, the non-
magnetic characteristics became more important for some special structures. This paper
focuses on the Canadian experience in the use of FRP as prestressing reinforcement for
concrete highway bridges. The two bridges described in this paper were built in 1993
and 1997 and monitored to examine the long-term behavior of FRP as prestressing
reinforcement in severe environment and under normal traffic conditions. Brief
summaries of the research work conducted by testing scale models to simulate the
behavior of the girder and the deck slab are presented. Design considerations and details
of the safety features used for each bridge are provided. The paper describes construction
of the precast girder and the methodology used for prestressing the FRP tendons and
rods. Test results of recent field testing of the bridges under truck loads are also
presented.
SIGNIFICANCE OF RESEARCH
This bridge is the first prestressed concrete highway bridge built in Calgary, Canada and
was completed in 1993. The bridge is a two span skew bridge made continuous for live
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The bridge acts as a simply supported span for its own weight and the weight of deck
slab. Each girder is pre-tensioned with 26 or 22-l/2" (12.7 mm) diameter steel strands to
resist these loads. The two simple spans made continuous with post-tensioning
consisting of 12-1/2" (12. 7 mm) 7 wire strands. Four of the twenty-six bulb-tee precast
girders were pre-tensioned by two types of CFRP tendons. Four of these girders were
pre-tensioned using Carbon Fiber Composite Cables (CFCC), 5/8" (15.2 mm) in
diameter, produced by Tokyo Rope, Japan. The other two girders were pre-tensioned
using 5/16" (8 mm) Leadline bars, produced by Mitsubishi Kasei, Japan. Fig. 2 shows
the configuration of the CFRP pre-tensioned strands used in the bulb-tee girder. The
CFCC strand has an area of0.176 in 2 (113.6 mm2). The cable is produced using carbon
fibers of polyacrylonitrile (PAN) type and epoxy resin [1]. The guaranteed ultimate
tensile strength is 250 ksi (1750 MPa), while the tensile elastic modulus is 20,000 ksi
(137 GPa). The average measured tensile strength, strain and elastic modulus based on
six samples are 310 ksi (2150 MPa), 1.6 percent and 19,400 ksi (137 GPa), respectively.
The linear density is 0.15 lb/ft (227 g/m) compared to 0.65 lb/ft (967 g/m) for steel.
The Leadline bar has an area of 0.073 in 2 ( 47.3 mm2). The guaranteed ultimate tensile
strength, strain and tensile elastic modulus are 285 ksi (1970 MPa), 1.3 percent and
21,300 ksi (147 GPa), respectively [2]. The linear density is 0.05 lb/ft (77 g/m). The
concrete compressive strength was 6,500 psi (45 MPa) at 28 days and 4,200 psi (30 MPa)
at time of jacking the tendons.
The main girders of Beddington Trail Bridge are pre-tensioned for dead load and post-
tensioned for live loads. The aim is to replace the steel tendons with CFRP tendons to
examine their behavior under typical environmental and loading conditions. The options
available to the authors, at the time, were to substitute the pre-tensioned tendons and/or
post-tensioned tendons. The replacement of the pre-tensioned strand will depend only on
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
the bond characteristic of the tendons, while substituting the post-tensioning tendons
requires investigating the method and reliability of the anchorage systems. Since it was
the first bridge ever built, the authors decided to replace the steel pre-tensioned tendons
only with CFRP tendons. A multi-channel fiber optic sensing system was used to
monitor the behavior of the bridge by measuring the strains of the CFRP tendons. In case
of the unlikely event of possible distress of CFRP tendons, safety features were included
[3]. One safety feature in this project is provided by adding holes in the web to support
an anchorage for external post-tension tendons, if required, as shown in Fig. (3).
The behavior of the bridge was investigated by testing four concrete T-beams. The
beams had total length of 21 ft. (6.3 m) and depth of 13 inches (330 mm). These tested
Experimental Results
A complete report of the test results can be found in reference [4]. The two beams
prestressed by Leadline and CFCC, tested under static loading conditions, behaved
linearly up to cracking and after cracking with reduced stiffness up to failure, as shown in
Fig. (4). After unloading and reloading, the behavior of the beams was elastic and the
permanent deflection was negligible. The stiffness of the beams after reloading were
identical to the initial loading.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The beam prestressed by CFCC tested under cyclic loading, showed elastic response for
two million cycles. This is attributed to the elastic behavior of the CFCC. The camber of
the beam prestressed by CFCC after the completion of two million cycles was equal to
the camber before the application of the cyclic loading. This indicates no loss of stiffness
due to cyclic loading. The beam was loaded to failure after completion of the two million
cycles. The beams followed the same path as the one tested under static loads. The
measured load and deflection at failure were 0.95 and 0.9 of the failure load and
corresponding deflection of the beam tested under static conditions, respectively.
The two prestressing materials are produced and shipped from Japan. Prestressing was
adapted to practice by splicing the composite material between two pieces of steel strands
using steel couplers as shown in Fig. (5). Applying this method allowed to use the same
prestressing equipment and minimized the length, consequently the cost of the FRP
reinforcements. Prestressing of all the strands for the typical girder is shown in Fig. (6).
The completed bridge is shown in Fig. (7).
Structural Monitoring
The bridge is instrumented with twenty fiber optic sensors attached to the CFRP tendons
and rods. Twenty fiber Bragg grating (FBG) were installed by Electro Photonics
Corporation. The network of FBGs is connected to a junction box for site monitoring as
shown in Fig. (8). To check the performance of the CFRP reinforcement and the FBGs,
the bridge was tested in November 1999 using three-axle truck load as shown in Fig. (9).
Test results indicated that 18 of the sensors were operative and the CFRP reinforcements
TAYLOR BRIDGE
Taylor Bridge is located in the Parish of Headingley, Manitoba, Canada and was opened
to traffic in October 1997. The 541 ft. long bridge (165 m) consists of 40 prestressed
concrete AASTO type girders. Four girders of the Taylor Bridge were prestressed by two
different types of carbon fiber reinforced polymer (CFRP) material using straight and
draped tendons. The girders were also reinforced by CFRP stirrups protruding from the
AASHTO type girders to act in composite action with the bridge deck. A portion of the
deck slab is reinforced by CFRP reinforcement. Glass fiber reinforced polymer (GFRP)
was also used to reinforce the barrier wall. The barrier wall is connected to the deck slab
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
with double-headed stainless steel bars. To obtain continuous information on the
behavior of the bridge and the performance of FRP as reinforcement and prestressing
tendons, the bridge is monitored to provide data to evaluate the long-term behavior and
durability of the FRP materials used.
Two different types of CFRP reinforcements were used. Carbon fiber composite cables
(CFCC), produced by Tokyo Rope in Japan, were used to pretension two girders, while
the other two girders were pretensioned using Leadline bars, produced by Mitsubishi
Chemical Corporation in Japan. Double-headed stainless steel tension bars were used for
the connection between the barrier wall and the deck slab.
Research Work
Due to a lack of codes and standards on the use of FRP as reinforcement and prestressing
materials for concrete bridges, an extensive experimental program was conducted over
five years before construction of the bridge. The program included testing of a large-
scale model of a bridge girder totally reinforced and prestressed with carbon FRP [5] and
a full scale portion of the bridge deck slab reinforced with CFRP under simulated traffic
loads up to failure [6] as shown in Fig. (lla) and (llb) respectively. Straight and draped
CFRP reinforcements were also tested under axial tension. Performance of CFRP as
shear reinforcement [7] including effect of bend and orieritation of the crack on the
tensile strength was investigated. Transfer and development lengths of the CFRP
reinforcement were also evaluated and a theoretical model was introduced [8]. In
addition, a research project [9] was conducted to examine the barrier wall and the deck
slab for steel-free bridge decks. The results of all these investigations were used to
design the Taylor Bridge.
Design Consideration
Bridge girders, prestressed and reinforced by CFRP, were designed to behave similar to
other girders of the bridge reinforced and prestressed with steel strains under service
loading conditions. The prestressing force and the eccentricity of the reinforcement were
kept the same for all girders. The initial prestressing level was 60 percent of the
Reinforcement detail of a typical bridge girder, prestressed with straight and drapped
tendons, well as reinforced with CFRP stirrups, is shown in Fig. (l3a). The figure shows
also the instrumentation used for structural monitoring of the girder. The completed
precast concrete girder is shown in Fig. (l3b).
Safety Features
Due to the limited experience with these types of CFRP reinforcements, an additional
safety detail was introduced based on the concept of providing an alternative load path in
the design of the bridge in order to avoid progressive collapse in case of failure of one of
its components. The alternative load path mechanism is illustrated in Fig. (14). The
cross diaphragms were designed to support the dead load of the bridge in case of the
unlikely event of failure of the two girders prestressed by CFRP. In addition, non-
prestressed reinforcements were provided in the girders prestressed by CFRP to develop
catenary action in case of breakage of the stressed reinforcement.
Monitoring System
A total of 63 fiber-Bragg grating (FBG) sensors and two multi-Bragg sensors were glued
to the reinforcing CFRP bars of the structural members of Taylor Bridge. FBG sensors
located at midspan were designed to monitor the maximum strain in the reinforcement
due to applied loads, while FBG sensors located at the girder ends were designed to
evaluate the transfer length of prestressing tendons. Due to the relatively high initial
prestressing strain (~8800 microstrain) and the limited full range of the FBG sensors,
most of the sensors were installed after tensioning the prestressing tendons. Some of the
FBG sensors were installed before prestressing to measure the initial prestressing strains
of the CFRP and steel tendons. The bridge is also being monitored by 26 conventional
electrical strain gauges mounted on the reinforcement to verify the readings of the optic
sensors. The multiplexing unit and the data logging system were installed in a heated
enclosure, located in the cross diaphragm under the bridge deck slab, as shown in Fig.
(15). A camera was installed at the bridge site to provide video information synchronized
with the optic sensors' signals as shown in Fig. (16).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Most of the available research on bond behavior of CFRP reinforcement was obtained by
testing specimens reinforced with single CFRP bars or strands; therefore, it is important
to assess the behavior in case of multiple CFRP tendons used for the bridge girder. FBG
sensors were installed on the CFRP tendons at the end of the bridge girders to measure
the effective stress level in the tendons after release of the prestressing force. Fig. (17)
shows the effective-to-initial prestressing ratio along five meters at the end of the bridge
girders prestressed by CFRP tendons. It can bee seen that the transfer length L~> is 340
mm (22.3 db) for CFRP tendons, where db is the bar diameter. Based on an independent
research program conducted to investigate the bond characteristics of CFRP tendons [8],
transfer lengths in the bridge girders were recommended as 22.5 db for the same CFRP
tendons. The observed values for transfer lengths showed good agreement with the
values determined based on the equations proposed.
Load Testing
The output signal of a FBG sensor installed on a CFRP prestressing tendon at midspan
was recorded every 0.24 second during a test loading conducted using a slow moving
truck, as shown in Fig. (18). The 36-ton truck made several passes over the girder span,
forward and backward. The FBG sensor was able to record the strain change experienced
by the CFRP tendon. The figure clearly identifies both the truck and the trailer loading as
two distinct peaks in the response curve even though the magnitude of the strains are
quite small. The direction of travel can also be detected by the relative magnitudes of the
peaks since the truck load is larger than the trailer load. Hence, the first event in Fig. (18)
represents a backward pass and the subsequent one is a forward pass over the span.
Taylor Bridge is shown in Fig. (19).
ACKNOWLEDGEMENT --`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The authors wish to acknowledge the support of the Network of Centres of Excellence on
Intelligent Sensing for Innovative Structures (ISIS Canada) and Manitoba Highways and
Transportation Department of the Province of Manitoba. The writers gratefully
acknowledge support provided by E-TEK Electrophotonics Solutions, Toronto, Ontario,
Canada, for providing the materials used in the Taylor Bridge. Special thanks are
extended to Walter Saltzberg, Chris Wade, Amr Abdelrahman, Amir Fam and Emile
Shehata for their assistance during the design and construction of the bridge.
REFERENCES
l. Tokyo Rope Mfg Co., Ltd. 1993. Technical Data on CFCC, Japan, 100 pp.
2. Mitsubishi Kasei Corporation. 1992. Leadline Carbon Fiber Rods, Technical Data,
Japan, 40 pp.
3. Rizkalla, S. and G. Tadros. 1994. "First Smart Bridge in Canada", ACI Concrete
International, Vol. 16, No.6, pp. 42-44.
'·
'·
.rJ
"'
[jj
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
..c:
-5 \\ ,,,
,,,, u
["'
~
2 '·'
~
'· '·
Abutment 13 Bulb T -Section Girders
DinU!Il¥iomin "111111··
100
90
Jll'l---"""
80 ~~
70
~ ~
60 ~ o/
~ ~~ ,
/ .. w
50
~ l./1
~
(J_ _U v
40 Experimental
30 Predicted
20
Ill '/
10 lh
0
~II
0 10 20 30 40 50 60 70 80 90 100 110 120
Deflection (mm)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 8: Junction box for the lead wires of the fiber optic sensors.
.. .-------------------------------------,
0~----------------------------------~
232.0 2320 2330
"" 2360
Fig. lla: Large Scale model of a bridge girder for Taylor Bridge.
20 Basedl~FCC ~Leadline
Girder with
E~I ({)
guaran,~eed Girder
strength .............. with Steel
~ -g 15
-ro
"E ~
Q) 0
E ..c 10
0 ~
~
service load
5
oL-----~----+-----+-----+-----
-5 0 5 10 15 20
Curvature (1/1 000 m)
Fig. 12: Flexural behavior of girder prestressed with CFRP and steel.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 13: (a) Reinforcement details of typical AASHTO bridge girder of Taylor Bridge
and (b) Completed girder ready for shipping to the site.
cross diaphragms
Fig. 14: Failure load-path mechanism used in the design of Taylor Bridge.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 15: Monitoring devices and the computer for Taylor Bridge.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 16: Camera used for monitoring traffic.
[~-------------~¥~~-------------~
0.8 .~:
0.6 i .................. ..
fpe
1; L1 = 340mm
!pi 0.4 I-Y~····································· · · · · · · · · · · · · · · · · · · · · · · · · ·················
0.2
~ ................................................................................
,.-······]·-······
1 · fr><· = the effective prestressing stress
j,,1 = the initial prestressing stress
15
10
w
2;
c:
£1
CJ)
5
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0
Synopsis: The Naval Facilities Engineering Service Center (NFESC) investigated a hull
concept for a proposed floating pier with a design life of 100 years. The hull concept
consists of bidirectional post-tensioned lightweight concrete panels reinforced with
carbon fiber reinforced polymer (CFRP) grids for crack control and toughness. CFRP
grid reinforcement could be a potential durable alternative to using stainless steel
reinforcement and could be cost-effective for marine application, if the use ofCFRP
reinforcement justifies minimizing the concrete cover and allowing minor cracking
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
during service load. The paper discusses construction and cost considerations developed
during production of 13 testing specimens of such a panel concept. Though the panel
concept seems feasible, the NFESC deferred the concept for the time being due to the
high cost ofCFRP, the difficulties in detailing panel connections using CFRP
reinforcement, and the unknowns related to the durability of CFRP reinforcement in
cracked concrete exposed to marine environment.
91
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
92 Wernli et al.
Markus Wernli is a project engineer at BERGER/ABAM Engineers Inc., Federal Way,
Washington. He holds a PhD. from the University of California in San Diego in the field
of advanced composites. His interests are the use of advanced composite materials in
infrastructure.
INTRODUCTION
The Naval Facilities Engineering Service Center (NFESC) is developing a pier concept
that offers a design alternative to the traditional pile-supported structure. The design goal
is that the pier has a maintenance-free service life of 100 years and is flexible in function
and location. The functional flexibility requires the ability to berth a range of Navy
surface combatants that have different demands on utilities and access. The flexibility of
location will allow the Navy to move or to reconfigure a pier as a Navy-base mission
changes. The mobility of the pier also enables building the pier off site, which lowers the
costs by accessing a larger construction market and minimizes the disruption to Naval
station operations during pier replacement at the site. Hence, the pier concept needs to be
minimally dependent on site parameters, such as soil condition, tidal difference,
seismicity, and climate.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
After an initial concept study, NFESC decided to pursue a two-story floating pier made
of 100m long (325ft) concrete pontoons or modules. The modules can be joined to form
the desired length of the pier, typically 400 m (1,300 ft) (Fig. 1). The modules can be
fabricated in a drydock anywhere along the coast, towed to the site, and joined. The
service interruption of an existing pier can be minimized to the time it takes to demolish
the old pier and to install the mooring for the new floating pier. The pier is referred to as
Modular Hybrid Pier (MHP).
DESIGN INNOVATION
NFESC considered several durable materials for the design of the modules, such as
concrete with high fly-ash content, carbon fiber reinforced polymer (CFRP)
reinforcement, stainless steel, and epoxy-coated steel reinforcement (1). The modules are
compartmented boxes whose hulls are made of bidirectional post-tensioned lightweight
concrete panels. The panels are manufactured at a precast plant to achieve higher and
more consistent quality, tighter construction tolerances, and more uniform surfaces.
The panel post-tensioning functions as main reinforcement resisting the local stresses
from hydrostatic pressure, as well as the global membrane forces from bending of the
entire pontoon. Under service loads, the hull panels are designed to have a compression
zone from either bending or decompression stresses at any section cut. In this manner,
the panels remain watertight. Limited tensile stresses are allowed in the tensile zone.
This requires rigorous service design criteria on maximum tensile strain, which typically
dominate the design of the panel. Supplemental grid reinforcement is added close to the
surface to improve cracking behavior and provide toughness during local overloading.
Because of the relatively small hydrostatic forces, the thickness of the panel is mainly a
function of the space required to accommodate the post-tensioning and grid
reinforcement with sufficient cover. The use of CFRP reinforcement allows a smaller
concrete cover attributable to the durable behavior of composites in a marine
environment. This is desirable for a floating structure as it reduces the panel thickness
and thus the weight and the draft of the structure. A typical marine structure is designed
with a concrete cover of 50 mm (2 in.). Reducing the concrete cover to 13 mm (0.5 in.)
could reduce the panel thickness by 25 percent to as small as 200 mm (8 in.) and still
accommodate all layers of reinforcement.
A prototype panel was designed using a fine CFRP grid reinforcement with minimum
concrete cover (Fig. 2). The bar spacing of the grid was minimized to improve crack
control. If the CFRP grid reinforcement proves to be effective as crack control
reinforcement, then larger reinforcement tensile strains could be permissible at service
level to better utilize the high strength of the composite material and to increase the
efficiency of the design.
Such prototype panels were tested at the University of Wyoming (2). The structural
testing program showed that the composite grid reinforcement can increase the ultimate
strength of the panel by 30 percent compared to panels reinforced with conventional mild
steel mesh. Crack widths could be kept below 0.25 mm (0.01 in.) at a load level up to 60
percent of the panel's ultimate load, while the composite material reached strains of more
than 0.2 percent.
While the current provisions by ACI 357 for offshore concrete structures (3) allows a
maximum strain of 0.06 percent in mild steel during service load to limit cracking, a
maximum design strain of the CFRP reinforcement of 0.2 percent could be acceptable, if
the CFRP reinforcement has proven adequate durability in a marine environment.
Therefore, the panel prototype would allow reinforcement strains more than three times
higher than the ACI 357 allowable. This approach could increase the service load level
of the panel by a factor of two and, for this reason, increase the efficiency of the hull
panels.
Part of the testing program in Reference (2) investigated the constructability of the
prototype panels. The test panels were fabricated using self-consolidating lightweight
concrete to allow concrete to encase the fine grid reinforcement and to minimize the need
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The objective of the testing was to evaluate basic constructability issues related to the use
of the CFRP grid material and the self-consolidating lightweight concrete mix. Two
constructability specimens were built with full-size post-tensioning ducts and spliced top
and bottom CFRP grid reinforcement layers. The specimen fabrication was to be
performed with production equipment in a precast plant rather than in a laboratory, using
full-size production mixer and a production stressing bed. The testing investigated the
effect of grid size on concrete flow to achieve complete embedment of the grid material
without voids or concrete segregation. The impact of grid splices on concrete flow was
examined. Finally, methods of handling, placing, and securing the CFRP grid material to
reliably achieve the minimall3 mm (0.5 in.) concrete cover were explored.
The grid material selected for use in the specimens was a CFRP grid with the brand name
NEFMAC. The grid is a machine woven continuous CFRP consisting of high-
performance carbon fibers in a vinyl ester matrix. It is made in various mesh and bar
sizes and is delivered in flat sheets. The bond between reinforcement grid and concrete is
achieved by interlock of the concrete at each crossbar of the grid. Two grid sizes were
used for the testing: a fine grid denoted as C6-50/50 (C6-2/2) with bar area of 17.4 mm 2
(0.027 in2) and a mesh size of 50 by 50 mm (2 x 2 in.), and a coarse grid denoted as Cl3-
1001100 (C 13-4/4) with bar area of 64.5 mm 2 (0.1 0 in2) and a mesh size of 100 by 100
mm (4 x 4 in.). The fine grid has 46 percent less material than the coarse grid. As an
alternative reinforcement to the CFRP grids, the structural testing program considered
specimens with CFRP reinforcement bars and with a conventional welded wire fabric.
These alternative reinforcements are not further discussed in this paper.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The lightweight concrete mix was designed to be durable and adequate for the marine
environment. The concrete was required to be self-consolidating, hence to have a high
workability to flow around the fine grid reinforcement without segregating with
minimum or no external compaction. The mix was made with lightweight aggregate and
normal sand and had a high content of fly-ash and additives to increase flowability and
cohesion. The maximum aggregate size was 9.5 mm (3/8 in.). The concrete reached an
average strength of more than 50 MPa (7.5 ksi) after 56 days.
The 11 structural test specimens were manufactured on a prestress bed with the
prestressing tendons typically being continuous through a pair of specimens. The
specimens measured 6.1 m by 610 mm (20 by 2ft) and had a thickness of200 mm (8 in.).
They were typically reinforced with four prestressed steel strands, a bottom
reinforcement layer over the middle third of the length of the specimen, and some minor
top reinforcement. In all structural testing specimens, the bottom reinforcement layer
was held in proper position by supporting it on concrete spacers and wedging it against
the prestress steel by concrete spacer blocks to prevent the grids from floating up. The
top reinforcement layer was supported by steel chairs at a spacing of 900 mm (3 ft).
The coarse grid was stiff enough to allow easy placement. The fine grid was very
flexible and tended to warp during tying, which made maintaining a constant concrete
cover within tolerances difficult (Fig. 4). Once the grids were in position, it was very
difficult to make any changes on both grid sizes, because the small bar spacing did not
allow adjustment access within the form (Fig. 5). The concrete of the constructability
specimens was deposited at one end of the form and flowed slowly across the form. The
concrete for the structural specimens was placed by moving the concrete delivery chute
across the form (Fig. 6). The concrete flowed around the grid reinforcement and
typically embedded all grids, whether fine, coarse, or spliced, without voids. Some
concrete testing cylinders and some sections of panel specimens that were cut after the
conclusion of testing showed aggregate segregation, which seemed to have occurred due
to the sensitive concrete mix rather than due to the layout of the reinforcement. As the
concrete filled up the forms, it tended to push the top layer of the grid up. The coarse
CFRP could resist the upward pressure, whereas the fine CFRP grids bulged up and had
to be pushed down to achieve the required concrete cover (Fig. 7). The fine grid would
have to be stiffened or tied more frequently than at the spacing of 600 mm (2ft) to stay in
proper position. For the stiffer reinforcement, such as the coarse grid and the CFRP bar
reinforcement, supports every 900 mm (3 ft) were sufficient to maintain proper concrete
cover.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The CFRP grids come in large flat sheets and are easily integrated as long as a panel
surface remains flat. Detailing with such grids becomes difficult where reinforcement
meets at an angle, such as in haunched panels or at panel connections. Due to the
compactness of the grids, it is almost impossible to interlace the grid reinforcement.
Also, the CFRP reinforcement cannot be bent on site and all connection reinforcement or
hoops have to be premanufactured to tight tolerances in order to meet the small concrete
cover requirement.
The CFRP grids are fragile, in particular the fine grids, and have to be handled carefully.
They are easily cut to length and placed. Walking on the reinforcement cannot be
permitted. Once the girds are placed, access into the form is very difficult, in particular,
for the fine mesh size. During the fabrication of the specimens, the installation of the
CFRP grids to achieve a consistent amount of concrete cover, especially for the fine grid
with small cover, turned out to be difficult and would need some practice and some
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
development of accessories, tools, and special procedures for efficient and accurate
layout, and achievement of consistent cover. Examples of accessories needed are spacers
and couplers, which could clip directly to the grid. The accurate positioning of the
reinforcement grid within the depth of the panel section is a major concern, because the
composite grid tends to move and to float up during concrete placing. The reinforcement
supports and spacers have to be spaced closer together on more flexible grids to prevent
the grids from localized floating. The structural testing in Reference (2) also indicated
difficulties creating a structurally sound lap splice of the grids at small cover. A
minimum concrete cover of 25 mm (1 in.) seems more reasonable than the 13 mm (0.5
in.) used in the tests to allow for larger construction tolerances for the grid placement.
Self-consolidating lightweight concrete mix can achieve the required workability and
compaction around all the tested reinforcement arrangements. However, the concrete
mix during testing was sensitive to the balance between water-reducing and cohesive
admixtures and tended to segregate. The problem of segregation was a problem of the
concrete mix design rather than of the reinforcement arrangement. Recent experience has
shown that considerable production testing and mix optimization is necessary to achieve
successful performance from self-consolidating lightweight concrete.
COSTS
A cost comparison for alternative grid reinforcement for the prototype panel is
summarized in Table 1. The comparison is based on similar stiffness, which is more
significant for crack control than strength and, thus, more relevant for this project. A
comparison based on strength would make the higher-strength, carbon-based composites
more competitive as the cost rises with strength. All composite reinforcement
arrangements seem to be at least three times more expensive than a stainless steel grid
with equivalent stiffness. In mass production, the cost of the prototype panel made with
two layers of black steel grids would arise to about $430 per m2 ($40 per sq ft). An
upgrade to stainless steel grid reinforcement would increase the panel costs by 10
If larger crack widths of 0.25 mm (0.01 in.) are admissible with the use of CFRP grids,
then twice as much service load can be resisted with the same panel section. As detailing
allows, the panel thickness could be reduced, which would save material costs
approaching the magnitude of the additional cost for the composite reinforcement. The
CFRP grids could, therefore, be cost-effective. However, similar results can be achieved
with any higher-strength, corrosion-resistant reinforcement material, such as a Duplex
2205 stainless steel at lower cost.
The effects of the marine environment on the durability of CFRP reinforcement is not
well known. Future work should focus on confirming the behavior and long-term
durability of prestressed concrete panels reinforced with different types of high-strength,
corrosion-resistant reinforcement. Besides accelerated aging tests on the reinforcement
material alone, rapid and long-term exposure tests on cracked concrete panels in marine
environment should be undertaken to confirm the durability of the panel concept.
Further investigations should also include more structural testing of the performance of
CFRP grid reinforcement. The bond transfer of the NEFMAC CFRP grid reinforcement
relies on the shear strength of the composite material at the crossbars. This strength is
primarily provided by the vinyl ester matrix. If the vinyl ester is cracked due to high-
tensile strain or due to fatigue or aging, the shear capacity of the grid crosslinks is
expected to drop significantly. More testing should investigate the anchorage of grids in
different stress environments, such as biaxial stress and fatigue loading. The grid also
forms a weak shear layer within the concrete section. The concrete section becomes
prone to delamination, in particular, in grid-splice locations. This could also become a
potential impact resistance problem. More bending and impact tests on panels with
spliced reinforcement should be considered.
CONCLUSION
The production testing showed that a prestressed panel with CFRP grid reinforcement can
be manufactured with a minimum concrete cover of 13 mm (0.5 in.) and that such a panel
can be cost-effective for marine design, if the use of CFRP justifies that crack widths of
0.25 mm (0.01 in.) can be admissible during service load. However, NFESC did not
consider the panel concept for the near-term future of the Modular Hybrid Pier and
instead selected a prestressed panel design following the ACI 357 service strain
requirements using a combination of stainless and epoxy-coated steel reinforcement with
a concrete cover of30 mm (1.2 in.) and 44 mm (1.75 in.), respectively. The decision was
based on aspects of costs, design, constructability, and durability.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
At the present time, the CFRP reinforcement grids are at least three times more expensive
than stainless steel reinforcement. Allowing cracks of 0.25 mm (0.1 in.) and a small
concrete cover of 13 mm (0.5 in.) with the use of CFRP reinforcement may increase the
reinforcement strain to three times as high as the allowable strain by ACI 357. This
could enable the panel thickness to be reduced to compensate for the cost premium of
CFRP over stainless steel and could make CFRP cost-effective. However, the minimum
panel thickness is limited by the space needed to accommodate all reinforcement layers,
including two directions of post-tensioning.
In Reference (2), the CFRP grid was shown to be structurally adequate for the panels, but
could also be replaced by pultruded CFRP rods or by a high-strength stainless steel mesh
with similar results, assuming larger crack widths could also be admissible with the use
of stainless steel reinforcement. The use of CFRP reinforcement instead of stainless steel
reinforcement cannot be cost-effective unless the CFRP price drops to the range of the
price for stainless steel or it can be shown that even higher service strain levels are
admissible.
REFERENCES
3. ACI 357, "Guide for the Design and Construction of Fixed Offshore Concrete
Structures," ACI 357R-84, American Concrete Institute, 1984, Reapproved 1997
CFRPGrid
Reinforcement e
e
=
=
N
Fig. 2: Panel concept with bidirectional post-tensioning and CFRP grid reinforcement.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 7: Concrete pour of constructability specimen with fine CFRP grid: Top
reinforcement layer is pushed up.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: A new connector system and a new method for the construction of partially
precast concrete sandwich panels are described. The new connectors are constructed
using glass fiber reinforced polymer and are used to tie two precast concrete layers
together through a layer of rigid extruded polystyrene insulation. In contrast to existing
connector systems that incorporate steel lattice girders, the new system effectively
eliminates thermal bridges and undesirable forced compatibility strains in the concrete
layers. In addition to providing energy savings for the building owner, the new system
and method can provide cost savings for the wall fabricator.
103
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Dr. Rex C. Donahey studied civil engineering and received his PhD from the University
of Kansas. He has taught structural engineering at Okahoma State University, the
University of Illinois, as well as the Unversity of Kansas. He has practiced civil and
structural engineering with the Marley Cooling Tower Company, Mission, Kansas, and
with Ellerbe Becket, Inc., Kansas City, and is currently Director of Engineering and
Research with Composite Technologies Corporation in Boone, Iowa. He is a registered
PE in Oklahoma, Indiana, and Florida. Dr. Donahey has published papers on bond
capacity of reinforcing steel as well as on composite beams with web penetrations.
Recent research has included development of new connectors as well as new design
methods for reinforced concrete sandwich panels.
INTRODUCTION
Since the introduction of DIN 4108-2 [1], DIN 4108-3 [2], and the German Energy
Conservation Regulation [3], increased thermal protection requirements have been
assigned to external walls constructed in Germany. In particular, it is now required that
thermal bridges shall be avoided. The following paper describes a newly developed glass
fiber reinforced polymer (GFRP) connector system used to construct a new type of
partially precast concrete sandwich panels. These panels are, in effect, used as stay-in-
place forms to which site-cast concrete is added to create a monolithic structural layer
over the full height and perimeter of a building construction. Because the new GFRP
connector eliminates thermal bridges within the completed sandwich panel, this new
connector and panel system provide significant insulation improvements when compared
with the insulation characteristics of existing wall types.
GENERAL
Multi-layer reinforced concrete wall panels with integral thermal insulation (concrete
sandwich panels) are currently used primarily as space-enclosing construction units for
commercial and industrial buildings. They are characterized by high resistance to weather
effects, outstanding durability, and numerous available cast-in finishes. In German
practice, the cross section of such prefabricated walls typically comprises at least one,
120 mm minimum thickness concrete layer (structural layer) on the interior side of the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
As can be inferred from the investigations of Cziesielski and Marquardt [4], external
walls constructed using multi-layer precast concrete sandwich panels represent an
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
economical alternative to the common block and brick-work type constructions
commonly used for home construction in Europe. In addition, the introduction of the
Energy Conservation Regulation [3], which requires that increased thermal protection be
provided by such walls, presents a challenge to the use of traditional systems. Since the
effect of thermal bridges in the external wall surfaces must also be considered [3] (also
refer to DIN 4108-3 [2]), the use of the common existing construction types may be
restricted unless one can provide a more exact verification, e.g. in accordance with DIN
EN ISO 10211-1 [5] or -2 [6]).
In German construction practice, the connection of the interior and outer shell of multi-
layer reinforced concrete sandwich panels predominantly is made according to DIN 1045
[7] using stainless steel anchoring systems (material specified as NR. 1.4571 according to
DIN EN 10088-3 [8]). For example, the anchor characteristics for one system are
prescribed in the German General Construction Approval Z-30.3-6 [9]. For the
dimensioning of such anchors, type-specific computations are available which have been
calibrated on the basis of construction-unit tests.
In the following sections, a newly developed system for multilayer reinforced concrete
wall panels is described. This system provides for the connection of the concrete layers
of partially finished precast walls during transport, during the placement of on-site
concrete, and during the service life of the building. As with the TM connector system,
this anchorage differs from conventional constructions by its use of GFRP, rather than
stainless steel, as the anchor material. Thus, the resulting concrete sandwich panels can
be used as external walls which are implemented as part of the structural system of the
In contrast to the surface layers in metal sandwich panels, the two surface layers of multi-
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
layer reinforced concrete sandwich panels are normally constructed with very different
cross sections. In German practice, only the inner, more rigid concrete layer is used to
carry structural loads in the building. The outside concrete layer serves exclusively as an
aesthetic element and for protection from the weather. Essentially, this layer is stressed
by wind loads, daily and seasonal temperature changes, temperature gradient over its
thickness, and by its dead weight.
The connection of the facing layer to the inner layer must therefore allow changes in
relative position of the two layers to be as unhindered as possible. This will allow
deformations resulting from temperature changes to be of equal magnitude at opposite
edges of the panel and will therefore help to avoid harmful, forced-compatibility strains
in the concrete cross section as well as in the sealant material used in the external joints.
Frequently, this requirement is satisfied by placing a rigid shear connector at the centroid
of the facing layer to transfer the dead weight of the facing layer to the inner structural
layer of the sandwich panel (Fig. 2). In order to prevent rotation of the outer shell due to
an inadvertent eccentric installation of the rigid shear connector, a so-called torsion
anchor may also be installed. Torsion anchors are flexible under horizontal loading but
rigid in the vertical direction, and therefore provide a nearly unhindered length variation
of the face layer. Finally, tension-rigid retaining anchors are installed in a rectangular
grid over the entire facing layer area. Because of their section properties, these retaining
anchors flex in response to in-plane movements of the facing layer. Therefore, although
they provide high resistance to separation of the two concrete layers, they provide very
slight resistance to the application ofloads acting in the plane of the panel.
An additional alternative system used for carrying the weight of the sandwich panel
facing layer is provided by the installation of two, horizontally flexible shear connectors
located symmetrically about the facing-layer centroid. In addition to the required grid of
retaining anchors, a horizontally rigid anchor is also normally installed (Fig. 3).
In the current art of constructing partially precast walls with integral thermal insulation,
the connection between interior and external wall layers is made using stainless steel
lattice girders or using conventional reinforcing bars in combination with conventional
precast sandwich panel anchor elements (described above). Substantial disadvantages are
associated with this construction, including the thermal bridge effect created by the lattice
girders (considered in General Construction Approval [ 16]), and the rigid constraint of
movement of the face layer relative to the structural layer (which leads to substantial
forced-compatibility strains).
In contrast, the new system described here is similar to that obtained with the TM
connection system [10] in that it incorporates flexible GFRP connectors (Fig. 4). These
connectors are evenly distributed over the entire panel surface and are combined with a
rigid extruded polystyrene foam thermal insulation (Fig. 5).
Due to the small heat conductivity of GFRP (approximately 0.5 W/(mK)) the thermal
bridge effect of the new connecting devices is negligibly small when compared to the
bridge effect of stainless or conventional steel lattice girder systems (with conductivities
of 17 W/(mK) or 50 W/(mK), respectively). This is graphically illustrated by the
computed temperature distributions represented in Fig. 5.
Since the resonant frequency of the spring-mass system within the wall depends on the
thickness and rigidity of the insulating material as well as the weight of the concrete
layers, the sound transmission for the current system will be equivalent to that for
completely prefabricated concrete sandwich walls. In accordance with Supplement 1 to
DIN 4109 [17], the calculated sound transmission reduction of concrete sandwich panels
with rigid insulating material is 2 dB less than the reductions for single concrete layers.
However, because of the large insulation effect of the interior layer concrete layer, this
decrease is not problematic.
For the transfer of gravity or bracing loads, the site-cast concrete (added to the partially
prefabricated wall) acts compositively only with the interior, prefabricated concrete layer.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The new connectors are provided with conical end anchors that are embedded within the
precast concrete layers. Pullout tests show that the governing force is limited by
anchorage in the concrete, not the tensile capacity of the connector. The allowable
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
tension capacity of the connectors can therefore be determined using, after multiplication
by the factor w = 0.7 as introduced by Eligehausen, Mallee and Rehm [19] for
consideration of load-carrying capacity in cracked concrete, with the calculated gross
load-carrying capacity of the connecting device, !IRk.c. [20] given by:
Herein f3wN is the nominal strength of the concrete according to DIN 1045 [7] in N/mm2
and hef the effective embedment depth of the connecting device in mm. Using this
equation, a maximum possible anchor grid of 500 x 500 mm can be calculated for
connectors used under the following conditions: 1) connectors are embedded in
prefabricated wall layers with concrete in strength class B 45, 2) connectors are installed
with a minimum embedment of 45 mm, 3) the fresh, site-cast concrete has consistency
KR according to DIN 1045 [7], 4) the concreting speed is limited to 0.8 m!h for the site-
cast concrete placement, and 5) calculations include partial safety factors YQ = 1,15 for
temporary effects for completing units in the construction condition and YMc = 1.8 for
concrete failure.
In the finished building, the GFRP connectors are subject, as with the TM connector
system, to tension/pressure loads caused by wind and temperature gradient in the external
layer, shear stress due to the dead weight of the facing layer, and shear stress caused by
the elastic-shear connection of the inner and the outer concrete layers. The flexible
connection between interior and external concrete cross sections is also loaded by the
occurrence of normal forces on the facing layer and by temperature-induced as well as
differential shrinking of the wall layers (Fig. 6). The connecting devices cooperate with
the pressure-rigid thermal insulation as represented in Fig. 7.
Referring to Fig. 7, much of the initial load is carried through the thermal insulation,
since, under small deformation of the composite system under shear stress, adhesion of
the concrete at the insulating material surface is still present (Load Curve 1). Upon
reaching the bond capacity of the insulation, a large part of the shearing stress is taken
over by the flexible GFRP anchors (Load Curve 2). Therefore, a sudden increase in the
The newly developed GFRP anchor is manufactured [22] from unidirectional, boron-free
silica glass fibers within a polyurethane matrix. The new connector geometry, its
anchorage ends, and other features were developed specifically for use in the new
partially precast wall panel system described herein.
Its section and characteristic material properties are given by:
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Cross-sectional Area A = 111 mm2 ,
Moment of Inertia I = 1036 mm4,
2
Tensile Elastic Modulus EAz = 45000 N/mm ,
2
Bending Elastic Modulus EAb = 35000 N/mm ,
Tensile Strength in Fiber Direction fJz = 710 N/mm2 ,
Compression Strength in Fiber Direction fJD = 440 N/mm2 •
The connecting devices are combined with a rigid extruded polystyrene thermal
insulation with the same mechanical characteristics as is used with the TM connector
system [10].
The load-carrying capacity of the GFRP anchors under shear load is shown schematically
in Fig. 7. The compression forces working at the wall level can be determined only on
the basis of construction unit tests. For GFRP with unidirectional fibers, the tensile
strength parallel to the fibers depends substantially on the material composition and the
relationship of the tension components. However, it is also a function of the combined
effects of simultaneously occurring tension, compression and transverse compression
[23].
During the calculation for the connecting devices, the reduction of their load-carrying
capacity due to sustained loading and cyclic fatigue strength must be considered. The
results of tension tests by Franke, Overbeck and Meyer ([24] and [25]), as well as the
shear tests of Ramm and Gastmeyer [26] show the influence of these demands (which in
this case result from dead weight and temperature change in the facing layer). From
Computation of the acceptable load-carrying capacity under combined creep loading and
fatigue stress follows the accumulation of damage rule [28] developed by Franke and
Meyer.
The production of partially prefabricated walls with the new connection system can be
accomplished on a tilting table in the following work process:
installing the reinforcing and the concrete of the exterior (facing) layer,
placing the insulation sheets on the consolidated concrete layer,
putting the shear connectors into the fresh concrete through the prepunched
insulation,
curing of the concrete layer,
installing the reinforcing for the interior layer of concrete, including the lattice
girders,
placing the concrete of the interior finished layer unit,
rotating and inserting of the facing layer with shear connectors into the fresh
concrete of the interior layer (Fig. 8).
However, with the use of the new connection system, the partially prefabricated walls can
also be produced without the need to rotate the facing layer. In this case, the interior
layers are first manufactured, the thermal insulation is placed on the upper reinforcing
bars of the lattice girder using spacer supports and the shear connectors are put into the
fresh concrete.
In a further refinement, during the placement of the outer layer of concrete, the insulation
can be supported by the GFRP connectors alone. With either procedure, the partially
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The erection procedures for the panels are accomplished as with completely prefabricated
concrete sandwich panels. With the new connection system, the placement of the site-cast
concrete within the partially prefabricated walls is made in a manner similar to that used
in conventional, poured-in-place wall applications.
Fig. 9 shows the cross section of a precast wall constructed using the new system and
which was subsequently placed in a residential foundation in Germany. Note that the
lattice girders are limited to the structural layer of concrete and the air gap that will
receive the site-cast concrete. Other projects include a two-story combined residence and
business structure constructed in Germany. It is anticipated that the majority of
applications will be in the construction of integrally insulated, reinforced concrete
basement walls-first because of the functional seal provided and second because of the
potential for simple, rapid construction.
A new connection system has been developed for the construction of insulated, partially
prefabricated concrete sandwich panels. The new system comprises GFRP connectors
that initially tie the two precast concrete layers together during transport and during
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
placement of the on-site concrete. Because the panels accept the placement of a
continuous layer of site-cast concrete, they effectively serve as stay-in-place forms in the
building construction. The panels therefore create a complete structural and energy seal
for the building perimeter. In service, the anchors provide a flexible connection system
between interior and external concrete layers, but are loaded by normal forces applied to
the facing layer as well as during temperature-induced as well as differential shrinking of
the wall layers. The system has been used to construct a business building and residential
basements in Germany.
NOMENCLATURE
REFERENCES
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
X4~~~~~--sUSPENDEDFACADEELEMENT
INSULATION LAYER
~~+--- SITE-CAST CONCRETE
LAYER
RETAINING ANCHOR
Fig. 2 Concrete Sandwich Panel with a Rigid Shear Connector at the Centroid
of the Fascia
ETAINING ANCHOR
CENTER ANCHOR
(HORIZONTALLY RIGID)
Fig. 3 Concrete Sandwich Panel with Two, Horizontally Flexible Shear Connectors
Located Away from the Panel Centroid
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
PRECAST CONCRETE
DECK ELEMENTS
(FIUGREE SLABS)
I
I
I
f-------------,--
,-------------,--
I o o I
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
I I
I I
I I
I I
! __o o I
NEW CONNECTING ~ LATTICE I
ANCHOR : /GIRDERS~
INSULATION LAYER I I
I I
I o o I
I I
I I
I I
I I
I o o I
I
I
1------------,--
,----~--------;;-r--
I
L ___________ LI __
TEMPERATURE DISTRIBUTION
v--- ---r-
WITH NEW
CONNECTING-r- _ _ 1
~NCHOR
CONVENTIONAL
LATTICE GIRDER
_/L----------- L--
SYSTEM
Fig. 5 Partially Precast Wall with only GFRP Connectors Crossing the Integral Thermal
Insulation (Partially Filled with On-Site Concrete)
!lllllllll w
H U2
Fig. 6 Carrying Behavior of the Wall Construction from Fig. 5 under Transverse Loads,
under Differential Temperature, and under Differential Shrinkage of the Interior and
External Concrete Layers
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
1- _, ~------1
- _, I
I
I
I
t I 0 I
t=---=' =
1- -!
I
I
I
L ______ J
I
I
I
.a
w II
do .1 1. • I
Q
SHEAR LOAD
SHEAR BOND
CAPACITY
2
REL.AllVE DISPLACEMENT, U
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 8 Fabrication ofNew Panel System with GFRP Connectors-Face and Insulation Layer
with GFRP Connectors Being Lowered into Structural Layer Containing Lattice Girders
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
121
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
122 Huckelbridge and Eitel
Arthur A. Huckelbridge, is an Assoc. Professor of Civil Engineering at Case Western
Reserve University. He is a registered PE in Ohio, and a member of ASCE. His research
interests include bridge health monitoring, fatigue lifetime assessment and structural
dynamics.
INTRODUCTION
Ohio ranks second in the nation for its number of bridges, thirty -eight percent of which
were listed as structurally deficient or functionally obsolete by the Federal Highway
Administration in 2001. Most concrete bridge deterioration is attributed to corrosion of
steel reinforcing bars, where concern of structural safety and cost of rehabilitation i s
escalating. Advanced composites (also known as fiber reinforced polymers or FRP) have
gained increasing attention in structural applications due to their durability, corrosion-
resistance, and high-strength to weight ratio. Other key advantages include their non-
conductivity, electromagnetic neutrality, ease of production, and ease of installation.
FRP products consist of high strength glass (GFRP), carbon (CFRP), or aramid (AFRP)
fibers bound together with a polymer matrix made of epoxy, polyester, thermoplastic, or
vinyl ester resin. Although CFRP and AFRP offer a high modulus, high fatigue strength,
and reliable durability, GFRP is the most cost-effective alternative for new or
replacement structures. On a cost-benefit basis, GFRP reinforcing bars are emerging as a
promising alternative to conventional, epoxy coated, galvanized, or stainless steel rebar in
corrosive environments.
research projects on the subject of composite bars, "so far, none seems to have quite
succeeded" (Nanni et al. 1996). The report conveyed that Japan and Canada have
surpassed the United States and Europe in composite bar use and research, and concluded
that more documentation of actual performance was necessary.
In 2002, the Federal Highway Administration (FHWA) reported that they are focused on
advancing FRP composite technology in order to rebuild, rehabilitate, and maintain the
infrastructure through their Innovative Bridge Research and Construction (IBRC) (US
DOT 2002). The FHWA established the IBRC with the passage of TEA-21.
Congressional TEA-21 funding from 1998-2003 subsidizes preliminary engineering
costs; repair, rehabilitation or construction costs; and performance monitoring costs of
bridges using innovative materials in order to develop cost effective materials, reduce
maintenance and life-cycle costs, develop construction techniques, and develop design
criteria. This investigation is focused on meeting the research need for comprehensive
performance data, in order to promote innovative material usage by monitoring the stress
The bridge is located in the Southeastern Lake Erie Snowbelt, where it is subject to heavy
application of deicing salts and an average of 53 freeze-thaw cycles each year (US AF
2000), resulting in accelerated concrete cracking and chloride contamination. The 11.6 m
(38 ft) wide bridge consists of two 13.7 m (45 ft) spans carrying two lanes of traffic on
Miles Road over Deerlick creek in Bentleyville, Ohio. The original 1956 bridge
superstructure consisted of a 230 mm (9 in) thick steel reinforced concrete deck with a 76
mm (3 in) asphalt overlay on five steel girders. The 2000 Average Daily Traffic Flow
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(ADT) was 9479, and 6% of the ADT is truck traffic. The bridge is one of 214 bridges
under the jurisdiction of the Cuyahoga County Engineer's Office (CCEO). The County's
1999 bridge inspection rating was 4, on a scale of one to nine, one indicating non-
functional and 9 indicating excellent condition. Repairs necessary to keep the bridge
operational were performed in 2001, resulting in an upgraded rating of 5. The steel
girders were found to be in good condition, but the deck had numerous holes in it with
temporary steel plate patches.
Supportive funds were received for the rehabilitation project in the amount of the GFRP -
steel cost differential ($82,024 total, $28,710 for actual material costs) from the FHWA's
TEA-21 IBRC grant, administered through the Ohio Department of Transportation
(ODOT). The bridge is an excellent candidate for the examination of the durability of
composite technology because it is one of two adjacent bridges on Miles Road being
concurrently redecked, the other deck immediately east of it utilizing conventional steel
reinforcement. The total cost of the GFRP RC bridge deck is approximately $500,000
($1570/m 2 or $135/ft2).
The replacement 215 mm (8.5 in) thick concrete deck is reinforced with GFRP bars from
Hughes Brothers, Inc. of Seward, Nebraska. The county's standard 31 MPa (4500 psi)
concrete deck mix was used. Given that the bridge lacks the typical asphaltic wearing
surface, a High Molecular Weight Methacrylate (HMWM) rapid-curing penetrating
sealant was used to prevent cracking.
MATERIAL PROPERTIES
Hughes Brothers Asian TM 100 GFRP bars were selected for this project. The bars consist
of seventy-five percent continuous E-glass (Electrical glass) fibers bound by Hetron 922
vinyl ester resin, an economical GFRP composition. E -glass fibers are the most
commonly used and most affordable fiber costing approximately $2.14/kg ($0.80/lb).
Vinyl-ester resins offer resistance to chloride salts, most acids and alkalis. The bars are
manufactured by pultrusion, where rovings of glass fibers held under tension are
impregnated with thermosetting resin. A rib pattern and sand coating improve
GFRP composites do not exhibit yielding, or plastic behavior; the tensile stress -strain
curve for composite bars with one fiber type is typically linearly elastic until failure. The
potentially catastrophic brittle failure has been a source of apprehension for designers.
Concrete crushing is a marginally more desirable failure mode then failure of the bars.
To avoid sudden and catastrophic failures, the manufacturer recommends that working
stress be limited to twenty-five percent of the minimum ultimate tensile strength of
composite bars, twenty-percent in harsh environments. It should be noted that composite
reinforcement cannot be used where moment redistribution is desired because plastic
hinges will not form, nor should it be used as compression reinforcement due to its low
and unreliable compressive strength.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The ASLAN™ 100 #5 (16 mm) and #6 (19 mm) bars used in this project have guaranteed
(mean minus three standard deviations) tensile strengths of 655 MPa (95 ksi) and 620
MPa (90 ksi), respectively. The guaranteed tensile modulus, or average modulus, is 40.8
GPa (5929 x 103 ksi). Note that the low modulus coupled with conservative design
methodology may lead to a considerably high reinforcement ratio, when compared to
conventional steel reinforcement. For this project the deck reinforcement ratios utilized
were 1.82% for top reinforcement, 1.67% for bottom reinforcement, and 0.83% for
shrinkage and temperature reinforcement.
Bond between concrete and composite bars is transferred primarily by frictional slip
resistance and mechanical interlock due to sand coatings and deformation patterns, and
also by adhesion resistance (chemical bond) and pressure due to concrete shrinkage and
swelling of the rods. The embedment length used to determine the relevant surface area
in a bond stress calculation has been studied primarily with pullout tests, which are
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
tensile strength and stiffness and, when under stress, accelerated creep rupture. Due to
the high alkalinity (as high as pH 13.5) found in concrete pore water, many studies have
focused on GFRP alkaline degradation. Accelerated aging investigations of the Asian TM
100 bars have been conducted at Penn State University, Iowa State University, and the
University of Sherbrooke. A 13 pH Calcium Hydroxide solution at 80 °C (176 °F) for 28
days was used to simulate 50 years of service life, resulting in a degradation of
approximately 16% in tensile strength and 4% in modulus of elasticity (Hughes 2001).
However, the moisture resistant vinyl ester resin, used in the Asian TM bars, can
significantly protect the glass fibers from hydroxide ions, with the exception of minor
penetration at the bars' cut ends.
Laboratory four-point bend tests on 203 mm (8 in.) one-way RIC slab sections at CWRU
resulted in the observed performance comparisons between GFRP reinforced and
conventional steel reinforced RIC sections indicated in Figure 1. It is obvious from these
comparisons that GFRP reinforced concrete has quite substantially lower post-cracking
flexural stiffness than steel reinforced concrete, for similar reinforcement ratios, and that
GFRP reinforced concrete lacks any "yield plateau," such as is typically observed in
under-reinforced steel reinforced concrete. The low elastic modulus of the GFRP
reinforcement, coupled with its reasonably high tensile strains at failure (on the order of
1.5%), does result in potentially large curvatures (implying readily visible displacements
and cracks) at failure, presuming that shear and bond failures are avoided. Thus there is
the potential for some visible evidence of impending flexural failure, even though no real
ductility is available.
Deflection and crack control of FRP reinforced concrete under service loads requires
close attention, however, as does shear capacity and adequate development of FRP
reinforcement, as has been mentioned previously. Based on the four -point bend tests
mentioned above, actual cracking curvatures can be considerably larger then those
The unique composite material characteristics also bring about constructability issues. All
bends in the bars must be made at the factory before curing; no field bending is
permitted. Bends are made by shaping bars over molds or mandrels prior to
thermosetting the resin matrix. Bent bars typically maintain 38% of ultimate tensile
strength due to stress concentrations at fiber bends (Hughes 2001). Strength reduction in
bent bars varies according to the ratio of bar diameter to the radius of curvature, the
bending technique, resin type, and fiber type. Due to these constraints, a limited amount
of bends and shapes are available. J-hooks and other bends are achieved in the field by
lap splicing; the additional placement time required for lap splicing is expectantly offset
by the lightweight of the bars. The low specific gravity of the bars may cause them to
float once concrete is poured, necessitating secure formwork. A dust mask needs to be
worn when cutting the bars and gloves need to be worn while handling the bars to evade
stray fibers.
The variable and unfamiliar properties of GFRP reinforcing bars necessitate the
importance of a solid foundation of performance results before they can be fully accepted
as a reinforcing material.
INSTRUMENTATION
The layout of the instrumentation is illustrated in Figure 2. The goal of the monitoring
system is to collect strain, temperature, and deflection data from the bridge for a
minimum of 2 years. On a quarterly basis, slow environmental data sampling over a 48 -
hour period and fast data sampling during a live load test will be conducted.
A total of 56 uniaxial foil strain gauges were installed in-situ to ensure some redundancy
if any gauges become damaged. The west span was instrumented due to its better
accessibility from below the bridge. The strain gauges were placed at the location of
maximum positive moment, in the region of 7116 span, and at the location of maximum
negative moment, at the center support. Weldable gauges and clamp -on strain
transducers were attached to the upper and lower flanges of each of the five girders at
these locations, for a total of twenty gages. Twelve bondable gauges were mounted to
the top layer of transverse bars concentrated above the two northernmost girders,
corresponding to the westbound traffic lane, and twelve were mounted to the bottom
layer of transverse bars midway between the three northernmost girders. Twelve
bondable gauges were mounted on the top and bottom layer of longitudinal bars, midway
between the two northernmost girders and near the ends of the flange of the mid -lane
girder.
The constantan alloy strain gauges, from Measurements Group CEA series, have 350-
ohm resistances, copper coated solder tabs, and an encapsulation of polymide film.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
To achieve a uniform surface on the reinforcing bars, an initial coating of adhesive was
used to fill the bars prior to gauge installation. Measurements Group AE -10 Adhesive
(two-component, 100%-solids epoxy) was selected because of its superior bondability to
composites, moisture resistance, and long-term durability. The coating system used on
the reinforcing bars was recommended by the gauge manufacturer for long -term saltwater
exposure and mechanical protection. It consisted of Measurements Group's M-COAT B
(nitrile rubber) on the lead wires and tabs, M -COAT FT (Teflon film) directly covering
the gage, and two layers ofM-COAT J (polysulfide liquid polymer compound) separated
by M-COAT FA-2 (aluminum foil).
Temperature probes collect thermal data. Since concrete temperature may rise above
ambient temperature on a sunny day due to incoming solar radiation and similarly may
fall below ambient temperature on a clear night due to outgoing radiation to the sky, one
temperature probe monitors ambient conditions and four temperature probes are
embedded in the concrete. The temperature probes were positioned near the bar strain
gauges, at locations of maximum positive and maximum negative flexure, in the top and
bottom reinforcement layers.
At the time of the live load tests, two potentiometers are positioned at 7/16 span to
measure beam deflections and five spring-loaded direct current displacement transducers
(DCDT's) are used at the center support to measure relative slab deflections. The
DCDT's are mounted under the slab, midway between girders and on the girder
diaphragms, which serve as a reference point for the deflections.
All instruments are monitored simultaneously with a computer -based data logger.
For conversion of recorded GFRP rebar strains to stresses, the elastic modulus was
verified through tensile testing of five #5 Asian TM 100 GFRP bars in the Bingham
Structures Laboratory at Case Western Reserve University using an MTS Universal
Testing Machine. The average modulus per ASTM D3916-02, taken as the initial linear
portion of the curve, was 37.9 GPa (5.50 x 10 6 psi), as shown in Table 2.
As strain gauges with the exact temperature compensation for the GFRP rebars were not
readily available, the observed in-situ strains need to be corrected for thermal effects.
The Aslan GFRP bars have a reported longitudinal CTE of approximately 9.07
microstrain ;oc (5.04 microstrain /°F). For this investigation, the maximum expected
gauge apparent thermal strain output will occur at below freezing temperatures and can
Thermal calibration was verified in the laboratory on four unstrained sample bars with
the same adhesive and coating system used in the field. A temperature range of 0 ° F (-
17.8° C) to 95° F (35° C) was used in the verification. Apparent strains of approximately
-1 microstrain I" F (-1.8 microstrain I o C) were indeed observed for the compensated
gages with a CTE of 6.5 microstrain I oF (11.7 microstrain I 0 C), as one would expect,
from the published CTE data for the bars. To effectively interpret thermal strain effects,
unstrained sample reference bars of each combination of bar and gage type are used in
the field during all tests. Real mechanical strains can be determined, when the
compensating gages are used in conjunction with the information collected by the
temperature probes, located adjacent to both the active and unstrained reference bars.
Prior to the removal of the old conventional steel-reinforced deck, five DCDT's at the
center support and twenty girder mounted strain gages were used to acquire benchmark
data for locating the neutral axis and general strain distributions of the bridge with the
original conventionally reinforced deck. Three load cases were utilized with a 186.6 kN
(41.96 kip) test vehicle traveling at a slow (~ 2 m/s) speed. For the first load case, the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
vehicle traveled on and off the bridge in the eastbound Ian e with the wheel loads
positioned approximately midway between the girders. For the second and third load
case, the truck traveled eastbound and westbound respectively with the wheel loads
positioned, as nearly as possible, directly above the girders. A roughly 3 Hz dynamic
signal was observed to be superposed on the quasi -static live load response. It is
presumed that the low amplitude dynamic load was due either to the idling speed of the
truck engine as it crossed the bridge at low speed, of possibly due to interaction with the
truck payload (a large tank of water).
Peak slab deflections, measured relative to the girders, occurred during the second load
case and were approximately 0.23 mm (.009 in), as shown in Figure 3. The observed
deck deflected shape at the center piper, as the loaded axle passed over the pier, is shown
in Figure 4. Top flange strain gauges measured very small strains of opposite sense to
the bottom flange gauges, suggesting that the composite neutral axis was located just
below the top flange (the original deck was not designed to be composite). The peak
live load strain of the bottom flange was about 124 microstrain, corresponding to a stress
of approximately 25 MPa (3600 psi), occurring in the first load case with the truck
wheels roughly centered between girders. The strain time histories for load case 2 are
shown in Figure 5. Detailed analysis of the observed distribution data resulted in an
actual estimate of0.82 for the lateral wheel load distribution factor, which is considerably
lower than traditional AASHTO criteria would suggest (i.e., sl5.5 = 1.27).
After completion and curing of the new GFRP reinforced deck, a preliminary load test,
utilizing a construction vehicle (concrete mixer), was performed primarily as a "check-
out" of the instrumentation. The total weight of the vehicle was approximately 286 kN
Top flange girder strains were limited to about 8 microstrain; the small magnitude
confirms that the top flange continues to be in close proximity to the composite neutral
axis of the new deck. Longitudinal reinforcing bar live load strains at the pier were all
tensile and peaked at 18 microstrain, corresponding to a tensile stress of approximately
0.74 MPa (107 psi). Longitudinal bar live load strains at 7/16 span were all compressive
and peaked at -40 microstrain, corresponding to a compressive stress of approximately
1.64 MPa (237 psi), as shown in Figure 8. Preliminary analysis of reinforcing bar and
girder strains at 7I 16 span suggest that the composite neutral axis is located somewhere
between the girder/slab interface and the bottom layer of reinforcing bars.
Transverse gauges located nearest to the north girder (farthest away from the load)
conveyed negligible strains, indicating a distribution factor again less than unity. Peak
transverse reinforcing bar live load strains at the pier were approximately 25 microstrain
for the bottom layer of reinforcing bars, corresponding to a tensile stress of 1.02 MPa
(148 psi), and -35 microstrain for the top layer, corresponding to a compressive stress of
1.43 MPa (207 psi), as shown in Figure 9. At 7/16 span the peak bottom layer transverse
strains were approximately 45 microstrain, corresponding to a tensile stress of
approximately 1.84 MPa (266 psi), and the peak top layer strains were approximately -5
microstrain, corresponding to a compressive stress of approximately 0.20 MPa (30 psi),
as shown in Figure 10.
Transverse slab curvature at the center pier was estimated from the rebar stra in data to be
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
approximately 0.47j.lmm· 1 (12j.lin- 1). Curvature at cracking per ACI 318 for a 215 mm
(8.5 in) slab is 1.22j.lmm- 1 (3lj.lin- 1) (ACI 1999). This indicates that the wheel load
associated with the 286 kN (64 kip) vehicle produced a transverse curvature in the slab
less then 40% of the ACI prescribed cracking curvature, which can be generally
considered a conservative estimate of curvature at first cracking.
A live load test of the FRP reinforced deck, utilizing the same test vehicle from the
benchmark tests on the steel reinforced deck, was subsequently conducted. The payload
was 209 kN (47.1 kips), about 12% more than had been used in the benchmark test. The
deck, which was still not open to traffic, had a deep snow cover, burying a fair amount of
construction material, so lateral positioning of the test vehicle was quite limited. As a
result the wheels were located close to the center and south -center girders, and relative
slab flexural deflections produced during this test were minimal.
The peak bottom-flange girder strains observed at the 7/16 span were approximately 84
J.lE, about 27% less than had been observed in the original benchmark test on the steel-
reinforced deck, even though the test vehicle was about 12% heavier. The top flange
CONCLUSIONS
The successful completion of this project should supply performance data that should
help facilitate a better understanding of composite reinforcement in new and replacement
bridge decks, as well as in other corrosive or electromagnetic environments. Composite
reinforcement in lieu of steel could allow for enhanced performance of our nation's
concrete bridge decks, unhindered by the very negative effects of salt-induced corrosion.
ACKNOWLEDGEMENTS
Funding for this project was provided by the Federal Highway Administration's TEA -21
Innovative Bridge Research and Construction program, through the Cuyahoga County
Engineer's Office. We wish to express our thanks to Mr. David Mach, Bridge Design
Engineer, Cuyahoga County Engineer's Office and Mr. Robert Klaiber, Jr., Cuyahoga
County Engineer, and Mr. Mark Tramontana, J.D. Williamson Construction Company.
REFERENCES
American Concrete Institute (ACI) (1999). Building Code Requirements for Reinforced
Concrete, ACI 318-99, ACI, Farmington Hills, MI.
American Concrete Institute (ACI) Committee 440 (2001). Guide for the Design and
Construction of Concrete Reinforced with FRP Bars, ACI 440.1R-01, ACI,
Farmington Hills, MI.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CTE x11J t'F 5.5 5.04 6.5 -4.0 to 0.0 -3.3to 1.1
Sample A 5.55 x 10• psi 38.2 GPa 5.62 x 1o• psi 38.8 GPa
SampleB 5.59 x I 06 psi 38.5 GPa 5.78 x 106 psi 39.9 GPa
Sample C 5.32 x 106 psi 36.7 GPa 5.68 x 106 psi 39.2 GPa
6 6
SampleD 5.16 x 10 psi 35.6 GPa 5.52 x 10 psi 38.1 GPa
Sample£ 5.90 x 106 psi 40.7 GPa 6.12 x 106 psi 42.2 GPa
Avg. 5.50 x to• psi 37.9 GPa 5.75 x to• psi 39.6 GPa
1411
120
[ ~
---- r--
v
/
v
I Elcr s El 9 /
f_ ./
~ El" .19 El 9
/
I ./ 1
-grade 80 Steel Rebar
•o
20
~( I
Slab ~1%
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
203mm @102
I I I I I
0.00001 0.00002 0.00003 o.OOIXU o.oooos o.ooooe 0.00007 0.00008 0.00009 0.0001
Curvature (1/mm)
W30xl2~
Longihtdinal Bar Strain Gages o Girder Sh·ain Gages (I{/! pier and 7/16 >tJan)
c=Trrut,.Yet·se Bru· Sh·aiu Gages Displacement Transducers(~ 1{ll pier)
0 Temperahtre Sensors Cjl Potentiometers (2 <?'!' 7il6 sprut)
0.001
-0.001
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
t::l
i-0.001 . -0.051 !;,
..=·.,.=
~· -0.003 t '-0.076
~ -0.004 r-----------------~1~1~ J
-0.005 -O.l:Z7
-0.006 -0.151
-0.007 -0.17&
-0.00&
0 1 4 6 8 10 11 16
Tbne (sec)
N NC
s
Figure 4: Slab Deflected Shape @ Center Pier- Old Deck
160
i ~--------------~~
.~L.------------------r--------r----------------------------------------~
1 3 4 5 6 7 9 10 1'1 12 13 14 15 16 17 18 19 20 21 22
Time(sec)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0
I:'
e.
Ig f"
il
il"
Q
I
-5
0 II 11 lti 10
Tbnt (ser)
Figure 6: Relative Slab Deflections @ Center Pier- New Deck w/ 286 kN kip
Test Vehicle
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
N s I
Figure 7: Slab Deflected Shape@ Center Pier- New Deck
60
50
NuthGWdt:r
OB3
Qtl
-
N~Cat
Ll (B)
Tl
U(B) ¢1> !6
~abr Gir4lr
-Bper.Mo•.Avg.
-Bper.Mo..A•g.
-Bper.Mo•.Avg.
(TS)
(Tl)
(T4)
on - 8 per. Mo•. Avg. (T6)
40
OB2
" -8per.M... Avg. (B4)
Bper.Mov. Avg. (B5)
JO
OBI
" -a•
B gages 1t1. bQtt¢m layer
~ zo
e... 10
:2 0
-10
-ZO
-JO
-40
0 z 4 6 8 JO JZ u 16 Jll zo 21 Z4
Thnt (stt)
Figure 8: Transverse Bar Strain Histories at Center Pier w/ 286 kN Test Vehicle
JO Ll2(B)
UI(B) c:::>Til
L9 <::;:.'lBll
zo
=·' B
<:::tBlO
s in bottom laver
10
i."' 0
:2"
-10
-20
-8p•r.M<W.Avg. (L7 U>p~7/16}
-8per. Mov. Avg. (L9 U>p layer 7116}
..JO +---------------'"-\------,J\f'!-'1--- -8per. Mov. Avg. (LJO lwtU>m layer 7116}
- 8 per. Mov. Avg. (LlllwtU>m ~ 7116}
- 8 per. M... Avg. (Ll2 botU>m layer 7116}
0 z 6 8 10 11 u 16 11 zo zz Z4 16
Tlmt (sec)
Figure 9: Longitudinal Bar Strain Time History at 7/16 Span w/ 286 kN Test Vehicle
10
70
Lll (B)
Lll (B) Til
60 L9 ~Btl
50 =B7 . . 810
B gages in bottom layer
-·10per. Mov. Avg. (B7)
Ie 40
-lOper.Mov.Avg. (B7)
-10pu.Mov.Avg. (BJO)
~ 30 -lOper.Mov. Avg.
-10 per. Mov. Avg.
(Bll) ----hr---~--\og.,------------------1
(BI2)
-10 per. Mo~. Arg. (/7)
zo 10per. Mov.Arg. (Tl2) ---fl---------"'"-\'ofM-..,-,,.,...=:------------1
10
-10
0 6 8 10 u 16 20 zz 16
Time (sec)
Figure 10: Transverse Bar Strain Time History at 7116 Span w/ 286 kN Test Vehicle
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
80
70
80
301
10
5 10 15 20 25 40 45
Time (sec)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 11: C and SC Girder Strains- FRP Reinforced Deck w/ 209kN Test Vehicle
130
. - - - - - - - - B r i d g e 178; West Span Bottom Flange Strains@ 7/16 spa~n~-----~
120 Comparisons Between Old Deck & New Deck _North
5121102
- NorthCenter 5121102
110
-Center 5121102
100 - SouthCenter 5121102
41, 970 lb eastbound vehicle, .steering axle In lead -South 5121102
- Center 11282103
- South 1128/03
80
1128103 New Deck - SouthCenter 1/28/()3
41, 060 lb eastbound vehicle, tand6m axle In lead -North Center 1128103
70
-North 1128103
60
20
10
-10
-20
··!
~L--------~--------------~-------------~
0 10 IS 25 30 35
Time (seconds)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
SP-215-8
Design Considerations of
GFRP-Reinforced CRCP
139
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
-
Morgantown, WV. He received his Ph.D. from Northwestern University, Evanston, IL,
in 1989. His research interests are structural dynamics, nondestructive evaluation (NDE),
wave propagation, and dynamic soil-structure interaction, and he has led the
advancement of applying NDE techniques to concrete, composites and ceramic materials.
He has published more than 90 technical articles with 36 referred papers. He is a
member of the ASCE Committees of Shock and Vibratory Effects of SEI, Dynamics
Committee and the Experimental Analysis and Instrumentation Committee of EMD. He
is also active in the ACI Committees of Lunar Concrete, NDE of Concrete, and FRP
Reinforced Concrete.
INTRODUCTION
Other than the advantage of eliminating steel corrosion, thermal and stiffness
compatibility between Glass Fiber Reinforced Polymer (GFRP) and concrete also offer
advantages when GFRP rebars are used in Continuously Reinforced Concrete Pavements
(CRCP), (1). Traditional CRCP, reinforced by steel rebars, has been applied for a few
decades, (2), and its performance has been reported by many researchers, (3-5). Up to
this point, there has been no precedent for the use of GFRP as reinforcement in CRCP,
and little related research has been conducted. It is therefore necessary to study the
mechanical behavior and to develop the design mechanism for the GFRP-reinforced
CRCP.
RESEARCH SIGNIFICANCE
This study shows the effects of substituting the conventional steel reinforcement
with GFRP reinforcement in the continuously reinforced concrete pavements, on the
critical design factors to determine the performance of the pavements. The results
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
provide the fundamental understanding of the design considerations of GFRP-reinforced
CRCP. The study also reveals the areas where further studies are recommended.
ANALYTICAL MODEL
R=~~ (1)
(2a)
and
I- cosh~f-x) (2b)
L
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
cosh/3-
2
where ( <Tc.s)avg and ( CTc.t)avg = average axial concrete stresses induced by concrete
shrinkage and temperature variation, respectively; p = reinforcement ratio (
reinforcement cross-sectional area divided by concrete cross-sectional area); E, =
Young's modulus of the reinforcement; !1T= temperature variation; ac and ar = CTEs of
the concrete and reinforcement, respectively;
(3)
where &c.s(t) = shrinkage strain at any time t (in days) for moist cured concrete, (8); and
( &c,s)uit = ultimate shrinkage strain for drying at 40% RH; and
2Gc IE,
/3= (4)
r} ln(R I r,)
where Gc = shear modulus of concrete. The negative values from Eq. (2) indicate
compressive axial stresses in the concrete, and the maximum axial stress in concrete can
be simply found at the midpoint, x = L/2. The analysis results were compared well with
experimental measurements of free-free slabs, (9).
In the shrinkage stress analysis, the Young's modulus of concrete, Ec, as well as
concrete shrinkage strain, &c,s(t), are employed as the time-dependent properties, and
therefore, the elapsed time, t (days) is the only variable. The time-dependent Young's
modulus of concrete can be evaluated by
fort> 28 (5b)
where Ec, 28 = Young's modulus of concrete at 28 days, (10). In the thermal stress
analysis, the concrete stresses at 28 days are estimated for different temperature
variations; Ec, 28 was therefore employed for the Young's modulus of concrete in the
analysis.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
As can be seen in Figures 3 and 4, the tensile stress levels in concrete caused by
both concrete shrinkage and temperature variation are far below the tensile strength of
normal concrete for a free slab. Other restraining forces acting on a given concrete slab,
such as friction from the subbase under concrete pavements or restraints from
reinforcement ties to neighboring slabs, need to be considered. When these restraints are
considered, the overall resulting tensile stress level in the concrete will increase, possibly
exceeding the concrete tensile strength and causing cracks in the concrete slab.
The CRCPs considered in the FE analyses have been assumed to have transverse
cracks equally spaced at 1.524 m (5 ft. or 60 in.) and were subjected to the concrete
shrinkage and temperature variation. Since the cracks in the CRCP are assumed to be
spaced equally, the longitudinal reinforcement at the crack is constrained in both
longitudinal and rotational directions. Therefore, only a representative slab-segment was
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
needed to model the CRCP. The length of the slab-segment, L, is equal to the crack
spacing. While longitudinal reinforcement of #6 rebars with a radius (r) of 0.95 em
(0.375 in.) was considered in the FEM analysis, the influence of transverse reinforcement
was ignored. The reinforcement at its ends is constrained in the longitudinal and
rotational directions. The bond between the concrete and reinforcement is assumed to be
perfect, and the bond-slip between the concrete slab and subbase were modeled by using
horizontal spring elements with a spring stiffness of236.409 kN/m (1,350 lblin.), which is
simulated for a flexible subbase or a lime-treated clay subbase, having the bond-slip
stiffness per unit area of 40.719 MPa/m (150 psi/in.) with respect to the concrete slab.
The spring stiffness is corresponding to a concrete element with a horizontal length of
3.81 em (1.5 in.) and a thickness of 15.24 em (6 in.). The underlying layers are modeled
as an elastic foundation with an elastic spring constant of 108.583 MPa/m (400 psi/in.).
The FE model of CRCP slab-segment has been assumed to be subjected to a concrete
Shrinkage at 28 days Of 355.6 fl-l: and the temperature Change c~n Of -27.8 °C (-50 °F)
and -16.7 °C (-30 °F) at the top and bottom of the slab-segment, respectively. The
temperature varies linearly from the top to the bottom. All the other model parameters
and material properties used are listed in Table 1, and the schematic details of the FE
model are depicted in Figure 5.
Two different types of concrete, with granite aggregate or siliceous river gravel
aggregate, have been adopted in the stress analysis of the CRCP slab-segment. The CTEs
It can be seen in comparison between Figures 6 and 7 that the tensile stress level
in concrete for the granite concrete with a CTE of 10.26 f.lc;oc (5.7 f.icrF) is lower than
that for the siliceous river gravel concrete with a CTE of 14.40 f.ic;oc (8.0 f.icrF). This is
because the difference between CTEs of concrete and reinforcement for the granite
concrete is smaller than that for the siliceous river gravel concrete. Therefore, the tensile
stress in concrete can be reduced by using concrete with lower CTE. Regardless of the
concrete CTEs, the maximum tensile stresses in concrete for both steel and GFRP
reinforced slab-segments are much higher than the concrete tensile strength, all the slab-
segments will presumably be cracked at its mid-span due to the given temperature
variation and shrinkage conditions. The concrete tensile strengths for the granite and
siliceous river gravel concretes are about 2.90 MPa (420 psi) and 2.55 MPa (370 psi),
respectively, with a concrete compressive strength,.fc', of about 27.58 MPa (4,000 psi).
The axial stress distributions in the reinforcements along the longitudinal (x-)
direction for two different concrete CTEs are shown in Figures 8 and 9. While the
reinforcements around the middle are under low compression, those around the crack (the
ends of the slab-segment) are under high tension. Because of the stiffness compatibility
between GFRP and concrete, the tensile stresses in the GFRP reinforcements are lower
than those in the steel reinforcements. The tensile reinforcement stress for the lower
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
concrete CTE (Figure 8) is also lower than that for the higher concrete CTE (Figure 9).
Figures 8 and 9 show that the reinforcements are subjected to higher tensile stresses than
the limits of their allowable tensile stresses. While the stress limit for the grade 60 #6
steel rebar is about 75% of its ultimate tensile strength, 379 MPa (55 ksi), (2), the GFRP
rebar's allowable tensile stress is about 20 % of its ultimate tensile strength under harsh
environment, 124 MPa (18 ksi), (13). The GFRP rebar's allowable stress range is
established for the prevention of any undesired early failure of GFRP rebars under
The FE analyses were also conducted for CRCP slab-segments with different
bond-slip stiffness between the concrete and the reinforcement. The bond-slip stiffness
between concrete and reinforcement varies from 1.344 x 109 N/m (7.675 x 106 lb/in.),
0.433 x 109 N/m (2.474 x 106 lb/in.), and 0.175 x 109 N/m (l.O x 106 lb/in.), where the
first value corresponds to the perfect bond case, the second one corresponds to the bond-
slip stiffness between the concrete and steel reinforcement, and the third one is arbitrarily
selected to about one half of the second value. These bond-slip stiffnesses correspond to
the bond-slip stiffness per unit area of 589.5 x I 03 MPa/m (2.172 x 106 psi/in.), 190.0 x
103 MPa/m (0.7 x 106 psi/in.), and 76.86 x 103 MPa/m (0.283 x 106 psi/in.), respectively.
Figure 10 shows that the concrete stress level from a larger bond-slip stiffness is higher
than that from a smaller stiffness, since the tighter bond between the concrete and
reinforcement provides more restraint to the concrete's volume change. It can therefore
be anticipated that using reinforcement that provides the concrete with a tighter bond can
cause more cracks in a certain length of CRCP, leading to smaller crack spacing and a
narrower crack width. Under the given conditions, only the maximum concrete stress for
the bond-slip stiffness of0.175 x 109 N/m (l.O x 106 lb/in.) is below the tensile strength
of the concrete. Figure 11 shows the axial stress distributions in the GFRP reinforcement
for the different bond-slip stiffnesses. The stress level around the crack decreases as the
bond-slip stiffness decreases. The tensile stress levels for all the bond-slip cases are
higher than the allowable stress limit for the GFRP rebar. Employing reinforcement with
a lower bond-slip stiffness might be one way to reduce the stress levels in both concrete
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
and reinforcement in the CRCP. It is noted that the bond between the concrete and GFRP
reinforcement can not be perfect, and for practical purpose it can be assumed to be
between the second and third bond cases. The stress development in the CRCP ts
sensitive to the bond between the concrete and the reinforcement.
The variation of crack width throughout the pavement depth for the crack
spacing, L, of 1.524 m (5 ft.) is shown in Figure 12. The crack width for the higher bond-
slip stiffness is narrower than that for the lower ones as a result of having more restraint
against the concrete volume change. Also, the crack width for steel reinforcement is
narrower than that for GFRP reinforcement at a bond-slip stiffness due to the higher
Young's modulus of the steel reinforcement. In the figure, the protrusions on the graph
are found at the mid-depth of the pavement, where the reinforcement is located, and the
graph is observed to be more protruded with the higher bond-slip stiffness. This is due to
the fact that the concrete volume change is restrained most at the mid-depth by the
reinforcement and is subjected to more restraint as the bond-slip stiffness increases. The
crack width at the top of the pavement is wider than that at the bottom of the pavement,
since the magnitude of the temperature drop (111) is larger at the top than at the bottom,
and the restraint from the subbase exists at the bottom. The crack widths throughout the
pavement depth for all the cases are below 1.02 mm (0.04 in.), which is the maximum of
allowable crack width to prevent the loss of concrete aggregate interlock.
The above results show that the bond-slip between concrete and reinforcement
can have a significant effect on the stress development in the CRCP, and on the
determination of crack spacing and crack width. Therefore, using reinforcement that
provides a proper bond to the concrete could favorably control the cracking. In addition,
the behavior of the CRCP can be adjusted by varying concrete material properties. For
example, if it is necessary to lower the concrete stress in the CRCP, the concrete with
smaller CTE and Young's modulus may need to be employed. For a sound design of the
CRCP reinforced with GFRP rebars, the bond-slip relationship between the concrete and
GFRP rebar at different temperature environments should be studied.
The previous FE results show that in most cases, not only do the concrete
stresses exceed the concrete tensile strength, but the reinforcement stresses around the
crack also exceed the limit of allowable working stress of GFRP rebar under the given
conditions. As previously shown, the lower bond-slip stiffness between concrete and
reinforcement, as well as the smaller CTE and Young's modulus of concrete, can lower
the stress levels in the concrete and reinforcement. Therefore, the bond-slip stiffness of
0.433 x 109 Nlm (2.474 x 106 lblin.), instead of perfect bond, is employed for a design
example, along with a concrete CTE of6.84 pt:fOC (3.8 pd0 F) and Young's modulus of
24.8 GPa (3.6 x 106 psi) for a limestone aggregate concrete. The same bond-slip
stiffness between concrete and subbase as previously used is applied for the example. In
addition, the CRCP segments at three different lengths of 1.067 m (3.5 ft.), 1.524 m (5
(i.), and 2.438 m (8ft.) are modeled to represent the ranges of the crack spacing and crack
width expected for a typical CRCP design. The other conditions for the example are the
same as previously adopted.
The axial tensile concrete stress distribution at the top of the slab-segment, along
the longitudinal direction, has been plotted for each slab length (Figure 13). As shown in
the figure, the concrete stress level of the longer slab-segment is higher than that of the
shorter slab-segment. The maximum stress levels in the concrete for the lengths of 1.067
m (3.5 ft.), 1.524 m (5 ft.) and 2.438 m (8ft.) are about 1.86 MPa (270 psi), 3.08 MPa
(447 psi) and 6.16 MPa (893 psi), respectively. The stress level for L = 2.438 m (8ft.) is
well above the tensile strength, about 2.41 MPa (350 psi), of the limestone concrete.
This slab-segment will crack at the middle. On the other hand, the slab-segments with
lengths shorter than 1.067 m (3.5 ft.) will not crack at the middle; the maximum stress
level for L = 1.067 m (3.5 ft.) does not reach the tensile strength of concrete. The stress
level for L = 1.524 m (5 ft.) is assumed to be roughly on the verge of the cracking point or
has slightly passed the point. Therefore, the crack spacing for the example is anticipated
to be approximately shorter than 1.524 m (5 ft.).
Figure 14 shows the axial stress distributions in the GFRP reinforcement for the
different slab-segment lengths. Around the crack, the stress levels in the GFRP reach
around 113.83 MPa (16.51 ksi), 192.70 MPa (27.95 ksi), and 387.59 MPa (56.21 ksi) for
the lengths of 1.067 m (3.5 ft.), 1.524 m (5 .ft.), and 2.438 m (8ft.), respectively. Except
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The friction between the subbase and the concrete pavement is also a factor
influencing the stress levels in concrete and reinforcement. By increasing the bond-slip
stiffness between the concrete and subbase from 236.409 kN/m (1,350 lblin.) to 24.271
GN!m (138,600 lblin.), which are the stiffnesses for the flexible or lime-treated clay
subbase and the cement-stabilized subbase, respectively, the GFRP stress level around
the crack decreases from 192.70 MPa (27.95 ksi) to 175.55 MPa (25.46 ksi) for L = 1.524
m (5 ft.). The subbase with increased friction applies an increased reaction force to the
concrete as concrete volume change occurs, which also releases the force provided by the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
SUMMARY
The analytical results for a freely-supported reinforced concrete slab are first
compared with the Finite Element (FE) results. This shows a good agreement and
verifies the applicability of the FE model. Due to the low Young's modulus of the GFRP
rebar, the tensile stresses for the GFRP-reinforced concrete are shown to always be
smaller than those for the steel-reinforced concrete, leading to less potential for cracking
in the GFRP-reinforced concrete. Therefore, the use of GFRP rebar as the reinforcement
might lead to larger crack spacings followed by wider crack widths in the CRCP. In
addition, using reinforcement that provides the concrete with a looser bond can also cause
lower tensile stresses in the concrete, leading to larger crack spacing and wider crack
width.
ACKNOWLEDGEMENTS
REFERENCES
I. Chen, H. L., and Choi, J. H., "Effects of GFRP Reinforcing Rebars on Shrinkage and
Thermal Stresses in Concrete," Proceedings of the 151h ASCE Engineering Mechanics
Conference, Columbia University, New York, NY, June 2002.
3. Won, M.; Hankins, K.; and McCullough, B. F., "Mechanistic Analysis of Continuously
Reinforced Concrete Pavements Considering Material Characteristics, Variability, and
Fatigue," Research Report 1169-2, Center for Transportation Research, The University of
Texas at Austin, 1991.
4. Suh, Y. C.; Hankins, K.; and McCullough, B. F., "Early-Age Behavior of Continuously
Reinforced Concrete Pavement and Calibration of the Failure Prediction Model in the
CRCP-7 Program," Research Report 1244-3, Center for Transportation Research, The
University of Texas at Austin, 1992.
5. Kim, S. M.; Won, M. C.; and McCullough, B. F., "CRCP-9 Computer Program User's
Guide," Research Report 1831-3, Center for Transportation Research, The University of
Texas at Austin, 2001.
6. Zhang, J.; Li, V. C.; and Wu, C., "Influence of Reinforcing Bars on Shrinkage Stresses
in Concrete Slabs," ASCE Journal of Engineering Mechanics, V. 126, No. 12, 2000, pp.
1297-1300.
7. Cox, M. A., "The Elasticity and Strength of Paper and Other Fibrous Materials,"
British Journal ofApplied Physics, V. 3, 1952, pp. 72-79.
9. Chen, H. L., and Choi, J. H., "Effects of GFRP Reinforcing Rebars on Shrinkage and
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
10. Mosley, W. H., and Bungey, J. H., Reinforced Concrete Design, Macmillan Education
Ltd., 1990.
13. ACI 440.1R-01, Guide for the Design and Construction of Concrete with Reinforced
FRP Bars, American Concrete Institute, 2001.
APPENDIX NOTATION
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(note: * granite coarse aggregate and ** siliceous river gravel coarse aggregate used)
Representative Prism
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
L
(a)
X<
···t)r
.......................................••••••••.••••.••.••••.•..••.•••.••••••.•.••••.•.••••••••.(;:/~
L
(b)
~
'~ 7.6 Cf"l
365.8 C"1
SCCTIDN A A
Figure 2-25.4 em (10 in.) thick CRCP with #6 long. rebars (r = 0.95 em) spaced at
15.24 em and #5 trans. bars (r = 0.79 em) spaced at 1.22 m
- 0.2
,f
!.
1
.!
0.15
! 0.1
t
i 0.05
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0.01
0.008
0.006
0.002
0
-30 -25 -20 -15 -10 -5 0
AT(°C)
Roller
=
Ec, 0 (28) 355.611£
llT at top of slab= -27.8 °C
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
12.-------------------------------------------------------,
10
=
Ec,0 (28) 355.6 ~
llT at top of slab= -27.8 °F
=
ll T at bottom of slab -16.7 °F
(Temperature changes linearly from top to bottom.)
ac = 14.40 IJEf C
a,. = 11.88~J£fC
2 a, 9 =9.07 JJ£fC
I 300
! :zoo
100
0
20 40 60 80 100 120
-100
x (em)
/-+-steal -e-GFRP /
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
900
I 400
ii 300
~
:zoo
100
0
20 40 80 100 120
-100
x (em)
/-+-steal -e-GFRP /
-l
!!.4
J
- 3 e c.• (28) =355.6 /JE
i 2
liT at top of slab= -27.8 °C
liT at bottom of slab = -16.7 °C
(Temperature changes linearly from top to bottom.)
ac = 10.26 IJEf' C
a,g =9.07 !JEf'C
Figure 10- Axial tensile stress on concrete surface vs. longitudinal location for
different bond-slip stiffnesses (p = 0.00739 and L = 1.524 m)
250
ii
~ 50
0 ~ AI 140
'""
-50
x (em)
)-+-perfectbond -e-0.433E9N/m ......... o.175E9N/m)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
! 7.62
a. = 10.26 !1£1" c
11.88 !l£!"C a,.=
t 10.16 a ,,9 = 9.07 jl£!" C
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
!!l 12.7
ii 15.24
>
t. 17.78
20.32
22.86
25.4 +------------~~-......-'-'l---tl>-----~-----1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Crack Width (mm)
Figure 12- Pavement depth vs. crack width for different bond-slip stiffnesses
(p = 0.00739 andL = 1.524 m)
7~--------------------------------
=
e ..,(28) 355.6 11£
AT at top of slab= -27.8 °c
AT at bottom of slab= -16.7 °C
(Temperature changes linearly fro
top to bottom.)
a c =6.84 11£!" C
a,9 =9.07 j1£/°C
Figure 13- Axial tensile stress on concrete surface vs. longitudinal location for
different slab-segment lengths (p = 0.00739; bond-slip stiffness= 0.433E9 Nlm)
ljl
) ) )
~~----------------------------------------------------~
x(cm)
Figure 14- Axial stress in GFRP rebar vs. longitudinal location for different slab-
segment lengths (p = 0.00739; bond-slip stiffness= 0.433E9 N/m)
2: 12.7
ii 15.24
i;
D. 17.78
20.32
22.86
25.4 +------------+---r-------&--,,.---------.-------,---------.1.----------l
0 0.2 0.4 0.6 0.8 1.2 1.4
Crack Width (mm)
Figure 15 -Pavement depth vs. crack width for different slab-segment lengths
(p = 0.00739; bond-slip stiffness= 0.433E9 Nlm)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: This paper presents eight field applications of an innovative composite pile in
marine environments. The pile consists of a concrete-filled circular fiber-reinforced
polymer (FRP) tube. No internal reinforcement was necessary in this system since the
FRP tube acted as reinforcement in the axial and hoop directions, in addition to its
function as a permanent formwork for the concrete. The field applications utilized the
innovative piles in a variety of configurations including fender piles, dauphins, supports
for marine structures and for bridge pier protection. In these applications, the pile is
subjected to a variety of loading conditions including different levels ofbending and
axial compression loads. Prior to using the new piles in field installations, a large
experimental program was initiated to examine the structural performance of the piles in
bending, under axial loads and under combined bending and axial loads. The test results
were used to verify analytical models, which were then used to develop simple design
charts to be used by the designers. This paper describes the different field applications of
the composite pile, a summary of the experimental program including test results, and the
design charts.
161
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
INTRODUCTION
Piles have traditionally been fabricated using steel, concrete, and timber. These piles
have limited service life and high maintenance costs when used in harsh marine
environments due to corrosion, degradation, and marine borer attack (1 ). Fig. 1 shows
examples of deteriorated conventional piles (2). It has been estimated that repair and
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
replacement of piling systems costs the U.S. over $1 billion annually (3). High repair and
replacement costs have led North American highway agencies and researchers to
investigate the feasibility of using composite materials for transportation and civil
engineering structures including pile foundations (4).
With the increasing demand for the new CFFT composite piles, a large experimental
program has been conducted over the past six years to study the structural performance of
CFFT under axial compression, bending and combined loading (5), (6), (7) and (8). The
studies examined different laminate structures of the FRP tubes, different wall thickness,
and the effect of central hole inside the concrete core. The studies have resulted in the
The following sections provide description of the CFFT composite piles, various
potential applications in marine environments and the fabrication process of the piles.
Actual field projects are then described, followed by a summary of the experimental
program used to evaluate the structural performance of the piles. The design charts
developed to asses the design engineers are also presented and discussed.
The CFFT pile consists of a GFRP tube filled with concrete. The GFRP tubes have a 51
percent fiber volume fraction and fabricated using the filament winding method. The wall
thickness of the tube consists of alternating layers of E-glass fibers embedded in epoxy
resin, oriented in the circumferential and longitudinal directions at angles of 88 and 3
degrees with respect to the longitudinal axis, respectively. The FRP shell behaves as a
membrane subjected to a bi-axial state of stresses acting in the longitudinal and hoop
directions and the fibers oriented in both directions are envisaged to replace the
conventional longitudinal and helical steel reinforcement in reinforced concrete piles. A
0.64 mm thick liner is provided at the inner surface of the GFRP tube to facilitate the
removal of the tube from the mandrel after fabrication as well as isolate the glass fibers
from the concrete core. Two-part marine grade coating is applied to the outer surface to
provide UV protection in full sunlight and protection in both the splash zone and total
submersion conditions in salt water. Table 1 provides the geometric and mechanical
properties of five standard size CFFT composite piles, including diameter, structural wall
thickness, elastic modulus and tensile strength in the axial and circumferential directions,
compressive strength in the axial direction, and Poisson's ratio of the tubes.
The concrete core incorporates the following mix ingredients per unit cubic meter: 360
kg of type 10 cement, 89 kg of CONEX (expansive agent used to compensate for any
volume reduction due to shrinkage), 803 kg of sand, 998 kg of 3/8 in. stones, 154 kg of
water, 4.8 kg of superplasticizer, and 5 percent of air content. The designed slump is 8
inches to provide good workability conditions. The 28 days design compressive strength
of the concrete is 42 MPa.
The CFFT has several unique advantages. The GFRP tube provides a permanent
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
formwork for the fresh concrete. At the same time, it provides flexural reinforcement and
confinement for the concrete core. The concrete core is protected by the impermeable
GFRP tube from corrosive agents, marine boars and biological attacks. On the other
hand, the pile contains no toxins or heavy metal coatings that can leech into sensitive
tidal water. The pile is non-conductive, which is ideal for electromagnetically sensitive
installations. The GFRP tube could be produced with roughened external surface to
increase the skin friction for additional load capacity or uplift resistance. The piles are
The CFFT composite pile can be used in marine environments in a verity of applications.
The most common applications include:
(a) Fender Piling: In front of marine structures, piles are used extensively as vertical
fenders, as shown in Fig. 2(a). The fender piles act as a buffer to absorb and dissipate the
impact energy of the ship during berthing and prevent vessels from going underneath the
pier.
(b) Dauphins: Dauphins are groups of piles, as shown in Fig. 2(b), placed near piers and
wharves to guide vessels into their moorings, to keep them away from structures or to
serve as mooring points.
(c) Light Structural piling: The piles can be used to support the loads oflight-duty piers
and wharves. In this application, bracing between piles is normally used to increase the
strength and stiffness of the foundation of the structure as shown in Fig. 2(c).
(d) Bridge Pier Protection: Piles and dauphins can also be used to create protective
structures for bridge piers and to guide vessels into the channel and away from bridge
supports as shown in Fig. 2(d)
GFRP hollow tubes are delivered to the site as shown in Fig. 3(a). Wooden plugs are used
to seal the ends of the tubes. The plugs are secured in place using metal straps, which are
connected to the composite tube using s.crews. The plugs on one end have a hole to
facilitate pumping concrete into the tube as shown in Fig. 3(b). The tubes are placed in an
inclined position on a steel frame. Ready mix concrete is pumped from the upper end of
the tube until it is totally filled as shown in Fig. 3(c). External vibrators are attached to
the supporting frame, as shown in Fig. 3(d), in order to prevent forming of air pockets
inside the concrete core. After the tubes are filled with concrete, the hole in the upper
plug is sealed. The number of lifting points depends on the length of the pile. The tubes
are placed on a flat smooth surface and left until the concrete is hardened. Later, the piles
are transferred to the installation location. The CFFT composite piles can be driven by all
conventional equipment used for traditional piles. Normal driving helmets may be used
as would be for steel or concrete piles.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The following sections describe eight selected field applications of the CFFT composite
pile in marine environments. The unique features of each project are highlighted.
The New York State Canals, a division of the New York State Thruway, is rehabilitating
the Erie Canal to make it more user friendly for pleasure craft. This program includes the
construction of additional pier and docking facilities. A total of 120 CFFT composite
piles of 323 mm diameter and 16.2 m long were used to support the new facilities as
shown in Fig. 4. The piles were used in a configuration similar to that shown in Fig. 2(c).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Flint, Allen, White & Radley PC, of Henrietta, NY, designed the dock facility at
Brewertown (Lock E-23). Construction and pile driving was performed by Slate Hill
Contractors. The design was challenged by unique subsoil conditions. The drivable soils
were only 6.1 to 7.6 m deep before rock was encountered. Although the soils were stiff
enough to support the finished structure, there was concern that ice would heave the piles
out of the soil. The composite piles were produced with a roughened surface over the
bottom 7.6 m for maximum skin friction. The top of the pile was produced with a smooth
slick surface to allow the ice to slide up and down the pile, minimizing the likelihood of
the pile heaving up. Three winters have passed and the engineering team is able to report
to the owners, that the structure is in good condition without any movement of the piles.
The project was completed in 1998.
The New York and New Jersey waterfront is contaminated with at least three different
types of marine borers, which are capable of causing severe damage, even to the heavily
treated wood piles. Thousands of existing wooded structures are vulnerable to marine
borer attack. At the same time, State and local environmental agencies have banned the
use of toxic wood treatments. The Brooklyn Navy yard is in the process of a privatization
program and many waterfront portions of the facility are in need of serious repair.
Standards and specifications have been established for the piling materials that are
acceptable to the owners, and the CFFT composite piles have been selected for these
projects. A total of 38 CFFT composite piles of 323 mm diameter and 25 m long were
used in a structural application to support a wharf extension as shown in Fig. 5, using a
configuration similar to that shown in Fig. 2(c). McLaren Engineers Nyack, NY designed
this project and has been instrumental in forwarding the use of composites on waterfront
infrastructure projects throughout the NYC area and on a national level through
workshops and ASTM committee work. The project was completed in 1999.
The Municipal Marina for the City of Belmar, New Jersey, is located on the Shark River,
near the ocean inlet along the south Jersey shore. The marina, which houses pleasure craft
and commercial head boats, is in a rebuilding process, where it is making a transition
US Naval Station Ingleside in the Gulf of Mexico houses the Navy's mine sweeper class
vessels that come for routine degaussing (a process that removes electronic fuzz and
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
static from the very sensitive detection equipment on board). At such facilities, steel and
iron are not allowed to be used in construction of the pier. Composite CFFT piles were
the most suitable for such requirements. Gee & Jenson as well as Whitney, Bailey, Cox
and Magnani consultants were responsible for the design of the piers and composite piles.
Southern Division NA VF AC coordinated the design efforts and the project was built by
Orion Construction. The project utilized the composite piles in different applications
including fender piles as shown in Fig. 2(a), structural battered (inclined) and vertical
piles as shown in Fig. 2(c) and dauphin clusters as shown in Fig. 2(b). A total of 180
CFFT composite piles of 367 mm diameter and 22m long were used as shown in Fig. 7.
The project is now over five years old and has survived a major hurricane without
damage. The project was completed in 1998.
As the Navy moves to consolidate the Pacific Fleet at Naval Station San Diego and the
surrounding support activities, there has been a need to upgrade waterfront structures.
Many piers and wharves have been repaired and/or replaced. Pier 5000, located at Point
Lorna, at the southern tip of the San Diego Bay, is one of four major piers that serve as
home base for several submarines. A total of 36 CFFT composite piles of 323 mm
diameter and 21.3 m long were used for the intermediate fender piles and for the corner
clusters on this pier, as shown in Fig. 8, in a configuration similar to that shown in Fig.
2(a). Because of the aggressive nature of this fender application, the designers, Blaylock
Engineering, have specified an abrasive resistant rub strip consisting of a 50 mm thick
wall HDPE pipe slid over the piles, where the berthing vessel comes in direct contact, for
protection against abrasive abuse. The project was completed in 1998.
The Navy is in the process of renovating the ammunition pier at Port Hadlock, on Indian
Island, Washington. Deep draft transport ships utilize this facility, and accordingly, long
fender piles are necessary to protect the pier. A total of 138 CFFT composite piles of 418
mm diameter and 25.9 m long were used as shown in Fig. 9 in a configuration similar to
that shown in Fig. 2(a). The tubes were filled with concrete at the production yard of
Bellingham Marine. The piles were transported to the job site by way of barge to
minimize costs. The piles have been fitted out with a 38 mm thick HDPE sleeve to
protect from abrasion abuse. Blaylock Engineering was the consultants for the Navy and
Oceaneering International was the general contractor organizing the project. The project
was completed in 2002.
Many piers and wharves are going through significant rehabilitation in San Diego as the
Navy consolidates the Pacific Fleet. Many of the repair projects involve an upgrade of the
fendering system. Several design criterions are considered in the selection of the fender
system. In particular, life cycle value evaluation has become a genuine priority for these
major maintenance projects. Blaylock Engineering, were the consultants for the Navy on
this project at Pier Bravo, located on Coronado Island in the San Diego Bay. Marathon
Construction was the marine contractor who performed the extensive repairs to the pier.
A total of 160 CFFT composite piles of 323 mm diameter and 19.8 m long were used as
shown in Fig. 10, using a configuration similar to that shown in Fig. 2(a). In order to
meet very tight scheduling demands and also to eliminate transportation costs, the fender
piles were fabricated at the job site. The project was completed in 2002.
The project consists of self supporting fender walls used to protect the piers of the Lake
Ponchartrain Bridge. The three and five pile support clusters will be spaced at 2.4 m in a
configuration similar to that shown in Fig. 2(d). The pile clusters will be connected with
ten horizontal wales, spaced at 0.5 m. Stainless steel cables and fittings are used for
connections. A total of 350 CFFT composite piles, 323 mm diameter and 19.8 m long
have already been fabricated locally by Boh Brothers, New Orleans at a rate of 50 piles
per day, as shown in Fig. 11. The project will begin the construction phase in the spring
of 2003. Design was completed by Whitney, Bailey, Cox and Magnani, Baltimore, MD.
The Greater New Orleans Expressway Commission will oversee the final design package
and be responsible for the construction of the fender approach walls. The project is
expected to be completed in summer 2003.
CFFTPILES
A comprehensive experimental program has been conducted over a period of six years to
evaluate the structural performance of the CFFT composite piles. During this period, test
The test piles were fabricated using the same technique specified for fabrication of the
production piles. The process was described earlier in this paper and shown in Fig. 3.
Typically, the specified compressive strength of the concrete used in the CFFT composite
piles at 28 days is 42 MPa. At the time of tests, however, the concrete compressive
strength of the test specimens was 60 MPa.
This test simulates applications such as fender piling, where the pile is mainly subjected
to lateral loads and bending. Two identical specimens were tested as simple beams, using
four-point bending as shown in Fig. 12(a). The span of the beams was 5.5 m while the
distance between the two applied loads was 1.5 m. Specimens were instrumented within
the constant moment region to measure the longitudinal and circumferential strains along
the depth. Mid-span deflection and applied load were also measured. The piles were
tested to failure to determine the flexural capacity. Failure occurred by rupture of the
GFRP tube in the tension side of the pile within the constant moment region as shown in
Fig. 13(a). Fig. 14 shows the load-deflection behavior of the piles. The behavior is
almost linear and the cracking moment is relatively low compared to the ultimate
moment capacity.
of the GFRP tube and concrete core, has been developed (13). The model is capable of
predicting the moment-curvature response of the pile in bending, which is used to predict
the load-deflection response as shown in Fig. 14. The model predicted well the flexural
response of the pile and was also verified using several other test specimens (7).
This test represents pile applications such as vertical structures support of piers and other
structures, where the pile is mainly subjected to axial compression loads. Two identical
specimens, 1 m long each, were tested to failure under concentric axial compression
loads as shown in Fig. 12(b). Axial and circumferential strains were measured using both
displacement and electric strain gauges along the perimeter of the tube, at mid-height.
The piles failed by fracture of the GFRP tube under a bi-axial state of stresses including
axial compressive and hoop tensile stresses as shown in Fig. 13 (b). The axial stress in
the tube is a result of the direct axial loading on the tube, while the hoop stresses results
mainly from confinement effects, which is attributed to expansion of the concrete inside
the tube under high axial stresses. Fig. 15 shows the axial load-axial strain behavior of
A theoretical confinement model has been developed to predict the stress-strain behavior
of FRP-confined concrete, which is used to predict the axial load - strain behavior of the
pile (5). The model is based on equilibrium and radial displacement compatibility
between the outer shell and the concrete core and accounts for the effect of the axial loads
induced in the FRP tube and the resulting bi-axial state of stresses developed in the FRP
tube. Fig. 15 shows the predicted axial load-strain response of the tested pile. The model
has also been verified using several other tests (5).
In several marine pile applications, the pile is subjected to both lateral and axial loads,
therefore, an experimental study was conducted to test several 1.8 m long portions of the
pile under combined bending and axial compression. Rigid steel caps were installed at the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
top and bottom of the test pile specimens, as shown in Fig. 12(c). An eccentric axial load
was applied at different eccentricities (on different specimens) until failure occurred,
either on the tension or the compression side of the pile. At large eccentricities, tension
failure occurred by rupture of the GFRP tube as shown in Fig. 13(c). At small
eccentricities, compression failure occurred by crushing of the GFRP tube as shown in
Fig. 13(d). Due to the nature of the load application, the applied axial force and bending
moment were coupled. The total maximum moment at mid-height is composed of the
primary moment, based on the initial eccentricity, and the secondary moment due to the
lateral deflection at failure at mid-height. Fig. 16 shows the axial load-bending moment
interaction diagram of the pile based on the measured axial loads and total bending
moments at failure as well as the beam and column test results. The behavior reflects very
well the transition from tension to compression failure through the balanced point.
A theoretical model has been developed to predict the full interaction curve of the
composite pile (8). The model accounts for variable confinement of concrete as a result
of the gradual change of the eccentricity of the applied load, which also result in a change
of the bi-axial state of stresses developed in the GFRP tube. The model utilizes the
classical lamination theory for the GFRP tubes and accounts for the gradual reduction of
stiffness as a result of the progressive failure of different layers. The predicted
interaction curve of the pile is given in Fig. 16 and shows good agreement with the
experimental data points.
Simple, yet accurate, design charts for the CFFT composite pile were requested by the
design engineers for the different field projects. Based on the comprehensive
experimental and analytical studies, theoretical models were developed and verified as
described earlier. The models were then used to develop design charts for the four
different sizes of the CFFT composite pile given in Table 1. The charts were based on the
material mechanical properties specified by the manufacturer of the GFRP tubes,
provided in Table 1, as well as the specified concrete compressive strength at 28 days (42
MPa).
Fig. 17 shows the moment-curvature design charts for the four sizes of the CFFT
composite piles, when subjected to pure bending. The charts provide the full response up
to failure. Also shown in the same charts are the cut-off lines representing different levels
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
of factor of safety (FOS) of 1.5, 2 and 3. For a given design moment, the most suitable
size based on the desired safety factor can be determined. The corresponding curvature is
then used to calculate the deflection of the pile using the moment-area method.
Fig. 18 shows the axial load- strain design charts of the composite piles, when subjected
to axial loads as short columns. The figure also shows the cut-off lines based on safety
factors of 1.5, 2 and 3. However, in most applications, slenderness effect will produce
additional bending moments in the piles. In this case, the interaction charts in Fig. 19 can
be used.
The full axial load - bending moment interaction design charts of the different sizes of
the CFFT composite piles are provided in Fig. 19. For a given design bending moment
and axial load, the charts will assist the designer in selecting the appropriate size of the
pile. The charts also show the locus of the balanced failure points on the curves.
It should be noted that these charts are developed for a specific CFFT composite pile
used for the field applications specifically stated in this paper, due to the lack of a general
design guidelines and simple analytical tools. However, efforts by ACI Sub-Committee
440-J (FRP Stay-in-Place Formwork) are underway to develop a unified design approach
and guidelines for this type of structures. Once this tool become available, there will be
no need for a product-specific design charts.
1. Concrete-filled FRP tubes (CFFT) have indeed gained large potential for marine pile
applications. The system is becoming more popular as it provides several advantages
over conventional piles. Non-corrosive characteristics, elimination of internal steel
reinforcement and utilizing the FRP tube as reinforcement and permanent formwork are
the most unique features.
2. Eight different field projects have been successfully completed using the CFFT
composite piles along the east and west costs of the United States. The piles were utilized
3. A total of 1514 CFFT composite piles have been used in the eight field applications
combined. The piles ranged from 323 to 418 mm in diameter and from 15 to 26m in
length.
4. The composite piles in the field projects have been in service for several years and
some went through severe hurricanes. To date, based on regular inspections, the piles are
performing very well.
5. A large experimental and analytical research program has been conducted to study the
structural performance of the piles using full scale specimens of the most popular size
(323 mm diameter). The data was used to verify analytical models, which were used to
develop design charts of the different sizes.
While this paper presented a number of field installations that the authors have been
involved in, it should be mentioned that several other field installations by other
researchers and industries have also been successfully completed, using CFFT.
ACKNOWLEDGEMENT
REFERENCCES
(2) Iskander, M.G., and Stachula, A. (1999) "FRP Composite Polymer Piling:
An Alternative to Timer Pilling for Water-Front Applications." Geotechnical
News, 4(Dec):27-29.
(3) Lampo, R. (1996) "Federal Interest Gives Recycled Plastic Lumber a Leg up"
ASTM Standarization News, 26-31.
(4) Iskander, M., and Hassan, M. (1998) "State of the practice review in FRP
Composite Piling." Journal of Composites for Construction, ASCE, 2(3 ), 116-
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
120.
(5) Fam, A.Z. and Rizkalla, S.H. (2001) "Confinement Model for Axially Loaded
Concrete Confined by FRP Tubes", ACI Structural Journal, 98(4):251-461.
(7) Fam, A.Z. and Rizkalla, S.H. (2002) "Flexural Behavior of Concrete-Filled Fiber-
Reinforced Polymer Circular Tubes", ASCE Journal of Composites for
Construction, 6(2): 123-132.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(8) Fam, A., Flisak, B. and Rizkalla, S. (2002) "Experimental and Analytical
Investigations of Beam-Column Behavior of Concrete-Filled FRP Tubes", ACI Structural
Journal, accepted for publication.
(10) Pando, M., Filz, G., Dove, J., and Hoppe, E. (2002) "Interface Shear Tests
on FRP Composite Piles." Paper accepted to ASCE International Deep
Foundations Congress, Orlando, Florida, February 2002, 15 pp.
(11) Seible, Frieder (1996) "Advanced Composites Materials for Bridges in the 21st
Century," Proceeding of Advanced Composite Materials in Bridges and Structures,
Montreal, pp.17-30.
(12) Karbhari, V. M., Seible, F., Burgueno, R., Davol, A. and Zho, L. (1998) "Structural
Characterization of Fiber-Reinforced Composite Short-and Medium-Span Bridge
System," Proceeding ofECCM-8, European Conference on Composite Materials, Vo1.2,
Ed. by I. Crivelli Visconti, 3-6 June, pp.35-42.
(13) Fam, A.Z. (2000) "Concrete-Filled Fiber Reinforced Polymer Tubes For Axial and
Flexural Structural Members," Ph.D. Thesis, The University of Manitoba, pp. 261.
Table I. Properties of the GFRP tubes used for CFFf composite piles
Geometrical and Mechanical Properties Standard Sizes
of the GFRP Tubes 10.7 in. 12.7 in. 14.4 in. 16.5 in.
Outer diameter (mm) 273 323 367 418
Structural wall thickness (mm) 3.7 4.4 4.7 5.2
Young's modulus (Axial tensile) (GPa) 27.6 27.6 24.1 21.4
Young's modulus (Axial compressive) (GPa) 27.6 27.6 24.1 21.4
Young's modulus (Hoop tensile) (GPa) 19.3 19.3 22.1 23.4
Axial tensile strength (MPa) 483 483 421 372
Axial compressive strength (MPa) 483 483 421 372
Hoop tensile strength, (MPa) 241 241 276 293
Poisson's ratio (Circ. I Long.) 0.19 0.19 0.16 0.15
(a) Corroded steel pile (b) Degraded concrete pile (c) Deteriorated timber pile
(a) Roughened surface of the pile (b) Canal House, Brewerton Lock
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(a) 25 m long pile lifted with single pick (b) Piles ready for driving
(a) Mooring piles (b) Structural piles to pin finger piers in position
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
160 .. , I
I
140
Theoretical
,...._ 120
g 100
-g 80
j
60
40
~oad
20
i-2---i
Deflection
0
0 50 100 150 200 2 50 300
Deflection (mm)
5000
~
-g
j
4000
\n,.-..m.,..]
Electric strain gage Experimental
3000
] 2000
1000
0~--~---~--~--~----~
4
Axial Strain X 10·3
6
r 10
..-··..
~
7000
6000
.. .. . . J. Theoretical
~ 5000
-g .•••.
j
3
4000
3000
..
~
2000 Experimental ~
1000
.. .. .,
0
0 50 100 150
-• --·""·
200 250 300
Moment (kN.m)
Fig. 16 -Interaction curve of the test pile
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
52oo~~~~~~~~1T~~~~~~~~7+~~~~
~ 150~~~~~=,~~~~~~~~~~~
J 100~~~~~~~~~~~~~~~~~~~~~~~~~
so~~~~~~~~~~~~~~~~~~~~~~~~~~
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Curvature (lim)
8000
7000
:.: ~ :.:
g-
6000
-.- t -, -
5000
'i0
-
' il
4000
3000
:·:.: -,-
~ 2000
1000
:. : ~ - -
0
0 0.002 0.004 0.006 0.008 O.Q1 0.012
Axial strain (rnrnlmm)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
8000~~~~~~~~~~~~~~~~~
---
~5000~~~~~~~~~~~ ,__...,......_
] 4000 +-i-_;_;_~_._._......_.:~........_-'-1-......_..-4.
c;
~
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
181
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
182 Pando et al.
Miguel Pando, is an Assistant Professor at the University of Puerto Rico, Mayaguez,
Puerto Rico. He received his M.Sc. from the University of Alberta, Canada in 1995 and
his Ph.D. from Virginia Tech in 2003. His research is focused on geotechnical aspects of
concrete-filled FRP tubes and soil-structure interaction.
Amir Fam, ACI member, is an Assistant Professor at Queen's University, Kingston, ON,
Canada. He received his M.Sc. and Ph.D. from the University of Manitoba, Canada in
1996 and 2000, respectively. He is Co-Chair of Sub-Committee 440-J, FRP Stay-in-Place
Formwork. His research is focused on structural applications of concrete-filled FRP
tubes.
INTRODUCTION
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Foundation piles are typically made of materials such as steel, concrete, and timber.
Problems associated with use of these traditional pile materials in harsh environments
include steel corrosion, concrete deterioration, and marine borer attack on timber piles. It
has been estimated that the U.S. spends over $1 billion annually in repair and
replacement of waterfront piling systems (1 ). Such high repair and replacement costs
have led several North American highway agencies and researchers to investigate the
feasibility of using composite piles for load bearing applications, such as bridge
substructures. Composite piles may exhibit longer service lives and improved durability
in harsh marine environments, thereby presenting the potential for substantially reduced
total costs. At present, the initial cost of composite piles is generally greater than the
initial cost of traditional piles. Also, questions related to performance such as driving
efficiency, axial stiffness, bending stiffness, durability, and surface friction, are typically
asked by engineers and designers. These questions exist because of the lack of long-term
track records and scarcity of well-documented field tests on composite piles.
RESEARCH SIGNIFICANCE
The study reported in this paper included installation of concrete-filled FRP tubes (CFFT)
piles as production piles to support the piers of two new bridges in Virginia, namely: the
Route 40 and Route 351 Bridges. This will allow monitoring the long-term performance
of these piles and to compare with conventional prestressed concrete piles. The field
tests presented in this paper were also undertaken to investigate the engineering
performance of CFFT composite piles as load-bearing foundation elements in bridge
support applications. The objectives of this study were to evaluate the driveability and
the axial and lateral load behaviour of CFFT piles and compare their performance to
The former bridge, which consisted of a steel truss supported by concrete piers, suffered
from excessive deterioration in both superstructure and substructure. Additionally, the
bridge was functionally obsolete with a roadway clearance of only 7.3 m (24 ft) curb to
curb and was rated as scour critical. The new bridge is 85.3 m (280ft) long and 15.2 m
(50ft) wide, and consists of a five span slab-type superstructure supported by four piers
and two end abutments. Each pier consists of a reinforced concrete beam-type elevated
pile cap supported by ten piles. Based on the design requirements, VDOT specified a 508
mm (20 in.) prestressed solid square pile, as shown in Figure 2(a). The design axial load
of each pile was 667 kN (150 kips). As a part of a composite pile study sponsored by the
Federal Highway Administration (FHW A), VDOT took the initiative to use the
alternative CFFT pile system for the entire group of piles supporting one of the piers of
the bridge. The objectives of the VDOT initiative were to examine the feasibility of using
composite piles for bridge substructures, particularly to examine the driveability of the
piles using conventional pile drivers, to assess performance under axial and lateral
loading conditions using full-scale field tests, and to compare composite piles to
conventional prestressed concrete piles.
Prior to construction of the new Route 40 Bridge, two piles were installed and tested in
the field, including a CFFT composite pile and a conventional prestressed concrete pile.
The following is a brief description of the two piles:
(a) CFFT Composite Pile: Details of the CFFT composite pile are provided in Figure
2(b ). The", FRP tube used for this pile is a Glass-FRP tube, type CP40, supplied by
Lancaster Composites Inc., of Lancaster, Pennsylvania. The 13.1 m (43ft) long tube was
fabricated using a filament-winding technique, where E-G1ass continuous fiber roving
were impregnated with polyester resin and wound over a rotating steel mandrel,
following a predetermined winding pattern. The average measured fiber volume fraction
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(b) Prestressed Concrete Pile: Details of the prestressed concrete pile are provided in
Figure 2(a). The prestressed concrete test pile, which is commonly specified by VDOT, is
a 508 mm (20 in.) square pile with a length of about 13.1 m (43 ft). The pile is
prestressed using fourteen, 12.7 mm (0.5 in.) diameter seven-wire strands of 1861 MPa
(270 ksi) ultimate strength, pretensioned to produce a prestress level of 5.6 MPa (0.809
ksi) in accordance with VDOT standards. The specified concrete compressive strength at
28 days is 40 MPa (5.8 ksi). The ties consist of No. 16M (No. 5) gage spiral wire.
Concrete test results indicated a 28-day compressive strength of about 43 MPa (6.2 ksi)
for this pile.
The prestressed concrete test pile and the production piles were fabricated at a precast
plant in accordance with VDOT standards. After fabrication and curing, the piles were
shipped to the bridge site. For the CFFT composite test and production piles, the FRP
tubes were shipped to the precast plant and filled with concrete while held in an inclined
position and supported by a steel beam along the full length of the tubes, as shown in
Figure 3 (a and b). Wooden end plugs were used to seal both ends of the tube. The plugs
were secured in position using metal straps connected to both the plug and the composite
tube in radial directions. Concrete was pumped into the upper end through a hole in the
wooden plug. The CFFT were handled using eight-point lifting devices along the length
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Both the composite and prestressed test piles were driven using a hydraulic impact
hammer, type ICE model 160S, which was also used to drive the production piles for the
Route 40 Bridge. The ram weight is 71.2 kN (16,000 lbs. A 190 mm (7.5 in.) thick
plywood pile cushion was used during pile driving. Figure 4(a) shows the pile driving
process. Table 2 provides some of the measurements obtained during pile driving. It can
be seen that the wave speeds, maximum compressive stresses, and maximum tensile
stresses are each similar in the different pile types. The measured stress levels are lower
than the allowable stresses recommended for prestressed piles. No standards are available
yet for driving of CFFT composite piles.
Prior to installation of the production piles for the new Route 40 Bridge, the two test piles
were subjected to axial and lateral loading using a Statnamic testing device. This system
applies the load to the pile by combustion of fuel within a closed chamber, in which gas
pressure builds and "launches" a heavy reaction mass upwards (or laterally) at a very
high acceleration close to 20 g's. When the reaction mass is launched upwards (or
laterally), an equal and opposite reaction force acts on the test pile (4 and 5). The
Statnamic setup is such that the load is applied in a somewhat gradual manner, as
opposed to an impact load like the one produced by a driving hammer. The applied force
is a time-dependent load with duration typically between 0.2 to 0.3 seconds. From the
recorded Statnamic force, one can estimate the equivalent static load by subtracting the
inertia and damping forces. This equivalent static force can be derived by using the
Unloading Point Method (UPM) proposed by Middendrop et al. (4). A summary of the
main results of the axial and lateral load tests are presented below. Details of the test
results and the instrumentation can be found in (2) and (3).
(a) Axial load Tests: Axial Statnamic load tests were carried out on both test piles, seven
days after pile driving. The axial load setup using the Statnamic testing system is shown
in Figure 4(b ). During the Statnamic test, several measurements are taken, including the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
applied load, the pile head displacement, the toe accelerometer readings, and strains. Pile
head displacement is measured using a laser beam device. Each test pile was subjected to
three loading cycles with increasing magnitude of applied load.
The equivalent static load versus pile head axial displacement for the prestressed concrete
and composite piles for the three cycles are shown in Figure 5(a). The behavior during
the last cycle in both piles shows that the geotechnical capacity of the piles was fully
mobilized, as evident from the rapid increase of the pile displacement near the end of the
test. The ultimate load capacities of the CFFT composite and prestressed concrete piles
were 4359 kN (980 kips) and 4190 kN (942 kips) respectively, while the head
displacements were 18 mm (0.71 in) and 13 mm (0.51 in.) respectively. Both piles
exhibited similar axial capacities, even though their shaft and end areas are slightly
different.
Figure 5(b) shows the equivalent axial static load versus axial strain behaviour of both the
composite and prestressed piles based on the peak load and corresponding strain during
each of the three load cycles. The axial strain is based on the average strain of the two
uppermost strain gauges (closest to the loading end), since they measured the maximum
axial strain induced in the pile. Figure 5(b) shows that the behaviour of the two piles is
relatively similar with the composite pile being slightly stiffer than the prestressed pile.
The figure also shows that the ultimate load is higher than the design load of the piles.
(b) Lateral load Tests: Following the axial load tests, lateral load tests were conducted
on both the composite and prestressed piles, using the Statnamic testing system shown in
Figure 4(c). The loading system is similar to that used for the axial load test; however,
the setup was placed horizontally in order to produce a lateral load. During lateral
The responses of the test piles were measured using displacement transducers and
subsurface accelerometers. For subsurface displacements, a string of eight lateral motion
.accelerometers were used to track the lateral movement profiles of the piles. The peak
lateral displacements obtained from the string of subsurface lateral motion accelerometers
provide the displaced shapes shown in Figure 6 for each of the four cycles of Statnamic
load applied to each pile. Both piles appear to form a hinge at a depth of about 5 m (16
ft.) below the point of loading. For the last two load cycles, the deflection profile along
the depth of the piles is almost bi-linear, with a sudden change in slope occurring at a
depth of about 4.8 m (16ft) measured from the loading point. This suggests that failure
may have initiated at the location of the sudden change in slope during the last two
cycles.
Based on the results of the field testing of the composite and prestressed concrete piles,
VDOT has decided to use the composite piles in the construction of Pier No. 2 of the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Route 40 Bridge (Structure No. 1006) over the Nottoway River in Virginia. The pier
consists of a reinforced concrete cap beam supported by ten CFFT composite piles. The
CFFT production piles are identical to the composite test pile in terms of the size,
laminate structure of the tube, and the methods of construction and driving. The installed
composite piles of Pier No. 2 are shown in Figure 7(a). After both the prestressed
concrete and CFFT composite piles were driven, and prior to casting the cap beams,
special preparation of the pile heads were necessary in order to facilitate connecting the
piles to the cap beams. Eight, 1 in. (25.4 mm) diameter holes were drilled through the top
flat surface of each pile, using a regular rock drill. The holes were 18 in. (457 mm) deep,
parallel to the longitudinal direction of the piles. In the prestressed concrete piles, the
holes were equally spaced in a 13 in. (330 mm) square pattern, whereas in the composite
pile, the holes were equally spaced in a 17.6 in. (447 mm) diameter circular pattern as
shown in Figure 7(b). These arrangements allow for a 3 in. (76 mm) concrete cover for
each hole. Eight, 48 in. (1219 mm) long, No.7 steel rebar were inserted in the holes and
epoxy resin was used to secure the bars inside the concrete. The formwork arrangement
of the cap beams were placed such that the bottom surface of the cap beam is 6 in. ( 152
mm) below the upper surfaces of the piles to allow for embedment of the piles inside the
cap beam. Figure 8 (a) shows elevation views of Pier No. 2. Figure 8 (b and c) show
cross-sections through and in between the CFFT composite piles, where No. 7 steel bars
are embedded 30 in. (762 mm) inside the cap beam. Figure 9 shows a general view of the
new completed bridge and a close up of Pier No.2 supported entirely by CFFT composite
piles.
In 2001, VDOT started a bridge replacement project at the Route 351 Bridge crossing of
the Hampton River in Hampton, Virginia. The general location of the bridge is shown in
Figure 1. In this project, VDOT has also decided to use a CFFT composite pile. Internal
longitudinal steel reinforcement was included in the CFFT pile to increase its flexural
stiffness. The test pile program involved one prestressed concrete pile and one reinforced
CFFT composite pile that were subjected to axial and lateral load testing. Two
instrumented piles, consisting of a prestressed concrete pile and a CFFT composite pile,
have also been installed in the new bridge to study long-term performance and load
transfer behaviour. The focus of the following sections is on description of the piles, pile
driving, dynamic testing, axial static load testing, and lateral static load testing.
The former bridge, built in the early 1940's, was 228.45 m (749.5 ft) long and had a
roadway 10.35 m (34 ft) wide. The rated capacity of the bridge was for 15 ton trucks
(133.5 kN). The bridge was composed of 16 equal spans supported on steel beams. The
superstructure was supported by 15 piers and 2 end abutments. The foundations of the
piers and abutments consisted of groups of timber piles with reinforced concrete pile cap
beams. The bridge suffered from excessive deterioration in the superstructure and
substructure, which led to the decision of replacement. The proposed new bridge is 228.5
m (749.7 ft) long and 12.8 m (42ft) wide, and consists of a thirteen span superstructure
supported by twelve piers and two end abutments. The bridge deck is supported by
AASHTO Type II prestressed concrete beams for most of the bridge. However, for the
first 4 spans on the west side of the bridge, steel plate girders will be used. The bridge
piers, except for Pier No. 2 located under highway I-64, consist of an elevated reinforced
concrete beam-type pile cap supported by a group of eight prestressed concrete piles.
Based on the design requirements, VDOT specified a 610 mm (24 in.) prestressed solid
square piles for the pier foundations. The design axial load for the piles in the piers was
890 kN (200 kips). A fully instrumented CFFT composite pile and a prestressed concrete
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
pile are installed in the actual bridge. The new bridge is expected to be finished and open
to traffic by May of2003.
Prior to construction of the new Route 351 Bridge, a reinforced CFFT composite pile and
a conventional prestressed concrete pile were installed and tested in the field. The
following is a brief description of the two piles:
(a) Reinforced CFFT Composite Pile: The reinforced CFFT composite test pile consists
of an 18.3 m (60ft) long, FRP tube filled with concrete and reinforced with steel bars.
The FRP tube, fabricated by Hardcore Composites LLC, Inc., consists of a GFRP tube
with an outside diameter of 622 mm and a wall thickness of 10.7 mm. The polymer is
Dow vinyl ester Derakane resin, and the fiber reinforcing consists of 4 plies of stitch-
bonded glass fiber fabric. Each ply includes four layers of fibers oriented at 0, +45, -45,
(b) Prestressed Concrete Piles: Similar to the Route 40 Bridge project, the performance
of the CFFT composite test pile was compared with that of a conventional prestressed
concrete test pile. At the Route 351 Bridge, the prestressed concrete test pile is similar in
configuration to that shown in Figure 2(a) but different in size and reinforcement. The
pile is 610 mm (24 in.) square and 18m (59.1 ft) long. The pile is prestressed with 16,
12.7 mm (0.5 in.) diameter, 7 wire strands of 1860 MPa (270 ksi) yield strength steel.
The strands were strained to produce an effective prestress of 5.5 MPa (0.803 ksi) after
losses. The spiral ties consist of No. 16M (No.5) wire with a pitch of 150 mm (6 in.).
The 28-day strength of the concrete was specified to be at least 35 MPa (5,080 Psi), and
the test results show that the strength of concrete was 55 MPa (7,980 Psi) at the time of
the load tests. Figure 11 shows the predicted moment-curvature response using a
procedure described in (7).
Figure 11 shows that the flexural strength of the reinforced CFFT pile is higher than that
of the prestressed concrete pile. This is attributed to the higher steel reinforcement ratio
of the CFFT (2.47 percent) compared to the prestressed pile (0.43 percent). Additionally,
the CFFT composite pile has a GFRP tube that provides a 5.24 percent reinforcement
ratio.
The prestressed concrete test pile was fabricated in general accordance with VDOT
standards at the precast plant of Bayshore Concrete Products Corporation in Chesapeake,
Virginia. After fabricating and curing, the piles were shipped to the bridge site. The FRP
tubes were fabricated by Hardcore Composite Inc. and shipped empty to the casting yard
ofS. B. Ballard, Inc. in Virginia Beach, Virginia. The two FRP tubes were 18.3 m (60ft)
and 21.3 m (70 ft) long, for the test pile and the production pile, respectively. Prior to
concrete pouring, a steel rebar cage was inserted into the FRP tubes as shown in Figure
12(a). The piles were then raised with a crane and secured in a vertical position with the
aid of steel sheet piles and steel bracing. The bottoms of the FRP tubes were fitted into 3-
inch grooves in a concrete slab. The bottoms of the FRP tubes were also bolted to the
concrete slab by means of L-shape angles. The concrete was then poured at the upper end
as shown in Figure 12(b). After fabrication, piles were kept in the same vertical position
for two weeks prior to shipping to bridge site.
The test piles at the Route 351 Bridge were installed at a test site adjacent to the bridge
site. Both test piles were driven using a single-acting diesel hammer type ICE 80S as
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The two test piles were subjected to static axial and lateral loading. A summary of the
main results of the axial and lateral load tests are presented below. Details of the test
results and the instrumentation used can be found in (2).
(a) Axial Load Tests: Static axial load tests were performed on the test piles in
accordance with ASTM Dll43. The compressive test loads were applied by jacking
against a steel frame anchored by eight reaction piles. The load was applied by three
calibrated hydraulic jacks operated by a single hydraulic pump. The load applied by this
jacking system was measured using a 4448 kN ( 1000 kips) calibrated load cell. The
deflections were monitored using different instrumentation including dial gages and
LVDT's. Figure 13(b) shows the axial load test setup. The axial load test was carried out
by increasing the axial load using constant load increments of89 kN (20 kips). After each
load increment, the load was held constant for a time interval of 2.5 minutes in
accordance with the Quick Load procedure included in ASTM Dll43.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 15(a) shows the measured axial load versus pile head displacement diagrams for
both test piles. It can be seen in this figure that both piles exhibit similar initial response,
however, the CFFT pile starts to experience large settlements at a lower axial load. The
pile failure loads obtained were 3095 kN and 2260 kN, for the prestressed concrete pile
and the CFFT composite pile, respectively. The two test piles have different embedded
shaft areas and different toe areas. The strain gages that were installed in the test piles
permitted calculations of load transfer from the piles to the ground. The resulting load
transfer diagrams are shown in Figure 15(b). From these diagrams, the shaft capacities
(skin friction components), are 78 and 65 percent of the total capacities for the
prestressed concrete and CFFT composite piles, respectively.
(b) Lateral Load Tests: Following the axial load tests, static lateral load tests were
performed in accordance with ASTM D3966, using the same reaction system as used for
the axial load tests. A calibrated jack and a calibrated load cell were used.
Displacements near the pile head were monitored with dial gages and LVDT's. Deformed
shapes of the piles at several lateral load magnitudes were obtained from standard
inclinometer devices installed in the piles.
In November 2002, VDOT has started the pile installation process at the Route 351
Bridge crossing of the Hampton River in Hampton, Virginia. The VDOT has decided to
install one fully instrumented CFFT composite pile to compare its long term structural
and geotechnical performance to that of a fully instrumented conventional prestressed
pile. The two instrumented piles were installed at the center of two adjacent piers, Pier
No. 10 and Pier No. 11 as shown in Figure 17. The connection details of the piles to the
reinforced concrete cap beam is very similar to that used in the Route 40 Bridge as shown
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
in Figures 7 and 8, except that the piles were embedded 12 in. (300 mm) inside the cap
beams and that the connecting steel bars are embedded 60 in. (1.52 m) inside the piles.
Construction of the bridge is currently on going and is expected to be completed by May
2003. Long term monitoring of the instrumented piles will be carried over the next few
years.
In 2000, the Virginia Department of Transportation (VDOT) replaced the old Route 40
Bridge (Structure No. 1006) over the Nottoway River. Ten new composite piles,
consisting of concrete-filled Glass-FRP tubes (CFFT), were used to support one of the
reinforced concrete cap beams that directly support the superstructure. Prior to
construction, full-scale field tests were undertaken to investigate the feasibility of
construction, handling, and drivability, as well as the structural performance of the new
composite piles in comparison to conventional prestressed concrete piles. In 2002, VDOT
has also decided to expand their investigation of the new pile by using fully instrumented
CFFT pile with internal steel reinforcement in the new Route 351 Bridge over the
Hampton River for long term monitoring. Based on the testing, analysis, and
construction experience, the following conclusions can be drawn:
1. The use of concrete-filled FRP tubes (CFFT) as piling for bridge piers is practical and
feasible. The FRP tube serves as a permanent formwork and reinforcement at the same
time.
2. Both the CFFT composite and prestressed concrete piles performed similarly during
pile driving, as demonstrated by driving records, PDA traces, wave speed generated by
the driving hammer, and the measured stresses. No discemable damage was detected in
the piles after installation, based on PIT testing performed on the piles before and after
pile driving.
3. Both the CFFT composite and prestressed concrete piles performed similarly under
the axial load tests. Full geotechnical capacity was mobilized in both cases befor~:
ACKNOWLEDGEMENT
REFERENCES
(2) Pando, M. (2003) "A laboratory and Field Study of Composite Piles for Bridge
Substructures" Ph.D. Dissertation, Virginia tech, pp. 396.
(3) Fam, A.Z., Pando, M.A., Filz, G.M., and Rizkalla, S.H. (2003) "Precast Composite
Piles for Route 40 Bridge in Virginia", PC! Journal.
(4) Middendrop, P., Berminghammer, P. and Kuiper, B. (1992). "Statnamic load testing
of foundation piles," Proceedings, 4th International Conference of Applications of Stress-
wave Theory to Piles, The Hague, The Netherlands, 581-588.
(6) Fam, A.Z. (2000) "Concrete-Filled Fibre Reinforced Polymer Tubes For Axial and
Flexural Structural Members," Ph.D. Thesis, The University of Manitoba, pp. 261.
(7) Collins, M. P., and Mitchell, D. (1997). Prestressed concrete structures. Response
Publications, Canada.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Pile Type
Measurement
Prestressed CFFT
Wave speed 12150 rus· 11840 fUs
Max compressive stress
2.55 ksi" 2.78 ksi
measured during driving
Max tensile stresses
0.7 ksi 0.42 ksi
measured during driving
Tension < 1. 02 ksi No standards
Allowable Stresses
Comp. < 4.5 ksi available
Pile Type
Measurement
Prestressed CFFT
Wave speed 12459 fUs' 12400 fUs
Max compressive stress
1.59 ksi" 2.35 ksi
measured during driving
Max tensile stresses
0.81 ksi 1.23 ksi
measured during driving
Tension < 1. 02 ksi No standards
Allowable Stresses
Camp. < 4.5 ksi available
· 1 ft/s = 0.305 mls " 1 Ksi = 6.89 MPa
,,
H:=~~~~//······~~J:.f
Hampton Roads _j.~\
Bridge-Tunnel \\
Route 40 Bridge
43ft.
(a) Prestressed concrete pile
GFRPtube
43ft.
(b) Composite pile
(a) Pumping concrete (b) Rear end showing wooden plugs (c) Lifting the pile at 8 points
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(a) Pile driving (b) Axial statnamic test (c) Lateral statnamic test
Figure 4-Pile driving and axial and lateral Statnamic tests at Route 40 Bridge
1l .0.3
.!1!
ft .0.4
'6
:iii
X -o.s
..,"
~
~ .0.6
~ .0.7
Prestressed pile -
Composile pile -
.0.6
I.., 1000 A
~
\...;;)
(J 800
""s
<J)
~..
c.!!!
600
..
.2:
::l
400
Prestressed pile
~ 200
Figure 5-Field test results of the axial load tests at Route 40 Bridge
BNo.7 8No.7
Prestressed pile
.
13!n.
t]I
I 13;n
20in.
.
-
I
20in
24.6in.
(a) CFFT Piles after driving (b) Steel bars for connection
Figure 7-Route 40 Bridge piles before and after fitting with steel bar connectors
T~
6No.7
No.4@ 10in.
.
I 2 No.4
.5
~
"'
:1l
2No.4
2No.4
l
---- ll:T:
2No.8
,--- -"- ---
' 4 No.8
I•
i :
- ·-··~-'·- ...
(a) Elevation of Pier No. 2 (b) Cross-section (a-a) (c) Cross-section (b-b)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 9-Route 40 Bridge in Virginia (Pier no. 2 fully supported by CFFT piles)
12001-r-------------------,
Reinforced CFFT
1000
~
z 800
c
5l 600 Prestressed
E
0 concrete pile
~ 400
200
0 +---.---.----r--~--~---4
0 5 10 15 20 25 30
Curvature (rad/1000m)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(a) Inserting steel cage into GFRP tube (b) Casting concrete
Figure 13-Pile driving and axial load test at Route 351 Bridge
]:
~~
5 ~;;-----1----i
~---,---,~-,~~
Ill
~
~ 10+-~~--+---+---+---+-~
§
~
~ 15
~
20~~~~~~~~~~~~
0 10 20 30 40 50 60
Pile blows per 0.25 m
Reinforced CFFT
composite pile
4oL=::::::::::;;~l_..........__j
(a) Load- displacement behavior
Load (kN)
5
:[
~10
CD
c
15
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
250
z 200
~
"C
Ill
.Q 150
!!!
s 100
~
50
0
0 10 20 30 40 50 60
Lateral deflection at ground surface (mm)
g 4
~
·a. 6
0
c.
.s 8
~
1i 10 .-'- 51.6kN
-a~ 12
- 96.0kN
·:.....:..-;:·144.8kN.
·~ 186.1 kN
14 ~ 230.1kN'
·~ 270.5kN
16
0 20 40 60 80 100 0 20 40 60 80 100
Deflection (mm)
(b) Deflected shape along the depth
Figure 16-Lateralload test results at the Route 351 Bridge
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CHAPTER3
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
EXTERNAL REINFORCEMENT
PART 1: BRIDGES
Synopsis: This paper presents the application of carbon fiber reinforced polymer
(CFRP) composites to strengthen an existing reinforced concrete bridge in Texas. The
composites were used to increase the flexural strength of the bridge so that it would
comply with the load rating required to qualify for roadway widening. The existing pan-
girders in the bridge required a 53% increase in the nominal flexural strength to meet the
H-20 load rating required by the Texas Department of Transportation. Two
strengthening systems were utilized in different spans of the bridge. The design
methodoiogy of these designs was validated in the laboratory by conducting full-scale
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
tests of strengthened bridge components. Results from the laboratory tests indicated that
the target required flexural strength was achieved with both designs. The bridge will be
inspected periodically to identify any distress that may occur in the composite laminates.
203
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
204 Breiia and Steves
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
INTRODUCTION
A structure that had typical design details common of pan-girder bridges used in
short-spans in Texas built in 1947 was identified as a potential candidate for
strengthening using bonded FRP composite sheets. The bridge is used for the crossing of
F.M. 1362 over Sue Creek in Burleson County, Texas. Inspection of the structure
revealed concrete cracking and spalling at several locations along the bridge, consistent
direction of the bridge. The in-place concrete compressive strength was determined
testing extracted concrete cores and through the use of a rebound hammer. A lower-
bound value equal to f; = 27.6 MPa was determined from these studies and used in all
calculations. The reinforcing steel was assumed to have a yield stress ( JY ) equal to
227.5 MPa.
The existing bridge required widening to meet current geometric and traffic
safety considerations. In order to qualify for widening, internal policies within TxDOT
require a bridge to have a minimum inventory load rating of H-20. Load rating was
conducted using the AASHTO Manual for Condition Evaluation of Bridges (1) with a
rating factor computed from the actions generated from an H-20 truck on the bridge:
RF = C- A1 D (Eq.l)
A2 L ( 1 + I)
where
RF = Rating factor for the live-load carrying capacity of the bridge.
C = Strength reduction factor times nominal capacity of member.
A 1 =Factor for dead load effects= 1.3 for inventory rating.
D = Dead-load effect on the member.
A 2 =Factor for live-load effects = 2.17 for inventory rating.
L =Live-load effect on the member.
I= Impact factor= 0.3
The rating of the bridge was determined using the smallest rating factor
calculated from all the members that contribute to the structural capacity of the bridge:
RT = (RF) W (Eq. 2)
where
RT= bridge member rating, tons.
W = weight of the first two axles of the design truck used in determining the live
load effect, tons.
Results of the load rating on the existing bridge are summarized in Table 1.
Strengthening of the pan-girders was required to increase the rating of the bridge to H-20.
The amount of strengthening required was determined solving Eq. 1 for the live-load
The composite strengthening solutions that were selected for this project were
very similar to schemes used in the experimental phases of this project (2, 3). The goal
of the research project was to develop methods to strengthen existing reinforced concrete
structures using composites. Two composite systems were selected to strengthen each
span of the bridge. One of the composite systems consisted of unidirectional carbon
fibers applied to the concrete surface using a wet-layup procedure (4). The other system
consisted of carbon fiber pultruded plates that are bonded to the surface of the concrete
using an epoxy paste (5). The published properties of the composite materials
corresponding to these two systems are listed in Table 2. The bridge configuration and
support conditions provided the opportunity to test two different composite systems in the
same bridge. Span A was strengthened using the pultruded plate system and span B was
strengthened with the wet-layup system. Strengthening the two spans with different
systems was possible because each span was simply supported at the ends and therefore
responds independently from the adjacent span.
The required area of composites was calculated using a moment-curvature
analysis of the strengthened cross sections. The analysis was conducted using nonlinear
material properties for the concrete and reinforcing steel, and linear-elastic properties for
the composites. The concrete was modeled using the uniaxial stress-strain model
proposed by Hognestad (6) and the reinforcing steel was modeled using a bilinear curve.
Failure of the reinforced concrete section was assumed to occur when the concrete
reached a value of0.0035. However, debonding of the composites was expected to occur
prior to concrete crushing so a limiting value of composite strain equal to 0.007 was used
for design. This value represents a lower bound estimate of measured strains during tests
of strengthened beams tested in the laboratory (2). The area of composites was adjusted
until the required strength was approximately reached. A plot showing the moment-
curvature response of the strengthened pan-girders using the final configuration for the
two different composite systems is shown in Fig. 2. The horizontal dashed lines in the
figure indicate the values of the live-load moment that is developed when the strain in the
composites reach a value equal to 0.007.
The results of the design were subsequently verified using a design procedure
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
based on equilibrium considerations of the strengthened cross section (Fig. 3). Failure of
the strengthened cross section was assumed to be controlled by debonding of the
composite laminates when these reached a strain equal to 0.007. Equations 3 and 4 were
solved simultaneously to give the required areas of composites for the two composite
systems used in the bridge. These equations are only applicable when the neutral axis
depth lies within the flange of the section illustrated in Fig. 3. The section width should
be adjusted in cases where the compressive stresses extend below the flange of the
section.
where
Mn req = Required nominal capacity of the strengthened section.
As= Existing area of reinforcing steel.
/y = Specified yield stress of the reinforcing steel.
f"c = 0.9 f'c·
Cr01 = Neutral axis depth due to total loads (dead + live).
Af = Required area of CFRP composite = bp times tP.
& CFRP = Strain in the CFRP composite at debonding from the concrete surface
equal to 0.007, measured to the centroid of the composite laminates.
EP = Tensile modulus of elasticity of the CFRP composite.
d = Distance from extreme compression fiber to centroid of existing reinforcing
steel.
b= Flange width measured from the centerline of the two pans adjacent to an
individual girder.
dp = Distance from extreme compression fiber to centroid of CFRP composite.
As mentioned previously, the limiting strain in the composite laminates was
determined from results of laboratory tests of small-scale specimens. In these specimens,
transverse composite straps were used to ensure that premature failure did not preclude
reaching the design strain in the composites. Therefore, transverse straps were placed
along the entire length of the laminates in the girders at a spacing equal to half the girder
depth to avoid premature failure of the composites by debonding.
It is recognized that this design procedure is approximate mainly because a
single value of strain in the composites was used to determine the resultant force in the
composite laminates. The resultant force in the composites was assumed to be
concentrated at a distance dp from the top of the girders. In reality the strain within the
laminates varies with depth from the neutral axis so the limiting value of 0.007 was
assumed to be the strain at the centroid of the laminates when debonding occurred.
The final designs for the two composite systems used in the bridge are shown in
Fig. 4. It should be noted that both schemes involved placement of the longitudinal
composite laminates on the sides of the girder stems instead of bonding the laminates
entirely on the tension face. Most of the pan-girders contained a concrete surface
irregularity along the bottom of the stem caused by the placement of adjacent pans. To
avoid significant surface preparation, the composite laminates were placed away from
this region of the girders.
The increased live loads on the bridge also caused an increase in the shear in the
existing pan-girders. The shear strength of the existing girders was calculated in
accordance with the 1996 AASHTO Specifications for Highway Bridges (7). The shear
strength exceeded the new demands calculated with the increased live load on the bridge
so shear strengthening of the pan-girders was not necessary.
CONSTRUCTION PROCEDURE
Strengthening and widening the existing bridge was conducted in three stages.
The first stage consisted of repairing and strengthening the existing pan-girders. The
strengthening operation took place while the bridge remained in operation. The concrete
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
cracking and spalling that was detected in the existing structure prompted the engineer of
record to require that all cracks be epoxy injected and concrete spalls repaired prior to the
application of the composite laminates. This repair work is also typically recommended
by the manufacturers of the composite systems to ensure adequate bond between the
composites and the surface of the concrete. After cracks were injected, the composite
systems were applied following recommendations from the manufacturers. A picture
showing the span strengthened using the pultruded plate system during construction is
shown in Fig. 5. The ease of application of the composite plates and lack of necessity for
heavy construction equipment is evident from this picture. Underside views of the final
configuration of each span of the bridge are shown in Fig. 6.
The second and third stages consisted of increasing the width of the bridge by
connecting new pan-girders to the existing (strengthened) edge girders. To maintain
traffic on the bridge with minimal disruption, work proceeded by widening one side of
the bridge at a time. New girders and bent cap extensions were connected to the existing
structure using dowels drilled into the existing structure. The region of the bridge that
was widened after strengthening the existing girders is shown in Fig. la.
As part of this project, laboratory tests were conducted before strengthening the
bridge to validate the design of the strengthening systems. The tests were conducted on
full-scale specimens consisting of two girders having identical dimensions as the pan-
girders in the bridge. A specimen was fabricated and tested with each of the two
different strengthening systems. Specimen J-1 was strengthened using the pultruded
composite system (5) and specimen J-2 was strengthened using the wet-layup composite
system (4). The composite application techniques that were used in the laboratory
mimicked the expected construction procedure in the field. The specimens were first
positioned on concrete blocks and loading was applied to generate cracking equivalent to
service conditions. The stiffness of the unstrengthened specimens was determined from
the low amplitude test to compare with the stiffness of the specimens after strengthening.
Contrary to the construction sequence followed in the field, the cracks in the
laboratory specimens were not injected before applying the composites to the pan-girders.
This decision was made to evaluate whether the transverse straps that were placed along
the longitudinal composites would be effective in delaying debonding even in cases
where moderate cracking was present. Strengthening was conducted following the
recommendations from the manufacturers of the composite systems.
The two specimens were tested to failure to evaluate the increase in flexural
capacity and validate the design procedure of the strengthening schemes. Both
specimens exhibited very similar behavior during testing. The failure of the specimens
was characterized by debonding of the CFRP composites from the surface of the concrete
ACKNOWLEDGEMENTS
Funding for this project was received from the Texas Department of
Transportation under project 0-1776. The laboratory tests were conducted in Ferguson
Structural Engineering Laboratory at the University of Texas. The research project was
conducted under the supervision of Dr. Michael E. Kreger and Dr. Sharon L. Wood.
Their comments and suggestions throughout the project are appreciated.
REFERENCES
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
3. Brefia, S.F., Wood, S.L., and Kreger, M.E., "Increasing the Flexural Capacity of
Typical Reinforced Concrete Bridges in Texas Using Carbon Fiber Reinforced
Polymers," Research Report # 1776-2, Center for Transportation Research,
University of Texas at Austin, 2001.
8. ACI Committee 440, "Guide for the Design and Construction of Externally Bonded
FRP Systems for Strengthening Concrete Structures," ACI 440.2R-02, American
Concrete Institute- Emerging Technology Series, Farmington Hills, MI, 2002.
RF 0.588 Eq.1
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
-
N
N
...
=
CD
=·
I»
I»
Table 2 Properties of CFRP Composite Systems •
=
CL.
Composite
.. Composite Thickness Strength Modulus
Rupture
Strain
...en
CD
System Type <
CD
in. mm ksi MPa ksi GPa J.le en
Pultruded
0.047 1.19 348 2,400 22,500 155 14,000
Sika P1atet
Woven Fabrict 0.040 1.02 139 960 10,600 75 13,000
Unidirectional
MBrace 0.0065~ 0.165t 550 3,790 33,000 230 15,000
Not for Resale, 08/10/2015 01:23:41 MDT
Licensee=UNI OF NEW SOUTH WALES/9996758001
Fiber
Notes:
* Properties published by the manufacturers.
**Sika System: CarboDur S812 Pultruded Plates (5) and
SikaWrap Hex 103C Woven Fabric
MBrace System: MBrace CF 130 (4)
t The pultruded plates were used for the longitudinal composites and the woven fabric was used for the
transverse composites.
t Thickness of one ply of composite material used for design calculations. Actual thicknesses after fabrication
ranged from 0.03 to 0.06 in. (0. 75 to 1.5 mm) per ply.
Daneshlink.com
Field Applications of FRP Reinforcement: Case Studies 213
Table 3 Comparison of Test Results with Design Moments
Specimen
. Mtest Mnreq
Mdesign
Span A Span 8
~~----~~~~--~----1
1.60 m (5'- 3") 6.55 m (21'- 6") existing bridge 1.60 m (5'- 3")
widening widening
Bridge Cross Section
#16M@250mm
Side Elevation
#13M@ 305 mm
'~t~~=H #10M@381mm
178mm
Cross Section
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
300 408
250 340
if 200 272 i
:!!. ~
c 150 204 i'
•E •
& 100 136 ~
::&
50 68
I 10mm
(24")
l~B I 610mm
I~'
I I
~
< 1 - 4) 203 mm wide- 2 ply both sides 8 (3")
Master Builders
MBraceCF 130
C JC'
76 mm x 1050 mm
4~ ";m
400mm
(16") 1_ 6 II_.,--__,, j_(3" x 41.25")
125mm
't::::::l:=-!1 T 80mm
(3.125")
(5")
Master Builders
SikaWrap Hex 103C MBrace CF 130
76 mm x 1050 mm 200mm(8") 76mm
(3" X 41.25") 2 ply, both sides (3")
-Specimen J-2
-300 H-20 (New Rating) 408
¢,!
~
-c
Q.
:.;;: z.)It,
G)
200 -::::=.---y----. 272 i
H-11.8 (Original Rating) G>
E
0 ------------11------------· ~
:::E H-20 (Service Loads) :::E
100
--------,------------- 136
0 . ... . 0
0 2 4 6 8 10 12 14
Deflection (in.)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: This paper reports on the use of externally bonded fiber reinforced polymers
(FRP) laminates and Near Surface Mounted FRP bars for the flexural strengthening of a
concrete bridge. The bridge selected for this project is a three-span simply supported
reinforced concrete slab with no transverse steel reinforcement, load posted and located
on Martin Spring Outer Road in Phelps County, MO. The original construction
combined with the presence of very rigid parapets caused the formation of a wide
longitudinal crack which resulted in the slab to behave as two separate elements. In
order to clarify the behavior of the structure, load tests were performed and a finite
element method (FEM) analysis undertaken. The FRP strengthening was designed to
avoid further cracking and such that the transverse flexural capacity be higher than the
cracking moment. Both FRP techniques were easily implemented and showed
satisfactory performance.
219
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
220 Casadei et al.
Paolo Casadei, is a PhD candidate in the Department of Civil Engineering at the
University of Missouri-Rolla. He received his B.Sc. in Civil Engineerring from the
University of Bologna, Italy. His research interests include repair of reinforced concrete
structures and validation of in-situ diagnostic cyclic load testing. He is a member of ACI,
ASCE, and ICRI. He is an EIT in the United States.
Nestore Galati, is a doctoral student in Composite Materials for Civil Engineering at the
University of Leece, Italy, where he received his B.Sc. in Materials Engineering. He
holds a M.Sc. degree in Engineering Mechanics at the University of Missouri-Rolla. His
research interests include repair of masonry and reinforced concrete structures. He is an
EIT in the United States.
Renato Parretti, is an ACI member and senior structural engineer with Co-Force
America responsible for numerous FRP design projects throughout the world. Mr.
Parretti holds a B.S. in Civil Engineering from the University of Florence, Italy. He is a
registered PE in Italy, and an EIT in the United States.
Antonio Nanni, is the V & M Jones Professor of Civil Engineering at the University of
Missouri-Rolla. He is a registered PE in Italy, FL, PA, MO and OK. He is an active
member in the technical committees of ACI (Fellow), ASCE (Fellow), ASTM and TMS.
He was the founding Chairman of ACI Committee 440 - FRP Reinforcement and is the
current Chairman of ACI Committee 437 - Strength Evaluation of Existing Concrete
Structures.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
INTRODUCTION
Over 40 percent of the nation's bridges are in need of repair or replacement< 1l. Budget
constraints prohibit many states from repairing or replacing all of these bridges;
consequently, states are forced to post load restrictions on their bridges as a temporary
solution until more funds become available for repair or replacement. Advanced
composite materials made of fiber reinforced polymers (FRP) have a high potential for
providing a solution to this problem.
The bridge selected for this project is located on old Route 66, now Martin Springs Outer
Road, in Phelps County, Missouri (see Fig. 1). This bridge was commissioned in 1926
and was originally on a gravel road. In 1951, the last miles of US Route 66 through
Phelps County were concrete paved. In 1972, Route 66 was replaced by interstate 1-44.
Commissioning of I-44 led to a significant decrease in traffic along Route 66. Load
posting of this bridge (a load restriction posting of S-16 trucks over 13 tons ( 11. 79 tons in
Sf units) 15 mph (24.14 km/hr), except for single unit trucks H-20 weight limit to 19 tons
(17.24 tons in Sf units), and all other trucks weight limit 30 tons (27.21 tons in Sf units))
was approved around 1985 and had a significant impact on the local economy.
This bridge is a three-span simply supported reinforced concrete slab. The total bridge
length is 66ft (20.12 m) and the total width of the deck is 22.5 ft ( 6.86 m). Fig. 2 shows a
detailed geometry of the bridge. Based on visual and Non Destructive Testing (NDT)
Given the very good concrete condition of the bridge, the structure was an ideal candidate
for strengthening using CFRP composites(l)(J). Two different strengthening schemes
were adopted in this project for evaluation purposes, bonded carbon FRP laminates
installed by externally wet lay-up and near surface mounted (NSM) rectangular FRP bars.
The paper describes the use of both strengthening techniques.
BRIDGE ANALYSIS
Load Combinations
For the structural analysis of the bridge the ultimate values of bending moments and
shear forces are computed by multiplying their nominal values by the dead and live
factors and by the impact factor according to AASHT0(4)(l) Specifications as shown in
Eq.(l):
(1)
where Dis the dead load, Lis the live load, j3d=1.0 as per AASHT0( 4} Table 3.22.1A, and
1 (maximum 30%) is the live load impact calculated as follows:
50 50
1= = =0.34::;30% (2)
L+l25 22+125
and L=22 ft (6. 70 m) represents the span length from center to center of supports.
Prior to the design of the strengthening, the analysis of the bridge was conducted by
considering a HS20-44 truck load (which represents the design truck load as per
AASHT0(4) Section 3.7.4) having geometrical characteristics and weight properties
shown in Fig. 4. According to AASHro<4l Section 3.6.3 fractional parts of design lanes
shall not be used for roadway widths less then 20ft (6.09 m).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Slab Analysis
The dead load was computed considering the self-weight of the concrete slab plus the
permanent weight of the top layer of asphalt. The weight of parapets has been computed
according to the geometrical propt:;rties of Fig. Sc and, for simplicity, distributed
throughout the width of the slab. Table 1 presents a summary of these values.
Computations for the design lane and the design truck load have been carried out and it
has been found that the design truck load is the controlling loading condition.
For the flexural analysis, the critical loading condition corresponds to the case when the
truck has one of its rear axles at the mid-span of the member (see Fig. 6). The factored
ultimate moment demand is computed for the entire slab in Eq.(3):
M = 1.3xmnL + 1.3xl.67xl.3x~L
2
(3)
u 8 4
Mu = 1.3(5.9:)(22) + .1.3x 1.67x 1~3x (32)(22) = 493 .?k _ ft (669 kN -m)
2
(4)
For the shear analysis, the critical loading condition is when one rear axle is closer to one
support and the other is 14ft (4.27 m) away over the span (see Fig. 7). The factored
ultimate shear demand is computed for the entire slab in Eq.(S):
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The cracking moment of a unit strip has been computed (see Eq.(7)) to design a
strengthening scheme able to ensure that <I>Mn,transv. is larger or equal than the cracking
moment.
BRIDGE STRENTHENING
The strengthening design follows the previous considerations and has the purpose of
giving the bridge a moment capacity in the transversal direction equal or greater to the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
cracking moment computed in Eq.(7), in order to avoid further crack openings and
deterioration of the concrete due to water percolation through the cracks.
Two different FRP strengthening techniques have been adopted: ( 1) externally bonded
CFRP laminates installed by manual wet lay-up, and (2) Near-surface mounted CFRP
rods embedded in pre-made grooves and bonded in place with an epoxy-based paste. The
main difference between these two techniques belongs to the surface preparation
necessary before the application of the strengthening that in tum depends upon the
conditions of the concrete substrate on which the laminates and bars are bonded.
Before surface preparation for FRP application, the central crack was repaired in order to
re-establish material continuity and assure no water percolation through the crack. For
this purpose, the crack was sealed using an epoxy-paste and then injected with a very low
viscosity resin as shown in Fig .. 8a-b. Once the crack had been repaired, FRP have been
applied following the design provisions.
The design of both FRP technologies is carried out according to the principles of ACI
440.2R-Oi6> (ACI 440< 6> in the following). The properties of the FRP composite
materials used in the design are summarized in Table 3 and Table 4. The reported FRP
properties are guaranteed values.
The tjJ factors used to convert nominal values to design capacities are obtained as
specified in AASHT0(4) for the as-built and from ACI 440(6) for the strengthened
members.
ffu =CEf;u
(8)
5Ju =CE&;u
where fiu and &fu are the FRP design tensile strength and ultimate strain considering the
environmental reduction factor (CF) as given in Table 7.1 (ACI 440(6)), and/fu and &ju
represent the FRP guaranteed tensile strength and ultimate strain as reported by the
manufacturer. The FRP design modulus of elasticity is the average value as reported by
the manufacturer.
The material properties of the laminates that have been used are listed on Table 3. The
design for externally bonded laminates called for a total of six, 12 in (30.48 em) wide,
single ply CFRP strips overlapping at center span for 10ft (3.05 m). The strips were
evenly spaced over the width of 20ft (6.09 m) and ran the entire width of the slab, as
shown in Fig. 9. The moment capacity provided with this strengthening scheme is equal
to ¢Mn=16.5 k-ft (23 kN-m). The CFRP laminates were applied by a certified contractor
in accordance to manufacturer's specification<9 J (see Fig. 10).
The material properties of the NSM and epoxy paste that have been used are listed on
Table 4. The required number of near-surface mounted reinforcement was determined to
be two CRFP tapes per slot on a 9 in (22.86 em) groove spacing. The bars were
embedded in 17ft (5.18 m) long, -% in (19.05 mm) deep, and ~ in (6.35 mm) wide
grooves cut onto the soffit of the bridge deck as shown in Fig 11. The moment capacity
provided with this strengthening scheme is equal to ¢Mn=15.5 k-ft (21.01 kN-m). NSM
bars were applied by a certified contractor following the specifications<8l prescribed by
the University of Missouri- Rolla (see Fig. 12).
In order to validate the behavior of the bridge prior and after strengthening, static load
tests were performed with a H20 truck (see Fig. 13). Although H20 and HS20 trucks
differ in their geometry, the loading configuration that maximize the stresses and
deflections at mid span could still be accomplished (see Fig. 14).
Displacements in the longitudinal and transversal direction were measured using eight
Linear Variable Differential Transducers (L VDTs) and a data acquisition system under a
total of three passes, one central and two laterals. For each pass three stops were
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The results of the first load test, relatively to the stop No.3, are reported in Fig. 16. All
diagrams show the discontinuity caused by the longitudinal crack. The bridge performed
well in terms of overall deflection. In fact, the maximum deflection measured during the
load test is below the allowable deflection prescribed by AASHT0( 4 ) Section 8.9.3 (omax:S
L/800 =0.33in (8.38mm)).
A second load test was performed after the installation of the FRP materials. The
monitoring devices were placed at the same locations of the previous load test.
Test results of the second load test, as expected, show a slight improvement in the
deflection of the deck in both the longitudinal and transversal direction (see Fig. 17 and
Fig. 18, respectively).
FEM ANALYSIS
To validate the data obtained from the load tests, a linear elastic FEM analysis was
conducted. For this purpose a commercially available finite element program ANSYS
6.1 (II) was used.
The element SOLID65 was chosen to model the concrete. The SOLID65 element is a
brick element defined by eight nodes having three degrees of freedom at each node.
For this project, the material properties of concrete were assumed to be isotropic and
linear elastic, since the applied load was relatively low. The modulus of elasticity of the
concrete was based on the measured compressive strength of the cores obtained from the
slab according to the standard equation ACI 318-02( 5) Section 8.5.1:
Each element was meshed to be 3.5 in x5 in x6 in (8.9 em xJ2. 7 em xJ5.2 em). In order to
take into account the presence of the parapet and curb, an equivalent, less complex shape
was chosen. Boundary conditions were simulated as simply supported at both ends (see
Fig. 19). To take into account the presence of the longitudinal crack, the modulus of
elasticity of the central elements was reduced of a thousands times with respect to the
value expressed in Eq.(9). From in-situ inspection, the depth and width of the crack was
chosen to be equal to one element dimensions. The load was applied on 8 nodes
simulating the truck wheels; each force was equal to 4 kip (I 7.8 kN) for the H20 truck.
The experimental and analytical results for the central and right passes in the transversal
direction are reported in Fig. 20. The graph shows the good match in deflection between
the experimental and analytical results.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CONCLUSIONS
AKNOWLEDGEMENTS
Phelps County is the owner of the bridge: its commissioners and staff provided the
opportunity and helped in the implementation - Richard Pilcher PE, MoDOT's Official,
and Spencer Jones, PE of Great River Engineering, Springfield, MO, provided oversight.
The authors would like to acknowledge the financial support of the Missouri Department
of Transportation, the University Transportation Center (UTC) at the University of
Missouri-Rolla (UMR) for the research component of the project. Master Contractors
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
installed the FRP systems. Hughes Brothers and Master Builders provided the FRP
materials, as members of the NSF 1/UCRC at UMR.
REFERENCES
(I)
Stone, D.K., Tumialan, J.G., Parretti, R., and Nanni, A., (2002). "Near-Surface
Mounted FRP Reinforcement: Application of an Emerging Technology", Concrete
UK, V. 36, No. 5, pp. 42-44.
2
<l Alkhrdaji, T., Nanni, A., Chen, G., and Barker, M. (1999), "Upgrading the
Transportation Infrastructure: Solid RC Decks Strengthened with FRP," Concrete
International, American Concrete Institute, Vol. 21, No. 10, pp. 37-41.
3
<l Nanni, A., (1997), "Carbon FRP Strengthening: New Technology Becomes
Mainstream," Concrete International: Design and Construction.
(4)
AASHTO, 2002: "Standard Specifications for Highway Bridges", 17th Edition,
Published by the American Association of State Highway and Transportation
Officials, Washington D.C.
(SJ ACI 318-02, 2002: "Building Code Requirements for Structural Concrete and
Commentary (318R-02)," Published by the American Concrete Institute, Farmington
Hills, MI. .
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. I - Martin Spring Bridge
a) Side View
b) Section View
r---------------------~~~------------------~
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
/'/, /'////////////,!/'//////, /"/////////////////////////, /"///////////////////////////,
l
v......
~ t•·..o·~
I ~~---~~~0-~---------~~
)t ,
} LE5________ J§i_________~J
/'////,
"
/'////////J////////,.,; /'/////////////////////////,.,; /'//////////////////////////,
Fig. 4- Truck Load and Truck Lanes (US units; 1 kip= 4.45 kN; 1 ft =30.48 em)
*
~ 14'-0" TO 30'-0" ,r 14'-0" ---,~~ '
a) Design Truck (HS20-44) *
18.0 !dJI FOR MOMENT
26.0 kip FOR SHBAR
LS
J.--tt·~·__J l
J. 22'..0"
LS
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
~ (r--------------------~,)
Jf . 7 ·~ [i;:tJ::%t?1f:1;f ;-:•~I0\J':li:is·:i::1
l'i ,_ I .J. 12" / 24" r;;.1 24" r;:1 I
~l~ - - - - - - - - - )
1 1
2 Plies CF-130 Stri
12" wide @36" ole
Fig. 9 - Strengthening with Laminates on Span 1 and 3 (US units; 1 in= 2.54 em)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
c, _____________ , ___, 3.
--t
la ----------------------~
~
Pt =8k
Pz =16k
Fig. 14- H20 Legal Truck (US units; 1 kip= 4.45 kN; 1 ft =30.48 em)
I I 1~
!1 2 !3 !
Stops Locations of Rear Axle
j... / v
""
=
~ O.ot
.......= /
~
' f'~ Vl /
!0.012
0.014 - ' ~ v \..... L,V
/__
~
...
0.016
O.oJ8
0.02
-Left Pass
-Right Pass
-Center Pass " ~
i
......_, ~·
Fig. 16- Mid Span Deflection in the Transverse Direction, Stop No.3
(US units; 1 in= 2.54 em)
L
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
Field Applications of FRP Reinforcement: Case Studies 235
Slab Length (in.)
0 20 40 60 80 100 120 140 160 180 200 220 240 260
' ,,
0
0.002
0.004
Central Pass ,
-Longitudinal Direction (BEFORE) /) ~
0.006
"'~ ... v
"' ,,
"":' - 'Longitudinal Direction (AFTER) ~
,go.oo8
1= / /
0.01
ro...' ... ,.. ~/
'i! 0.012
l:l
0.014
' ......... ..... r-- .... f-
...........
....
_,. /
lo'
0.016
0.018
0.02
Fig. 17- Center Line Deflection in the Longitudinal Direction, Stop No.3
(US units; I in= 2.54 em)
~= 0.01
"""' ~ - ·Transverse Direction(AFTER)
"'
~
~ 0.012
0.014
0.016
~~ -- ~
O.Gl8
0.02
Fig. 18- Mid Span Deflection in the Transverse Direction, Stop No.3
(US units; 1 in= 2.54 em)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0.004
0.006
• ,· .. ..
' • ~.
...
• ' ,., '"
y
j I
--
ruck'tule
l
~
:..t
Tmck'u
forr!pt : l
,
'\
,
, ,
, ,
0.02
-- Right Pass Analytical
'"-""'""
. .
Fig. 20 - Comparison of Experimental and Analytical Results in the Transversal
Direction (US units; 1 in= 2.54 em)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: Since its inception in 1995, the ISIS Canada research network has developed
design procedures and innovative techniques for the rehabilitation and repair of existing
concrete structures using fiber reinforced polymer (FRP) materials. In co-operation with
various industrial partners, ISIS Canada has conducted many field application projects,
successfully transferring ISIS technology into practice in the field.
This paper provides a review of four recent field applications in Western Canada,
utilizing externally bonded FRP for the repair and strengthening of bridges. The projects
include flexural strengthening of a bridge deck under lateral bending, shear strengthening
ofl-shaped AASHTO girders for two bridges, and the repair and strengthening of
concrete corbels supporting a single girder pedestrian bridge. Some construction costs
and the time required to complete each project are presented, as well as practical details
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
specific to each application.
239
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
240 Hutchinson et al.
Robin Hutchinson is an Engineer-In-Residence and Senior Instructor with the Dept. of
Civil Engineering at the University of Manitoba, Canada. She received her Ph.D. from
the University of Manitoba in 1999.
Gamil Tadros, MAC/, is Technical Applications Consultant, ISIS Canada and Structural
Consultant, Speco Engineering Ltd. Dr. Tadros is an active member in Canadian codes
(or the use of FRP. His field of activity includes bridge analyses, rehabilitation, and
design, structural analysis, prestressed concrete, and FRP design. He is a member ofACI
Committee 440, Fiber Reinforced Polymer Reinforcement
Jadwiga Kroman, MAC!, is Senior Engineer with The City of Calgary, Canada.
Ms. Kroman 's field of activity includes managing, design and analyses of bridges and
parking structures. Ms. Kroman's particular interests include advancement of methods
(or concrete structure upgrading and restoration using FRP, as vehicles for optimization
of bridge structure life cycle and efficiency.
INTRODUCTION
The ISIS Canada research network has developed design procedures and standards for the
rehabilitation and repair of existing concrete structures using fiber reinforced polymer
(FRP) materials. Through various field applications across Canada, this technology has
been successfully transferred into practice. This paper provides a review of four recent
field applications in Western Canada, utilizing externally bonded FRP for the repair and
strengthening of bridges.
Flexural strengthening of the Country Hills Boulevard Bridge in Calgary, using CFRP
laminates is presented, followed by the shear strengthening of I -shaped AASHTO girders
using CFRP sheets. This shear strengthening technique is described for two projects, the
Maryland Bridge in Winnipeg, Manitoba and the John Hart Bridge in Prince George,
British Columbia. And finally, the repair of concrete bridge corbel supports using CFRP
sheets for the Jacques Lodge Pedestrian Bridge in Calgary Alberta, is discussed.
Visual monitoring of these bridges has been conducted in a regular basis. The increase in
capacity due to FRP strengthening was limited to 50% of the original capacity. The
criteria and design procedures used in these projects are consistent with those included in
ISIS Canada Design Manual No. 4 (Neale 2001 ).
This two-lane bridge, shown in Fig. 1, was constructed in 1975 to carry Secondary Road
564 over Highway 2, between Edmonton and Calgary, however it became a busy and
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The City of Calgary appointed CH2M Limited to analyze the bridge and determine
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
appropriate strengthening measures. Design and construction criteria included the need
to strengthen the bridge to carry current design live loading, and to maintain one -lane
traffic on the bridge at all times during construction.
The bridge is a two-span composite steel box-girder structure with a 44-degree skew.
Each span is 57.9m long, the overall width is 10.06m and the clear roadway is 8.53m.
There are three 1.575m wide by 1.67m deep box girders supporting a 140mm thick
reinforced concrete deck slab. The bridge was designed for AASHTO HS20-44 loading.
The strengthening criteria was to review the strength of the superstructure and bearings
for the CS-750 truck which applies 750 kN of live load according to The Canadian
Standard CAN/CSA-S6-88 (1988). A finite-element grid analysis program was used to
model bridge behavior under a moving CS-750 live load. The model was analyzed for the
effects of self-weight, superimposed dead load, and truck or lane load. The CS -750 truck
was applied at different locations across the bridge. Combined truck loading was also
considered.
Deck Slab
The deck slab is 140mrn thick with a 50mm asphalt-wearing surface. It spanned five
equal lengths of 1.575m, between girder flanges and it proved to be the main problem
area in the bridge rehabilitation work.
In the analysis, it was assumed that the 50mm of asphalt over the deck slab would be
removed and replaced with heavier overlay concrete. Over most of the slab area, analysis
showed that applied factored bending moments, Mr, were in the order of 42 kN.m/m'.
Strength was therefore not a factor for most of the slab, but crack avoidance, or
minimization, in .the negative moment region was an important issue. In two specific
areas in each span, the factored negative bending moments in the slab (53 kN.m/m')
exceeded the moment capacity provided by the existing reinforce concrete section (Fig.
2). Two alternatives were considered to strengthen the deck slab. The first alternative was
to provide slotted strips in sections of deck slab and strengthen these sections with
additional reinforcement. The other was to add epoxy bonded carbon fiber strips on the
top of the deck slab. The CFRP strips were applied in the capacity deficient tension zone
and anchored in the adjacent compression zone (Fig. 3)
The deck strengthening involved the installation of Carbon Fiber Reinforced Polymer
(CFRP) strips, lOOmm Sika Dur as shown in Fig. 3. The lOOmm wide x 1.2mm thick
The design for the number of strips to be used per meter is based on limiting the stress in
the strips at ultimate loads, to less than 25% of the ultimate capacity of the strips. The cut
off of the strips were chosen, as shown in Fig. 3, to be in the compression zone of the
deck slab to avoid peeling of the strip at its end according to Saadatmanesh and Malek
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(1998). The cost of CFRP including the installation is $160.00 per meter of the CFRP
plates.
General contractor for the rehabilitation project was Walter Construction (Canada) Ltd.,
with the CFRP application completed by Van Mason Coating Ltd. Engineers for the
project were CH2M Limited, of Calgary, and ISIS Canada was consultant for the project.
Externally bonded CFRP sheets have been used to upgrade the shear capacity of two
concrete bridges in Western Canada. The John Hart Bridge in Prince George, British
Columbia (BC), and the Maryland Bridge in Winnipeg, Manitoba both consist of
I-shaped prestressed concrete. AASHTO girders which do not meet current code
requirements for shear strength under today's heavier truck loads. In fact, the industrial
traffic Gross Vehicle Weight (GVW) has increased 3-fold since the construction of the
bridges in the early 1960's. These bridges are both located on main arterial roadways and
continue to be heavily used. In both cases, the fact that the bridges remained completely
accessible to traffic during construction was a major factor in selecting this particular
shear strengthening method.
The John Hart Bridge is owned by the BC Ministry of Transportation Central Northeast
Region, and consists of 7 simply supported 33 m spans with six 1500 mm deep
prestressed concrete girders per span. Diagonal CFRP sheets were applied over a 4 m
length at each end of all 42 girders, resulting in one of the largest strengthening
applications of its kind.
The shear capacity of the girders was increased by 15 to 20 % using a single diagonal
layer of Rep lark™ CFRP sheets. The CFRP sheets are manufactured by the Mitsubishi
Corp. of Japan, and were supplied by Specialty Construction Products of Winnipeg,
Manitoba. The specialty contractor responsible for preparing the concrete surface of the
girders and applying the CFRP sheets was Retro of Vancouver, BC. Fig. 4 (b) shows the
application of the diagonal CFRP sheets. The final component of the externally bonded
CFRP sheet system is a 2 to 4 mm thick protective coating.
Scouten and Associates of Prince George BC were consultants for the project, with Sure
Span of Vancouver BC as general contractors. The shear strengthening of all 84 girder
ends was completed within a 6 week period in September and October of 1999.
The City of Winnipeg, Manitoba, Canada has implemented a trial application of CFRP
sheets as a first step in the plan to upgrade the shear capacity of the Maryland Bridge.
The bridge consists of two five-span continuous prestressed concrete structures with 7
girders per span. Consultant for the bridge rehabilitation project is Dillon Consultants
Ltd., of Winnipeg.
In October 1999, two girder ends were strengthened by Vector Construction Ltd. using
the MBrace™ system manufactured by Master Builders Inc., and two girder ends were
strengthened by Concrete Restoration Services Ltd. using the Replark TM system
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
manufactured by Mitsubishi. The total time for construction was 1 week for each pair of
two girder ends, and the total cost of approximately $30,000 was similar for each pair of
girder ends.
Fig. 6 illustrates the tendency for the CFRP sheet to peel and straighten due to the shape
of the girder. Based on experimental results, this outward peeling force was accounted for
in the design of the strengthening scheme for both the Maryland and John Hart bridges.
For design, strain in the CFRP sheets is limited to 2 millistrain based on both bond and
scale model beam tests. Test results indicate that the shear contribution of the CFRP
sheets can be added to the contributions from the concrete and the internal steel stirrups,
in order to predict the overall shear capacity of the girder. (Hutchinson and Rizkalla,
1999)
The Jacques Lodge Pedestrian Bridge in Calgary, Alberta, consists of a precast concrete
"drop-in" girder supported by a cantilevered girder at each end. As shown in Fig. 7, the
concrete corbels supporting the main girder exhibited cracking due to unexpected friction
forces and a lack of sufficient internal diagonal steel reinforcement. The direction of the
fibers in the applied sheets was consistent with the ordination of the deficient reinforcing
steel.
A strut and tie model was used to design the CFRP strengthening scheme. The
configuration of the CFRP sheets is shown in Fig. 8 (a). Since the outer faces of the
corbel are vertical, a maximum strain of 4 millistrain was used for design, based on the
work of various researchers. (Maeda et. al. 1997; Hutchinson and Rizkalla 1999; Neale
2001) A single horizontal layer and a diagonal layer of CFRP sheets were required for
each side of the corbel, as shown in Fig. 8 (a).
In order to repair the concrete corbels, the entire main span was jacked up. After cleaning
Repair of the Jacques Lodge Bridge was completed in 6 weeks at a total cost of $20,000.
The consultant for the project was HMS Structural Engineering, and the contractor was
JAMOR Engineering Ltd. The bridge is owned by the Metro Calgary Foundation.
CONCLUSION
Externally bonded CFRP has been used to strengthen four bridge structures in Western
Canada. For all of the rehabilitation projects described in this paper, the ease of handling
of the CFRP materials resulted in reduced construction time, when compared with
conventional repair techniques. For the three heavily used highway bridges, at least one
lane of traffic remained open at all times during construction. The four projects described
in this paper demonstrate a successful transfer of ISIS technology into practice in
Western Canada.
The methods of strengthening the above bridges developed by ISIS Canada researchers
and were instrumental for the development ofiSIS Canada Design Manual No.4 (Neale
2001)
REFERENCES
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Hutchinson, R., and Rizkalla, S., "Shear Strengthening of AASHTO Bridge Girders
Using CFRP Sheets," A CI Special Publication SP- I 88: Proceedings of the Joint A CI Fall
Convention - FRP for RC 41h International Symposium, Baltimore, Maryland, October
1999, p. 945-958.
Maeda, T., Asano, Y., Sato, Y., Ueda, T., and Kakuta, Y., (1997) "A Study of Bond
Mechanism of Carbon Fiber Sheet," Proceedings of the Third International Symposium
on Non-Metallic (FRP) Reinforcement for Concrete Structures, Sapporo, Japan, October
1997, Vol. 1, pp. 279-286.
Neale, K., "Strengthening Reinforced Concrete Structures with Externally -Bonded Fiber
Reinforced Polymers," ISIS Canada Design Manual No.4, ISIS Canada, 2001.
Saadatmanesh, H., and Malek, A. M. (1998), " Design Guidelines for Flexural
Strengthening of RC Beams with FRP Plates", Journal of Composites for Construct ion,
Vol. 2, No.4, pp 158-164.
Compression Compression
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Fig. 8 (a): Application ofCFRP Sheets Fig. 8 (b): Bridge After Repair
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: The use of fiber reinforced polymer (FRP) composites to restore the original
capacity of impacted prestressed and reinforced concrete (PC and RC) girders is highly
effective. In many instances, FRP strengthening represents the only practical and cost-
efficient solution to such problems. The ease and rapid installation of the external FRP
reinforcement without traffic interruption is often the key parameter for the use of this
technology. In this paper, the flexural strengthening of one accidentally-damaged precast
PC girder of a bridge over the Gasconade River, Missouri is presented. The impact
resulted in the spalling of concrete and severing of two out of 3 8 prestressing tendons.
The location of the damage was near the mid-span of the girder. The design was based
on restoring the ultimate flexural capacity of the member. This was accomplished with
the installation of properly anchored carbon FRP laminates (3 plies) using manual lay-up.
249
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Antonio Nanni: is the V & M Jones Professor of Civil Engineering at the University of
Missouri-Rolla. He is a registered PE in Italy, FL, PA, MO and OK. He is an active
member in the technical committees of ACI (Fellow), ASCE (Fellow), ASTM and TMS.
He was the founding Chairman of ACI Committee 440 - FRP Reinforcement and is the
current Chairman of ACI Committee 437 - Strength Evaluation of Existing Concrete
Structures.
Chip Jones: is the Project Manager for Emery Sapp & Sons, Inc. Mr. Jones holds a B.S.
in Civil Engineering from the University of Missouri-Columbia. He is a registered PE in
MO. He is actively involved with the AGC of MO.
INTRODUCTION
Repair of impacted prestressed and reinforced concrete (PC and RC) structures using
traditional and emerging technologies has been the subject of several studies (1-7). Fiber
reinforced polymer (FRP) systems composed of fibers embedded in a polymeric matrix,
exhibit properties which make them suitable for their use as structural reinforcing
elements. FRP composites are characterized by excellent tensile strength in the direction
of the fibers and by negligible strength in the direction transverse to the fibers. They do
not exhibit yielding, but instead are linear-elastic up to failure and characterized by a
range of low to high modulus of elasticity in tension and low compressive properties.
FRP composites are corrosion resistant and are expected to perform better than other
construction materials in terms of weathering behavior. Several types of fibers have been
developed for use in FRP composites. For this application, carbon fibers, which are
recognized to be the stiffest and most durable, were used.
OBJECTIVES
The objective of this project was to restore the original ultimate flexural capacity of an
accidentally impact-damaged PC girder of Bridge A5657 (see Figure 1) located on Route
28 over the Gasconade River, South of Dixon, MO. Two prestressing tendons in the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Similar projects involving the strengthening of damaged bridge girders have been
conducted (8), including two in the State of Missouri (9, 10). In these projects, carbon
FRP (CFRP) laminates were proposed for installation by manual lay-up to restore the
original ultimate capacity of the impacted girder.
A commercially available FRP strengthening system was selected for its high strength
and excellent performance under sustained and cyclic loading (11 ). The system includes
primer, putty, fiber sheets, and impregnating resin (i.e., saturant) for installation by
manual lay-up.
Material properties of the FRP reinforcement reported by the manufacturer, such as the
ultimate tensile strength, do not account for long-term exposure to environmental
conditions, and were considered as initial properties. FRP properties to be used in all
design equations are given as follows according to ACI 440 guidelines (12):
ffu = CEJ;
(1)
cfu = CEc;u
where fiu and Efu are the FRP design ultimate tensile strength and strain considering the
environmental reduction factor (CF) as given in Table 2 (see Table 8.1 of reference 12),
and/1u and ciu
represent the FRP guaranteed tensile strength and strain as reported by the
manufacturer. Table 3 summarizes the FRP laminate properties calculated using the net
fiber area when an environmental reduction factor CE of 0.85, corresponding to exterior
exposure condition, is selected.
The design of the concrete cross-section of the girder was carried out according to (12).
Typically the FRP external reinforcement is dimensioned to verify the following
equation:
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(2)
The equation for the nominal moment capacity of a composite (PC plus deck) section
strengthened with FRP flexural reinforcement is given as follows:
(3)
where FFRP and FP, represent tensile forces in the FRP laminate and prestressing tendons,
respectively (see Figure 4).
The stress values in each of the materials depend on the strain distribution and the
governing failure mode. Because of the number of variables involved, there is no direct
procedure for determining the strain distribution and failure mode. Instead, a trial and
error procedure is necessary. This procedure involves first estimating the depth to the
neutral axis, e, and determining the corresponding failure mode based on this assumption.
The estimated depth to the neutral axis may be confirmed or modified based on strain
compatibility, the constitutive laws of the materials, and internal force equilibrium. The
computed factored moment capacity of either the composite section before damage,
impacted, and repaired member is summarized in Table 4.
The rehabilitation of this impact-damaged girder called for concrete repair and
application of CFRP laminates· as shown in Figure 5a). The flexural strengthening
consists ofthree 24 in (60 em) wide plies with lengths of lOft, 11ft, and 12ft(3.00, 3.35,
and 3.65 m), respectively, applied to the bottom of the girder with fibers aligned along its
longitudinal axis. The triple-ply laminate was centered over the damaged area.
Ten strips, 8 in (20 em) wide and spaced at 16 in (40 em) on centers, were then U-
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
wrapped around the bulb of the girder over the previous installation (see Figure 5b). The
purpose of the U-wrap is to prevent delamination of the FRP plies applied to the bottom
surface of the girder. Figure 6 and Figure 7 show the installation of the FRP laminates
and the U-wrap used to hold in place the flexural reinforcement, respectively.
FRP reinforcement was no longer needed at the cross section where the damaged strands
become again fully effective. For the damaged strands, it is reasonable to assume a linear
transition between the point of zero stress and full prestress transfer over their
development length. The development length, ld, is expressed as follows (see Figure 8,
and (13 and 14)):
where db is the strand diameter (in in),/ps is the stress in the prestressed reinforcement at
nominal strength of member (in ksi), and fse represents the effective stress in the
prestressed steel after losses (in ksi). In tum,J;;s andfse are given as follows (14):
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
where hu is the ultimate tensile strength of the prestressing strands, /'p is the factor for
type of prestressing strand equal to 0.28, pP is the ratio of prestress reinforcement,
fJI=0.75, m = pfy I fc, w' = p'jy I J;, p =A) bd, p' =A~ I bd, and P1
represents the final prestressing force after losses. The term in the square brackets in
Equation (5) should not be less than 0.17, as suggested by (14). Using this controlling
value, one can obtain hs=253 ksi (1744 MPa). From Equation (4), the development
length for the strand can be calculated equal to 5.5 ft (I. 70 m).
The length of the first and longest FRP ply was then set equal to I2 ft (3. 65 m), which
corresponds to 11ft (3.35 m) necessary to account for the tendons' development length
plus I ft (0.3 m), corresponding to 6 in (15 em) at each end, deemed necessary for the
FRP development length as suggested by (12). The remaining two plies were set to have
lengths of II and 10ft (3.35 and 3.0 m).
If one were to consider that the effect of the two-strand rupture be limited to the bulb
portion of the girder (conservative assumption), then the differential stress at each
extremity would be equal to:
where b I is the width of the bulb, and hI represents the height of an equivalent rectangular
section having the same area of the girder's bulb. Since the stress differential is relatively
small and localized, the effect of the unbalanced prestress force about the girder's vertical
axis of symmetry due to the strands being severed on one side of the girder may be
neglected.
Before carrying out the CFRP laminate installation, the damaged area of the girder was
restored with a rapid setting, no-shrinkage, cementitious mortar. The sequential
installation procedure was as follows:
• Surface Preparation: the bottom edges of the girder are rounded for proper
wrapping and the concrete surface is sandblasted until the aggregate is exposed
and the surface of the concrete is free of loose and unsound materials.
• Application of primer: a layer of epoxy-based primer is applied to the prepared
concrete surface using a short nap roller to penetrate the concrete pores and to
provide an improved substrate for the saturant.
• Application of putty: after the primer becomes tack-free, a thin layer of putty is
applied using a trowel to level the concrete surface and to patch small holes.
• Application offirst layer of saturant: the first layer of saturant is rolled on the
putty using a medium nap roller. The functions of the saturant are: to
impregnate the dry fibers, to maintain the fibers in their intended orientation, to
distribute stress to the fibers, and to protect the fibers from abrasion and
environmental effects.
• Application of fiber sheet: after the fiber sheet is measured and pre-cut, it is
placed on the concrete surface and gently pressed into the saturant. Prior to
removing the backing paper, a trowel is used to remove any air void. After the
backing paper is removed, a ribbed roller is rolled in the fiber direction to
facilitate impregnation by separating the fibers.
• Application of second layer of saturant: a second layer of saturant is applied and
worked into the fibers with a ribbed roller. After this, the second and third fiber
sheet can be installed by repeating the described procedure.
After fiber installation and curing, a layer of protective coat was applied to ensure UV
protection of the FRP and to warrant durability performance of the system.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CONCLUSIONS
This paper reports on the use of externally bonded FRP laminates for flexural
REFERENCES
(I) Shanafelt, G.O. and Hom, W.B, (1980): "Damage Evaluation and Repair Methods for
Prestressed Bridge Members," Transportation Research Board, Washington, DC, NCHRP
Report 226.
(2) Shanafelt, G.O. and Hom, W.B., (1985): "Guidelines for Evaluation and Repair of
Prestressed Concrete Bridge Members," NCHRP Report 280, Transportation Research
Board, Washington, DC.
(3) Olson, S.A., French, C.W., and Leon, R.T., (1992): "Reusability and Impact Damage
Repair of Twenty-Year-Old AASHTO Type III Girders," Research Report No. 93-04,
University of Minnesota, Minneapolis, MN, May.
(4) Aboutaha, R, Leon R.T., and Zureick, A.H., (1997): "Rehabilitation of Damaged
AASHTO Type II Prestressed Girder Using CFRP," Proc., II Symposium on Practical
Solutions for Bridge Strengthening and Rehabilitation, Kansas City, MO, April, pp. 293-
301.
(5) Zobel, R.S, Jirsa, J.O., Fowler, D.W., and Carrasquillo, R.L., (1997), "Evaluation and
Repair oflmpact-Damaged Prestressed Concrete Bridge Girders," Report CTR 1370-3F,
University of Texas, Austin, TX.
(6) Klaiber, F.W., Wipf, T.J., Russo, F.M., Paradis, R.R., and Mateega, R.E., (1999):
"Field! Laboratory Testing of Damaged Prestressed Concrete Girder Bridges," Iowa DOT
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Report HR-397, Iowa State Univ., Ames, Iowa, Dec., 261 pp.
(7) Dolan, C.W., Rizkalla S., and Nanni, A., Editors, (1999): "Fiber Reinforced-Polymer
Reinforcement for Concrete Structures"- Fourth International Symposium (FRPRCS4),
ACI Special Publication No. 188, American Concrete Institute, Farmington Hills, MI,
1182 pp.
(8) Nanni, A., (1997): "Carbon FRP Strengthening: New Technology Becomes
Mainstream," Concrete International: Design and Construction, Vol. 19, No. 6, June, pp.
19-23. .
(9) Nanni, A., Huang, P.C., and Tumialan, G., (2001): "Strengthening of Impact-
(IO) Schiebel, S., Parretti, R., and Nanni, A., (2001): "Repair and Strengthening of
Impacted PC Girders on Bridge A4845," Final Report RDT01-017/RI01-016, Nov. 2001,
Missouri DOT, Jefferson City, MO, 21 pp.
(12) ACI 440.2R-02, (2002): "Guide for the Design and Construction of Externally
Bonded FRP Systems for Strengthening Concrete Structures", Reported by ACI
Committee 440, American Concrete Institute, Farmington Hills.
(13) PCI Design Handbook, Precast and Prestressed Concrete, (1999), Fifth Edition,
Raths, Raths and Johson, Inc., Editor, Page 4-26. ·
(14) ACI 318-99, (1999): "Building Code Requirements for Structural Concrete (ACI
318-99) and Commentary (ACI 318R-99)", Reported by ACI Committee 318, American
Concrete Institute, Farmington Hills.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Ultimate design tensile strain ofFRP
FRP Guaranteed tensile strain
Strength reduction factor
Coefficient depending on the prestressing tendon type
Ratio on non-prestressed tension reinforcement
Ration of compression reinforcement, As /bd
Ratio of prestress reinforcement
Ginc Differential stress at the bulb extremities of the girder
(j) pfJJc
m' p'fJJc
t - - 3'-6" ------1
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
/DamagedA.rea
10
, ~JstCFRPP/y 1-----12'-------1
(a) Elevation View
----JJ'----~ ~'UICFRPP/y
f-----12' JrdCFRPP/y
(a) Elevation View
a) b)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 6 - Installation of the Flexural Reinforcement
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`--- 263
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
264 Basler et al.
Miklos Basler is the Key Account & Business Development Manager for
Structural Strengthening for Sika Services AG, Zurich, Switzerland. Miklos is a
member of ACI and the Swiss Society of Engineers and Architects, fib Task
Group 9.3 - FRP reinforcement for concrete structures and of the IABSE Working
Commission 8: Maintenance and Repair of Structures. He holds a B.S. C. E. from
the Faculty of Civil Engineering, University in Belgrade.
David White, P.E. is the Director of Marketing, Repair and Protection for Sika
Corporation, Lyndhurst, NJ, USA. He is a voting member of ACI Committee 440,
as well as Committees 503 and 548. David is also active in the International
Concrete Repair Institute Strengthening Committee and the American Segmental
Bridge Institute. He holds a B.S.C.E. from Columbia University in New York and
an M.S.C.E. from Polytechnic University in Brooklyn, N.Y.
Mario Desroches is the Area Manager for Sika Canada in Quebec, Canada. He
is the Director of the Quebec Chapter for ICRI. Mario holds a B.S.C.E. from
Montreal University, Ecole Polytechnique and has worked for Sika since 1991.
He personally has supervised over 60 FRP strengthening projects and has also
collaborated with ISIS Canada for FRP Research applications.
1. Introduction
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
In today's world, construction engineers are faced more and more frequently with
the task of strengthening existing Bridges in order to secure or even increase
their load bearing capacity. Many different strengthening techniques are
available, such as installing additional steel, external post-tensioning, bonded
reinforcement, increasing the concrete cross-section etc. The use of bonded
steel plates has been used successfully since the late 1960's. External plate
bonding is a method of strengthening which involves adhering additional
reinforcement to the external faces of a structural member.
The chemical resistance of CFRP plates against pollutants is very good. The
carbon fibers and the epoxy matrix have long-term resistance against concrete
pore water, de-icing salts and hydrous acid solutions. CFRP plates are available
in different widths between 50mm to 150mm and thickness of 1.2mm to 1.4mm.
CFRP plate system are particularly suitable for flexural strengthening, in-situ
rehabilitation as well as for pre-stressing.
The main function of the adhesive layer is to transfer the forces between the
composite material and the substrate, such that they will act monolithically.
The following properties are important for high strength structural bonding:
- High bonding strength - High cohesive strength
- Low creep - Good resistance against humidity and
alkalinity
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
There was little difference in the load-bearing capacity, mid-span deflection and
strain in the Carbo Our at debonding of test specimens with or without oscillating
loading during curing of Sikadur adhesive. Therefore, bridges can remain open to
traffic during application of CFRP plate system.
4. Case Studies
4.1 Bridges in the Republic of Macedonia
In 1999, the 1st TMCA (Transport Movement Control Agency) of the US Army in
Germany initiated the need for transportation of military equipment by means of
HETS (Heavy Equipment Transport System) military vehicles having nine axles
and a total weight under loaded conditions of 104.3 tons. The objective was to
achieve conditions for safe crossing of loaded HETS military vehicles over the
bridges on the National road M-2 between Kumanovo and the Bulgarian border.
The U.S. Army Corps of Engineers, Rock Island District under the leadership of
Mr Mark R. Hoague, P.E. and Mr Mohamed M. Moussa, PhD, P.E, carried out the
Design review.
The objective was to verify the model used by the Consultant for the bridge
analysis, to check and verify the results of the analysis and design due to loaded
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
HETS vehicle and to confirm the efficiency of the CFRP plates for bridge
strengthening. In addition, load testing was performed to confirm the
performance of the system. Test were made on bridges before and after
strengthening with the Composites, measuring stresses in re-bars, concrete,
CFRP plates as well as vertical deflection of bridges.
The test results proved that the proper behaviour of the bridges is in accordance
with the applied Design Approach and that the increased load bearing capacity
under service conditions up to the level of anticipated influences due to HETS, is
appropriate without any restriction of limitation in regular traffic flow.
Bonding CFRP plates to the soffits increased the flexural strength of the 28m
main girders. Using a special heating device, the epoxy resin was accelerated to
cure in four hour.s.
Designed for traffic load according to DIN 1072, SLW600 the construction of the
bridge was completed in 1995. To allow the crossing of the bridge for a truck with
the equipment for the Krsko Nuclear Power Plant it was necessary to increase
the flexural strength of the bridge deck by bonding on the bottom side CFRP
plates in both directions.
5. Conclusions
As a result of intensive theoretical investigations, research, and laboratory
testing, the strengthening of Bridges with advanced composites has now become
a state-of-the-art technique in the construction industry. Carbon fiber systems
are now being used more and more for bridge strengthening. Their excellent
long-term resistance, high corrosion resistance and ease of installation offer
unparalleled advantages to the bridge engineering marketplace.
6. Acknowledgement
The authors acknowledge the support of:
• Mr Kenneth C. Crawford, Chief Engineer, Support Operation Engineering
51
(SOE), 21 Theater Support Command (TSC) of US Army- Europe
• Prof. Dr. Tihomir Nikolovski, Project Manager, Working Group TMCA I
Bridges, Faculty of Civil Engineering, Skopje, Republic of Macedonia
• Mr Christoph Czaderski, EMPA, Dubendorf, Switzerland
in preparation of this paper for providing information on project i.e. test results,
project documentations, photos etc.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
7. References
[1] EMPA Test Reports 170'569e-1, 418931 E and 418931 E/1 . Sika
CarboDur Structural Strengthening Systems. Bonding of CFRP plates
under dynamic load. Static testing of pre-stressed narrow slabs post-
strengthened with CFRP strips. Dubendorf, 1998 and 2001.
[2] EMPA Test Report 402'017E/2: Sika CarboDur Structural Strengthening
Systems. Fatigue and Failure Test. Test beams B1 and B2, Dubendorf
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Table 3 Static failure test results
Test Specimen Ybr Ebr Fbr Comparison Fbr
[mm] %o [kNJ %
BO, reference beam 84 636 78
No fatigue test
T2, with no fatigue test 83 9.15 815 100
B 1/82, with fatigue test. 78 8.50 743 91
( .. ~~~1-t----t--------i
I 1~ ~~~~~~~~~~--~---+----~--4---~
oK;._...__ __ ......._ _
....
_.__.....,~ ..__....~.._ .___..~...._-4•
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Q,O :u 1:.0 1.2 1A 1.8 u
~.
Fig. 6-Bl-Debonded
CFRPplate
(-39) 7 (-46)
;f
i +4 Main girder - Strains
I
! ORI: Before strengthening
STR: After strengthening
Fig. 11-View
81-b ......
Fig. 14-Vicw
- .'!<>
II ;t-
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--l
fL- 1.0
--~-I--
>:<2>.?.6
12.00
,..----.1
4?0 I
1I
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CHAPTER4
SP-215-16
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: The objective of this paper is to present some of the real life applications of
fiber reinforced polymers (FRP) in Egypt. The authors of the paper have recently used
carbon fiber reinforced polymers (CFRP) with great success to repair several types of
structures, which include hotels, administrative buildings, and historical structures. This
paper will provide a brief summary of some of the projects repaired or strengthened
using CFRP.
279
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
280 Mahfouz and Rizk
Ibrahim Mahfouz, is Professor of Structural Engineering, Faculty of Engineering,
Shoubra, Zagazig University, Cairo, Egypt. He is·a registered consulting PE in Egypt,
and Canada. He has designed numerous buildings and structures in the Middle East, and
Canada. He is a member of ACI Committee 440, and the Director of the Egyptian
Standing Code Committee on FRP. He is Chairman of the Egyptian Committee on Repair
and Strengthening of Structures using FRP.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
for the strengthening of structures using externally bonded FRP. In Egypt, the use of
FRP in the construction industry has received special attention during the last decade.
This is demonstrated by the extensive research activities on FRP that were carried out by
researchers at universities and research centers in Egypt. In Addition, the supreme
Council of Egyptian Universities and the United States Agency for International
Development in collaboration with the Egyptian Contractor's Federation have funded
the establishment of the first specialized FRP research center in Egypt. As a result, the
use of FRP to repair and strengthen structures is now well accepted by the construction
industry in Egypt, which necessitated the need to develop the Egyptian Code of Practice
for the Reinforcing and Strengthening of Structures using FRP. In April 2002, the
Egyptian Minister of Housing and Urban Development established the FRP Standing
Code Committee to develop the Egyptian FRP Code of Practice. This allowed the authors
and a number of other engineers in Egypt to use FRP to strengthen structures. The
discussion in this work will focus on three structures, namely 1) Repair of a beach resort;
2) Strengthening of an office building; 3) Repair of two historical brick domes.
PROPERTIES OF CFRP
CFRP laminates and sheets were used in all the projects presented in this paper. The
ultimate strength, modulus, and ultimate strain of the CFRP sheets were 4275 MPa, 228
kN/mm2, and 0.19%, respectively. An epoxy resin was used for all the applications. Pre-
installation quality control tension tests were carried out on the CFRP sheets. Tension
tests were also performed during the installation process. All tests were carried out in
accordance with the procedures outlined in the ACI Committee 440K document< 1J. The
quality control tests are an essential part of the repair process. In fact, the test results
obtained show a significant difference between the strength of the CFRP sheets (fiber
with epoxy) and that of the fiber based on the fiber cross section only. The average
strength obtained from the test results expressed in terms of the ultimate load per unit
width of the composite laminate was 3200 N/cm, while that of the supplier data sheet was
Nine two story single bay transfer frames in a five-story beach resort were strengthened
using CFRP. The resort is one of the largest and busiest resorts in Egypt. Each frame acts
as a transfer frame to resist the loads from the discontinuous columns of the three stories
above it (see Figure 1).
A typical frame has a span of 40ft (12m), and the beams were 47 in (1.2 m) deep. (see
Figure 2). The frames lost 20% to 30% of their flexural reinforcement and 10% to 15%
of their shear reinforcement because of corrosion of the reinforcement. It was required to
complete the project in three weeks before the beginning of the summer season. As a
result of the time constraint, the use of CFRP was the best repair alternative if not the
only technique that can be used. The repair of frames was carried out in accordance with
the provisions of ACI 440F (2). The repair work consisted of four 3.9 in (10 em) wide
CFRP laminates attached to the bottom fiber of the girder along with two additional 3.9
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
in (10 em) wide CFRP laminates attached one on each side of the girder, and 3.9 in (10
em) wide U-shaped CFRP sheets @ 9.8 in (25 em) were used in order to enhance the
shear capacity of the beams and to prevent the delamination of the bottom layers (see
Figure 3). Two additional 2 in (5 em) wide CFRP sheets one on each side of the girder
below the flange were also used to prevent the delamination of the CFRP used for shear
reinforcement (see Figure 3). Figure 4 shows a typical frame during repair using CFRP.
Since all the repair work in this project is classified under the bond critical application
case, works dealing with removal of all rust of the steel reinforcement along with the
preparation of the concrete substrate and the treatment of concrete surfaces were carried
out prior to the installation of the CFRP laminates and sheets. Pull-off tests were
performed on the concrete substrate after removing the rust.
A bank (Faisal Bank, one of the largest banks in the Middle East) bought a high-rise
building in down town Cairo. Because of the change in the function of the building, one
hundred and twenty columns required upgrading. Figure 5 shows some of the columns
before being upgrade using CFRP. It was required to increase the axial load carrying
capacity of the columns by 25%. The columns were circular and had diameters that
varied from 12 in (30 em) to 30 in (75 em) and a height of 10ft (3 m). The use of steel
plates or concrete jackets was not an option because the bank needed all the clear space
between the columns. Also, the building was very expensive so the bank was unwilling to
sacrifice any space. In addition, using concrete jackets or steel plates in the basement,
would have made the parking of cars in impossible. Using steel or concrete jackets to
increase the load carrying capacity of the columns would have resulted in increasing the
stiffness of the columns a matter that is considered undesirable in the case of an
earthquake. Considering the nature of the building, it can be seen that the time for the
The design was carried out in accordance with the requirements of the ACI 440F< 2J
document. Since it was required to increase the column's strength and ductility without
affecting its stiffness, all the fibers were oriented in the direction perpendicular to the
longitudinal axis of the column. Depending on the strengthening demands and on the
dimensions of the columns, one or more layers of CFRP were used for upgrading. Some
columns required continuous wrapping, while discontinuous wrapping was sufficient for
other columns (see Figure 6).
Several brick domes in a number of ancient and historical structures (800 years old) were
repaired using CFRP. The domes were on the verge of collapsing. The use of CFRP
guaranteed the completion of the repair work in a very short time otherwise the domes
would have collapsed. Two cases are presented in this work, namely, EL-Aini Domes and
Sodoun Dome. In both cases, bricks suffered from severe deterioration; and the domes
were seriously cracked. Some cracks were in the body of the domes only, which made the
domes behave like a set of isolated arches. Other cracks extended to the supporting ring
beams and resulted in a probable loss of equilibrium that if left without repair would have
definitely led to a total collapse of the historic structures. The use of CFRP was
recommended for the repair work since CFRP possess the significant advantage of not
causing any alteration to the outside appearance of the structure after the repair work is
completed because the CFRP sheets used were very thin. In addition, the repair work
using CFRP does not result in a change in the color and the appearance of the structure.
In this regard, a special powder made of the same type of the brick was used as a topcoat
on top of the CFRP. Maintaining the outside appearance is invaluable when dealing with
the repair of historic structures. Another consideration in choosing CFRP sheets as the
repair method for the domes was that CFRP sheets required minimum equipment for
installation, which made them suitable for this application because accessibility to the site
was very restricted. Finally, the cost of using CFRP was found to be very competitive
when compared to traditional material for this particular case.
As a result of the cracks in the body of the dome, the structural system of the domes was
changed to that of a set of isolated arches. These types of cracks are not very serious, and
the repair of such cracks is only important to maintain the integrity ,between the various
structural components. The repair works will enable the structure to act as a dome when
it is subjected to additional loads or to a straining action in the future, but it will continue
to act as a set of arches under the current loads. On the other hand, the cracks that
occurred in the supporting ring beams are serious and may cause structural instability and
probable collapse. As such, repairing these types of cracks was essential and required
immediate attention.
The repair works for the cracks can be categorized as either contact critical or bond
critical, depending on whether the CFRP is installed on the outer or inner surface of the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Historic spherical brick domes having diameters of 39 ft (12 m) resting over cylindrical
brick walls suffered from sever multiple penetrating cracks that occurred in the body of
the domes (see Figure 7). The domes in the cracked condition behaved like a set of
isolated arches jointed at the top portion of the dome. In order to maintain the structural
stability, the supporting cylindrical portion of the structure was confined using two layers
of 7.8 in (20 em) wide CFRP sheets spaced at 20 in (50 em). Due to the severity and the
large number of the cracks in the body of the domes, it was decided to use additional
confining at various levels of the dome (see Figure 8). Although, all the repair work for
the El-Aini domes fell under the contact critical category, special treatment of the surface
of the dome using fiber cement mortar was performed at the locations of CFRP due to the
poor condition of the dome surface. Brick powder was used as a topcoat on top of the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CFRP, resulting in no alteration to the shape, color, and the configuration of the historic
structure.
A historic brick dome having an irregular external configuration was supported by a brick
cylindrical wall resting on the top of a number of brick posts (see Figures 9 & 10)
suffered from sever multiple penetrating cracks. The cracks occurred in the body of the
dome and in the supporting wall resulting in structural instability. In order to maintain
structural stability, the supporting cylindrical portion of the structure was confined by
installing two layers of 7.8 in (20 em) wide CFRP wrapping sheets having a spacing of
20 in. (50 em.). Due to the irregular configuration of the outer surface of the dome, the
stitching repair method using both U shaped CFRP sheets and CFRP laminated inserts
was used.
!-Recommended Test Methods For FRP Rods And Sheets, ACI Committee 440 K.
2- Guide For The Design and Construction of Externally Bonded FRP Systems For
Strengthening Concrete Structures, ACI Committee 440 F.
i
4M
RC GIRDER 12DX45 em
~
i RC GIRDER 80X30 em
T ~-------12M-------+
--
Figure 1: Single-Bay-Two-Story Frame repaired
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
I2-5CMWIDE
I CFRPSHEET I 10 CM WIDE-U SHAPE i
1 CFRP SHEET @ 25 CM j
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
by R. J. Watson
Synopsis: It is the intent of this paper to update developments in the fascinating field of
composites. Special topics will include, masonry walls, blast effect technology, and
bridge column and bent rehabilitation and arch bridge restoration.
"Technology is a queer thing, it brings you great gifts with one on hand and stabs you
in the back with the other" (C.P. Snow).
It remains for us to keep abreast of the latest developments in composite technology to
stimulate our thinking and keep us ahead.
291
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
INTRODUCTION
Under the URM Building Law of California, passed in 1986, approximately 25,500 URM
buildings were inventoried throughout the state. Even though, this number is a relatively
small percentage of the building inventory in California, it includes many cultural icons
and historical resources. The building evaluation showed that 96 percent of the buildings
needed to be retrofitted, which would result in approximately $4 billion in retrofit
expenditures. To date, it has been estimated that only half of the owners have taken
remedial actions, which may attribute to high retrofitting costs. Thereby, the
development of effective and affordable retrofitting techniques for masonry elements in
an urgent need.
For the retrofitting of the civil infrastructure, externally bonded Fiber Reinforced
Polymer (FRP) laminates have been successfully used to increase the flexural and/or the
shear capacity of reinforced concrete (RC) and masonry members. An alternative to the
use of FRP laminates is the use of near surface mounted (NSM) FRP bars or tapes. This
technique consists of placing a bar in a grove cut into the surface of the member being
strengthened. The FRP bar can be embedded in an epoxy-based or cementitious-based
paste, which transfers stresses between the substrate and the FRP bar. The successful use
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The use of NSM FRP bars for increasing the flexural and shear strength of deficient
masonry walls, in certain cases, can be more convenient than using FRP laminates due to
anchoring requirements or aesthetics requirements. Applications of NSM FRP bars does
not require any surface preparation work and requires minimal installation time compared
to FRP laminates.
Another advantage is the feasibility of anchoring these bars into members adjacent to the
one being strengthened. For instance, in the case of the strengthening of a masonry wall
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
with FRP bars, they can be easily anchored to columns and beams.
The paper presents two applications of FRP bars for the strengthening of URM walls. In
the first application, NSM FRP bars are used as flexural reinforcement to strengthen
URM walls to resist out-of-plane forces. In the second application, a retrofitting
technique denominated FRP Structural Repainting is described. In this technique the
FRP bars are placed into the horizontal masonry joints to act as shear reinforcement to
resist in-plane loads.
In both applications glass FRP (GFRP) bars were used to increase either the flexural or
shear capacity. The GFRP bars are deformed by a helical wrap with a sand coating to
improve the bond between the bar and the embedding paste. The bars are produced using
a variation of the pultrusion process using 100 percent vinylester resin and GRFP fibers.
Typically fiber content is 75 percent by weight. The bars are commercially available in
high volumes with stocking locations in several points throughout North America and
Europe.
Flexural Strengthening
FRP bars can be used as a strengthening material to increase the flexural capacity of
URM walls. The successful use of NSM bars for improving the flexural capacity of RC
members led to extending their potential use for the strengthening of URM walls. The
use of NSM FRP bars is attractive since their application does not require any surface
preparation work and requires minimal installation time.
Strengthening Procedure
The NSM technique consists of the installation of FRP reinforcing bars in slots grooved
in the masonry surface. An advantageous aspect of this method is that it does not require
sandblasting and puttying. The strengthening procedure can be summarized as: (1)
grooving of slots having a width of approximately one half times the bar diameter and
cleaning of surface, (2) application of embedding paste (epoxy-based or cementitious-
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
FRP structural repainting offers advantages compared to the use of FRP laminates. The
method itself is simpler since the surface preparation is reduced, sandblasting and
puttying is not required. In addition, the aesthetics of masonry can be preserved.
To ensure proper bond between the epoxy-based paste and masonry, dust must be
removed from the grooves by means of an air blower prior to filling the bed joints. A
masking tape or another suitable adhesive tape can be used to avoid staining. Stack bond
masonry allows installation of FRP bars in the vertical joints, if required. In this case the
faceshell thickness of the masonry units does not limit the groove depth, therefore the
groove can be deeper.
Final Remarks
• Masonry walls strengthened with NSM FRP bars exhibited similar performance
to walls strengthened with FRP laminates.
• For flexural strengthening, increments ranging between 4 and 8 times of the
original masonry capacity may be achieved.
In the year 2000, the Defense Department had begun a 15-year renovation program to
strengthen the Pentagon in the event of a terrorist attack. The outer E-ring of the
Pentagon was almost completely abandoned for renovations when an airplane struck the
building on September 11, 2001.
The renovation program included stiffening the load bearing walls, and columns with
Kevlar like fibers and spray applied poly-urea urethane similar to what is used on the
deck of pickup trucks today. The Fyfe Company already had developed a Kevlar like
product on the market, and quickly developed a poly-urea urethane product.
The cost of repairing the two-million square feet of damage to the Pentagon will total
approximately $700 million. The E-ring, the Pentagon's outer ring, at the point of
The Pentagon renovation design-build team includes the general contractor, Hensel-
Phelps, the architect, Weidlinger Associates and a number of experts in the military and
industry.
This paper will demonstrate the type of repairs to structures using composites and
pictorially describe some 11 projects where they have used successfully and in a cost
effective manner since they compete with conventional repairs.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Highway 401 -Ontario, Canada
Highway 401 is a principal east/west expressway that runs through Toronto, Ontario. It
is one of Canada's busiest highways and the substructural members of its road crossing is
shown in Figure 1. The Ministry of Transportation wanted a fast method of repair and
chose composites with GRFP. All deteriorated concrete was removed, FRP bars cleaned,
the surfaces restored and the finish composite coat which includes a final coat of ultra
violet protection is seen in Figure 2.
Figure 3 shows a typical deterioration to a column on the Long Island Railroad at the
Rockville Station. The problem was insufficient cover over the reinforcing bars. They
are 14 foot octagonal columns. Thirty-five columns were done initially, followed by
fifty. The rehabilitation procedure was to remove loose concrete back to good solid
matrix concrete then the reinforcing cleaned and the original contour restored. Figure 4
illustrates the completed work with GRFP after which a coating to resemble the original
concrete was applied. A silicone sealant was used at the base of the columns to seal off
the entry of free water.
Figure 5 shows a new post tensioned bridge on I-84 in Connecticut known as Bigfoot
Bridge. In its early life, it sustained severe stress cracking of the columns due to an
undetermined source and arrangements were made to repair them with CFRP due to the
high forces at work. Since it was necessary to do this work quickly, working in cold
weather, workmen performed inside of a polyethylene. enclosure (Figure 6). Figure 7
Here again, we see workman wrapping GRFP to columns on Interstate Route I-70 in
Dayton, Ohio (Figure 11). The completed columns are described in Figure 12. This was
accomplished as a means to prevent further section loss and strength of the columns.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Dupage County - Illinois DOT
The Florida Turnpike bridges have experienced deterioration of the bridge columns
presumably due to chloride attack as is seen in Figure 15. Figure 16 illustrates the
bottom portion of the column going down to the footing which has been wrapped with
GRFP. Workman are installing a special treatment to protect against the entry of
brackish water.
Summit County, Ohio desired to repair a badly cracked and deteriorated concrete arch
due to corrosion and age. It was desired to repair the concrete and strengthen the arch rib
member which suffered from structural cracking and exposure of reinforcing bar.
Typical condition of the spandrel concrete shows cracking and section loss as seen in
Figure 17.
A team of specialists in composite knowledge came in and completely restored the arch
and spandrel members at a fraction of the cost and time that would have been required to
use full depth concrete repairs. The CFRP that was used increased the shear/flexural
strength and ductility of the structural members. Ultra Violet protection was included in
the finish coat which was also designed to protect the concrete from the future ingress of
chlorides (Figure 18).
Figure 20 depicts the finished arch as new as the date it was originally built a half century
ago.
Big Tujunda
Big Tujunda is a half century old bridge in the State of California that had reached an
advanced degree of corrosive deterioration in the arch as well as the spandrel members.
For reasoning of restoring the structure to its original beauty and cross sections, the
California Department of Transportation decided to go ahead with complete rehabilitation
using composites.
Figure 21 describes the method of scaffolding that was used to complete the work while
Figure 22 illustrates the complete structure good as new in every respect. Los Angeles
Public Works Association set the terms of the contract. The entire cost of labor and
materials was $225,000 which was considerably less than conventional methodology.
Somewate Bridge
CONCLUSION
Since the inception of FRP for use to accomplish seismically retrofitting on bridge
columns in 1989, composites have now been successfully used in a wide variety of
bridges and structures. FRP now have the benefit of building codes which describe their
use and this has given engineers the comfort they need to specify them in their projects.
Figure 3 - Insufficient cover to the bars Figure 4 - Column has been restored with
resulting in sever deterioration of the composites. An ultra violet coating was
concrete, Long Island Railroad later applied, Long Island Railroad
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
splitting of the columns prepare the columns inside of
polyethelyne tents due to cold weather
Figure 7 - 1-84, Bigfoot Bridge, Connecticut- Carbon fiber composites were used
because of a need for great strength
Figure 16 - Florida Turnpike Bridge column Figure 17 - Akron, Ohio State Road -
has been repaired down to the footing typical condition of arch component
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: The Gentilly-1 nuclear power plant, in Quebec, Canada, was decommissioned
in 1978. Since that time, the containment structure has been used for the storage of the
moderately contaminated nuclear reactor. The enforcement of more rigorous
. environmental regulations, as well as economic considerations, have raised the
decommissioning period from 40 to 100 years, thus severely increasing the durability
requirements for the structure.
The containment structure, constructed of thick prestressed concrete, was in good
condition except for the secondary concrete. The latter is a keystone for the durability of the
structure because it fills the recesses and protects the terminations of the tendons against
corrosion. The differential shrinkage caused cracking and de bonding and, with freeze-thaw
cycling over the years, the secondary concrete had to be removed and replaced. The ring-
beam, at the top of the containment structure, was severely affected because the numerous
tendons of the roof terminate at that level.
The retrofit of the ring-beam consisted of replacing the secondary concrete with high-
quality shrinkage-compensated mortar and concrete, followed by FRP wrapping. The layout
of the FRP wrap was designed to mitigate the adverse effects of the new secondary concrete
shrinking-induced cracks. Most ofthe concrete cold joints were covered by the FRP wrap,
which was anchored on the dome roof to provide an effective support.
303
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
304 Demers et al.
Marc Demers, Eng., M.A.Sc. Research Associate, Department of Civil Engineering,
University of Sherbrooke, Sherbrooke, Quebec, Canada.
Gamil Tadros, P.Eng., Ph.D. President, SPECO Engineering Ltd., Calgary, Alberta,
Canada.
I. HISTORY
The Gentilly-1 nuclear power plant, located in the province of Quebec in Canada, is a
CANDU prototype system that was developed by Atomic Energy of Canada Limited
(AECL) to be delivered to the Hydro-Quebec power utility. It was designed in the late
1960s and its construction was completed in 1974. The operating tests of the reactor were
never deemed satisfactory for safe operation, and AECL decided to decommission it in
1978. The nuclear fuel was removed as well as all highly contaminated parts. Because the
reactor vessel was too large to be removed, it was decided to keep the containment
structure as a storage building for the decommissioning period of time. For technical and
economical considerations, it was concluded in 1984 that returning the site to a condition
of completely unrestricted access and usage was not immediately necessary. Also, as a
result of more rigorous environmental regulations, the decommissioning period was
increased from 40 to 100 years.
Despite the fact that Gentilly-1 is decommissioned, from the AECL and public points of
view the facility must nevertheless appear to be in perfect condition. Thus, there is a
stringent requirement on the overall appearance and long-term durability of the
containment structure.
The cause of the deterioration problem was the shrinking of the secondary concrete
placed after the cure of the structural concrete, leading to cracking and debonding. The
effects of this initial cracking were exacerbated by the poor quality of the installed
materials, detailing of the reinforcing steel, and the freeze-thaw cycling. The deterioration
of the secondary concrete of the ring-beam was rather rapid, and repairs were first done
in 1985. In 1993 AECL was obliged to repair the ring-beam again; the secondary
concrete was so deteriorated in some areas that it was necessary to remove it to prevent it
from falling down. As shown in Figure l, the appearance of the structure was very poor.
In 1996, the planned solution to cover the ring-beam with steel cladding was abandoned.
Due to the poor results of the past, the high costs of the repairs, and the lengthening of the
decommissioning period, AECL decided to conduct an extensive study on the integrity of
the containment structure to seek out the best solution.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The first step of the investigation of the containment structure consisted of a visual
inspection. The AECL inspection staff found that the south side of the ring-beam, which
has the greatest exposure to sunlight, suffered the most damage with extensive
deterioration of the secondary concrete. Questions concerning the safety of the structure
were raised since indications of corrosion and alkali-aggregate reactions were observed.
In 1998, AECL conducted a detailed assessment of the containment structure (I) and
other problems were also observed, such as:
local degradation of the structural concrete, which was thought to be accelerated
through contact with the poor quality secondary concrete;
the presence of contaminants in the structural concrete such as dust, and chips and
pieces of wood;
vertical cracks in the buttresses concentrated in the secondary concrete; and
corrosion of the horizontal post tensioning ducts and tendons.
However, this investigation confirmed that the overall structure was safe and that the
structural concrete, including the ring-beam, was in a satisfactory condition. The stresses
in the concrete and the prestressing cables were measured by various tests and found to
be in agreement with the expected levels. The mechanical, physical and chemical
properties of the concrete were tested, leading to a confirmation of the presence of alkali-
aggregate reactions that could exert stress levels comparable to those associated with the
prestressing cable network if the moisture of the concrete exceeds 80%. Other properties
of the structural concrete were found to be satisfactory. Finally, the corrosion observed
on the post tensioning cables was attributed to the construction method employed; it was
concluded that no further corrosion took place since the completion of the initial
construction.
Being assured of the good structural condition of the containment structure, AECL
decided to seek out a long-term solution to the problem of the secondary concrete of the
To assure the long-term durability of the ring-beam repair, the engineering team focussed
on three complementary aspects: first, the use of state-of-the-art concrete repair
technologies for the replacement of the secondary concrete; secondly, the use of FRP
materials to secure and protect the repairs; and finally, the implementation of a rigorous
quality control program for every aspect of the repair work. The in-field application of
the stringent contractual requirements has resulted in a high quality, durable repair of the
ring beam.
The experience of AECL and ISIS Canada was called upon in 1999 to prepare the
Technical Specifications document (3) for the repair work to be performed on the ring-
beam. The experts from ISIS Canada prepared the first draft, where innovative FRP
repair technologies were proposed. The engineers of AECL subsequently reviewed the
document to modify it to their standards, particularly with regard to the aspects of
rigorous quality control required by the nuclear industry. Finally, both groups discussed
minor changes and reached a consensus on the Technical Specifications document.
The long-term durability of the repair was a critical objective since the required service
life of the containment structure is being extended by 75 years. The major concerns about
the durability of the repair of the ring-beam are related to the freeze-thaw cycling. Any
crack filled with water will eventually damage the concrete. For that reason the shrinkage
of the repair concrete mix must be minimized, and the surface water should be eliminated
quickly.
The first step of the repair was the removal of the non-adherent or unsound concrete
(Fig. 2). The defective areas were determined using the conventional sounding method.
The perimeter of the repair area was saw-cut to obtain clean joints, and the depth was
reduced when reinforcing steel was encountered. For demolition, the rating of the
jackhammer was limited (15 pounds) to reduce micro-cracking of the sound concrete. In
many locations, the depth of unsound concrete was more than 400 mm and AECL
decided not to go deeper based on the fact that the repair concrete would be steel
reinforced and provide adequate protection. The demolition was completed with a sand
blast preparation to open up the concrete pores.
Two different procedures for the replacement of the damaged concrete of the ring-beam
were used to take into account the depth of the repair. With the first procedure, for repairs
of depths less than 50 mm, non-sag, shrinkage-compensated mortar was applied by hand
saturated, surface-dry concrete, and water cured for seven days. Curing compounds were
forbidden because they can reduce the adhesion of the FRP wrap and must therefore be
removed. The specified mechanical properties of the mortar and the concrete in the two
procedures were identical and as follows:
compressive strength at 7 days: 30 MPa;
direct tensile bond strength: 1.0 MPa;
no bleeding or segregation; etc.
In addition to the high quality concrete repair, the secondary concrete was wrapped with
surface-bonded glass FRP material to improve the durability of the ring-beam repair. The
FRP wrap is not intended to prevent the formation of shrinkage cracks since dry concrete
was mandatory for the FRP installation, and cracks appeared immediately at the end of
the water cure of the fresh concrete (Fig. 4). Nevertheless, the wrap is able to mitigate the
adverse effects of cracks because of the following characteristics and properties of the
FRP:
On installation, the primer component of the FRP system is liquid; it seals the cracks,
and the cured FRP material is waterproof. Accordingly, rainwater cannot access the
cracks behind the FRP wrap, reducing considerably the adverse effects of freeze-
thaw cycling.
The FRP material is non-corrosive, allowing it to be bonded on the surface as a
structural skin reinforcement to reduce the widths of possible additional cracks.
The FRP material exhibits a linear-elastic behaviour until failure. Thus, any relative
movements of the cracks tips are forced to return to their initial positions.
The high strength of the FRP material enables it to sustain the weight of any
delaminated concrete, in the unlikely event that this should occur.
The application of the FRP wrap on the ring-beam required a standard concrete surface
preparation standard for this type of material; that is, a profile correction, the rounding of
all comers to a 30-mm radius, a sand-blasting of the concrete, and dry conditions. The
installation of the FRP wrap followed the hand-applied wet lay-up procedure
recommended by the manufacturer. Basically, the application steps are as follows:
apply the primer, a low viscosity epoxy that penetrates the concrete pore structure to
maximize the bond;
level the surface with putty, an epoxy paste (Fig. 5);
install the vertical strips of the FRP wrap according to the following procedure:
• cover the surface of the FRP strip with the first layer of resin, a low sag epoxy
that impregnates the fibers;
• install the glass fiber fabric on the fresh resin, work the material to ensure the
penetration of the resin between the fibers (Fig. 6);
The configuration for the FRP wrap is an open grid, as shown in Figure 8, to allow
moisture exchange between the concrete and the atmosphere. Drainage of the massive
concrete is assured by the grid openings at the bottom of the ring-beam. The arrangement
of the grid is designed to cover most of the joints between the structural and secondary
concrete. To ensure that the vertical strips are structurally capable of sustaining the
weight of any possible delaminated concrete, they were anchored by providing the full
development length on the dome roof. The required performance-based properties of the
glass FRP were specified as follows:
minimum tensile strength in the fiber direction 500 MPa
minimum elastic modulus in the fiber direction 25 GPa
minimum ultimate tensile strain 2%
minimum thickness lmm
The achievement of long-term durability requires high quality materials along with a
proper installation. A quality control program was therefore mandatory for this project.
Accordingly, the Technical Specifications document included many inspection and test
requirements. The quality control program began with the selection of the contractor and
materials, and continues with the monitoring of the various sensors installed during the
repairs.
The contractor that was selected was required to demonstrate his experience in the repair
of similar large-scale concrete structures, as well as his knowledge in the rehabilitation of
civil engineering structures using FRP materials. Prior to any work on the site, the
demolition, the concreting, and the FRP installation, the contractor had to produce a
detailed procedure for the quality assurance plan. Also, the FRP materials selected were
limited to systems manufactured specifically for the strengthening of concrete structures.
Quality control for the concrete repair began with the preparation of qualification patches
built and tested to demonstrate the properties of the materials and the efficiency of the
procedures (Fig. 9). Various concrete mixes were applied, cured, and tested on the wall of
the containment structure at ground level. On the actual repair on the ring-beam, the
quality was controlled by sounding inspection to detect possible voids and delaminations,
and random pull-out tests to measure the adhesion to the structural concrete (Fig. 10).
Although the concrete strength and adhesion were higher than the specified values, it was
impossible to completely eliminate shrinkage-induced cracking (Fig. 4).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The long-term quality control is assured by monitoring the various strain and temperature
sensors both embedded in the repair concrete (Fig. 11), and bonded on the FRP surface
(Fig. 12). This project was used to compare field measurements from proven
technologies, such as vibrating wires and thermocouples, to those obtained with
innovative fiber optic sensors (FOSs). On the Gentilly-1 ring-beam, two types of
temperature FOSs (long gauge and Brillouin scattering) and six types of strain FOSs
(Bragg grating, long gauge, Fabry-Perot, Brillouin scattering, "smart patches", and low
coherence interferometry) were installed. The strains and temperatures are to be
monitored on a regular basis.
3. CONCLUSION
The secondary concrete of the ring-beam of the Gentilly-1 nuclear containment structure
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
was severely deteriorated. Although the strength of the thick-walled prestressed concrete
structure was excellent, repairs were needed for durability and aesthetic considerations.
The long-term durability of the repair was the primary objective since the containment
structure will store the moderately contaminated reactor for the next 50 to 80 years.
The long-term durability objective of the ring-beam repair was attained by using state-of-
the-art technologies for the replacement of the secondary concrete, together with the use
of FRP materials for securing and protecting the concrete repairs. A rigorous quality
control was implemented for every aspect of the repair work. The innovative application
of FRP wraps to mitigate the adverse effects of shrinkage cracks was successfully carried
out.
Close monitoring of the structure using various sensor technologies will provide
information on the behaviour of the repairs, thus allowing for prompt maintenance
interventions if required.
ACKNOWLEDGEMENTS
The support of Atomic Energy of Canada Ltd. and ISIS Canada is gratefully
acknowledged. The collaboration of Vector Construction Ltd., Gestion Pachar Inc.,
SOESCA Inc., Sika Canada Inc. and Master Builders Technologies Ltd. is greatly
appreciated.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 12-Thermocouple and fiber optic gauges bonded on the FRP surface
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Innovative Strengthening of
Antwerp Zoo's Reptile Building
Synopsis: The use of externally bonded FRP (Fibre Reinforced Polymers) reinforcement
for strengthening or rehabilitation purposes is becoming a well documented and often
applied technique world-wide. A particular application of this technique in Belgium, is a
rehabilitation project at the Antwerp Zoo, where an innovative strengthening system
based on multidirectional carbon fibre reinforced polymer (CFRP) composites was
applied. This system, offering new possibilities for strengthening of structures, consists
of PC CarboComp Plus laminates and sandwich prefab composite beams. A two-way
slab and supporting beams were successfully repaired and strengthened in only 38 days,
with a minimum of disruption (animals were not relocated and the building remained
open for public) and respecting the restraints typical for a protected historical building. In
a second phase of the project, masonry columns at the basement level were wrapped with
aramid FRP (AFRP) and the supported beams were strengthened with CFRP sandwich
prefab elements. In order to achieve sufficient efficiency of the wrapping, transverse
links were provided through the rectangular columns. Related to the first part of this
project, experimental verification of the system was performed.
317
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
-------------
318 Matthys et al.
Stijn Matthys, is a post-doctoral research assistant at the Magnel Laboratory for
Concrete Research. He holds a PhD and M.Sc. in Civil Engineering of Ghent University.
His main research interests concern design recommendations for FRP reinforcement and
structural health monitoring with fibre optic sensors. He is secretary of jib (International
Federation for Structural Concrete) Task Group 9.3 'FRP Reinforcement for Concrete
Structures'.
INTRODUCTION
As one of the most challenging applications in Belgium, this paper will discuss the
strengthening and rehabilitation of the Reptile building in the Zoo of Antwerp, involving
some new and innovative features.
The zoological garden of Antwerp has been constructed between 1880 and 1910 in the
centre of the city, next to the central railway station. Recently the decision has been made
to renovate and modernize the zoo, including the reptile building. This building, dating
from 1906, is among one of the first reinforced concrete buildings in Belgium and is a
protected monument. In a first phase, the strengthening project focussed on the floor
In a second phase, the masonry columns at the basement level supporting the structure,
needed to be strengthened as well as the supported reinforced concrete beam. The
columns are made from brickwork, among which 10 rectangular columns with
dimensions of approximately 400 mm by 950 mm, 12 rectangular columns of 600 mm by
940 mm and 12 square columns with a side length of about 420 mm. The length of the
columns is equal to 2.0 m. The columns are grouped in four rows, each row supporting a
reinforced concrete continuous beam. Each of these beam has a total length of 38.6 m,
including 12 spans of 2.5 up to 2.7 m, a width of 400 mm (two outer beams) or 600 mm
(two inner beams) and a total depth extending below the floor slab of 200 mm.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
STRENGTHENING CONCEPT
It was required to increase the service load (dead + live load) on the slab from 5 kN/m 2 to
15 kNim 2• Given the heavy deterioration of the concrete floor (two-way slab on beams), it
was originally planned to first restore the original bearing capacity and to strengthen it in
a later phase by installing steel beams underneath the floor. This option would have
involved extensive disruption in the use of the building (relocating the animals and
closing the building for visitors) and would have required damage to the external walls in
order to bring in the steel beams. The building being a protected monument, public
authorities were very reluctant regarding the latter aspect. Furthermore, the overall costs
involved with this original rehabilitation plan were rather high.
As an alternative to the steel beams, gluing CFRP laminates to the soffit of the concrete
beams, appeared not possible because the concrete compressive zone would not be able
to withstand the increased internal compression force and would crush. Furthermore, it
was difficult to increase the compression zone above the floor slab because of the
existing equipment on the first floor. Therefore, strengthening by means of an alternative
system was envisaged. The company TRADECC/ECC NV, who pioneered the external
steel plate reinforcement of civil structures in Belgium when they strengthened the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
executed in merely 38 days, without removing the aquariums and with almost no
influence on the daily activities of the zoo.
The service load was increased to 15 kN/m 2 with a significant factor of safety,
whereas no heavy machinery was required and the external walls were not ~amaged.
High ratio between useful bearing capacity and own weight, and high durability of
the strengthening system.
Following the successful application of this system in 2000 on the floor slab, it was
decided to apply this technique further in 2003 for the strengthening of the structure at
the basement level. In this case, the system of multi-directional CFRP laminates and
lightweight CFRP reinforced wooden beams was proposed to strengthen a continuous
reinforced concrete beam supported by masonry columns. To upgrade the latter columns,
wrapping with AFRP sheets was proposed. For the rectangular columns, given the high
aspect ratio, the efficiency of the wrapping would have been very low and extra
transverse links were needed.
Whereas traditional CFRP strips used as externally bonded reinforcement are typically
composed of unidirectional fibres, newly developed multidirectional CFRP laminates
were used which also contained fibres in the plus and minus 45° direction, allowing the
use of extra mechanical fixations such as bolts. These laminates are manufactured
according to specification, up to 600 mm wide and up to 20 mm thick, allowing also
variable thickness. Both surfaces are protected by a scrim cloth.
In addition, prefab composite CFRP/wood beams (Fig. 2) have been developed which are
bonded and anchored to the concrete. These composite beams are light (ca. 160 kg/m>')
and easy to install without special equipment. They consist of balsa wood with the fibres
oriented perpendicular to the longitudinal axis of the beam, so that transverse stresses are
efficiently transferred between the CFRP laminate underneath and the concrete member
to which the wooden beam is attached. In the fibre direction, the balsa blocks have a
compressive strength of 10 MPa and a modulus of elasticity of 3.3 GPa. The concept
behind these prefab elements is to increase the lever arm between the externally bonded
CFRP tension reinforcement and the compression zone of the concrete. In this way the
efficiency of the FRP EBR is increased in providing extra moment capacity and bending
stiffness to the existing concrete section.
Design concept
To restrict the stresses in the compression zone of the beams supporting the two-way
slab, the strengthening system was designed to increase the lever arm between the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
external tensile reinforcement and the compressive zone. This could be realized by using
multidirectional CFRP laminates anchored by bolts and spacer elements (see previous
section). In this way, increased stiffness was obtained to limit deformations. Hence, the
design was basically determined by the serviceability limit state (limitation of concrete
compressive stresses and deflections), resulting in relatively low strains in the CFRP
reinforcement. Indeed, the elongation of the laminates under maximum service load was
limited to 0.1 % in order to obtain acceptable deformations of the slab, taking into
account that the new basin to be constructed has glass side panels.
Prior to bonding of the externally bonded CFRP laminates, balsa blocks were glued
underneath the concrete beams with structural epoxy adhesive. The spacer elements had a
width of 300 mm and a total depth of 200 mm or 400 mm, for the 300 mm by 500 mm and
the 200 mm by 300 mm concrete beams respectively. In this way the depth below the slab
of all composite beams equalled 700 mm. For the 200 mm by 300 mm concrete beams,
the width was first increased to 300 mm with epoxy mortar (Fig. 3). This mortar layer
was additionally anchored to the slab with steel dowels. At the anchorage zone (the beam
ends), epoxy mortar blocks were used instead of balsa in order to anchor the CFRP
laminates and to prevent contact of the balsa wood with the side walls (Fig. 4). Beneath
the concrete/balsa beams, a CFRP laminate 300 mm wide and 8 m long was glued and
anchored with steel bolts at the end zones (Fig. 4). The thickness of the laminate varied
between 3.6 mm at the anchorage zone up to 7.2 mm in the middle 6 m. Next, extra
transverse strengthening was provided by wrapping the soffit, the side faces and part of
the concrete ceiling, with prefabricated L shaped CFRP laminates. This increased the
shear strength and allowed a better stress transfer between the different materials.
The repaired secondary beams, which transversely connect the main beams, have been
strengthened with externally bonded CFRP laminates, without providing a balsa spacer.
The CFRP reinforcement had a thickness of 2.4 mm and width of 200 mm. In a similar
manner to the main beams, transverse wrapping was provided with CFRP.
Design concept
The strengthening in the second phase was dictated by the need for flexural and shear
strengthening of 4 continuous reinforced concrete beam, supported by masonry columns,
increasing the service load (dead+ live load) from 210 kN/m to 300 kN/m. Similar to the
floor slab, the flexural design was governed by stiffness considerations, so that a CFRP
laminate with a spacer was chosen. For the shear design, a strengthening configuration
was chosen allowing high efficiency by providing fibres perpendicular to the expected
shear cracks and anchoring the CFRP shear reinforcement. For the masonry columns,
although the stress level was still acceptable under service conditions, the design of the
wrapping was dictated by the required safety level at the ultimate limit state. At the same
time, the benefits with respect to increased ductility and robustness were of importance.
Wrapping of the masonry columns has been provided by applying 1 layer of AFRP (PC
AraComp 330W), with the fibre oriented at 90° with respect to the column longitudinal
axis (Fig. 6). The AFRP fabrics have a width of 300 mm and a weight of 330 g/m 2 • They
were wrapped with an overlap of 150 mm. For the rectangular columns, given the rather
high aspect ratio, extra transverse links were provided after wrapping to increase the
effective confinement. This has been done by means of two multidirectional CFRP
laminates glued longitudinally at opposite column faces, and connected through the
column by means of treated bars (galvanized steel) with a diameter of 16 mm, spaced
every 300 mm (Fig. 7).
After the concrete repair and column strengthening, the beams were strengthened in the
span zones by means of prefabricated PC CarboComp sandwich composite beams, with a
length of 2.5 m. These 100 mm thick by 400 mm or 600 mm wide strengthening elements
consist of wooden planks with a 2 mm thick CFRP laminate at the soffit as tension
reinforcement, wrapped transversely with CFRP. These prefabricated elements were
glued and bolted to the concrete beam (Fig. 8). In Figure 8, the extra wooden plank at the
soffit is temporary and only for ease of installation. Multidirectional CFRP continuity
reinforcement was provided by gluing and bolting at the ends, to connect the
prefabricated strengthening elements over the column support zone (Fig. 9). Next, shear
strengthening was provided by means of 2 layers of CFRP with a weight of 300 g/m 2 and
oriented at an angle of 45° to the beam axis. The upper part of the CFRP is anchored to
the beam by means of a bolted multidirectional CFRP laminate (Fig. 9).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Preceding the first phase of the strengthening project, preliminary tests were conducted(s)
at the Magnel Laboratory for Concrete Research of Ghent University. Besides standard
quality control tensile testing, these comprised both bond shear tests on the laminates and
bending tests on the strengthened system. The bond tests aimed to verify the feasibility of
anchoring the PC CarboComp Plus laminates with steel bolts. To study the behaviour of
concrete members strengthened with PC CarboComp sandwich beams, 4-point bending
tests were performed on beams and slabs.
To further investigate the use of bolts on the structural behaviour of FRP EBR
strengthened members, a more extensive test programme was executed on a series of 5 T-
beams strengthened with multi-directional CFRP laminates (type PC CarboComp Plus)
using different bolt anchorage configurations<9). The overall object of this test programme
was to demonstrate the feasibility of delaying or avoiding certain debonding mechanisms
and obtaining a more 'pseudo-ductile' behaviour at ultimate. The latter aspect aims at
increased deflections at ultimate and less brittle failure due to extensive slip and a more
gradual debonding.
The adhesion strength of the CFRP laminate to concrete was verified by means of two
bond shear tests, according to the test set-up shown in Fig. 10 (the test area is located at
the left hand side, the right part acts as the fixed reaction side). With the first specimen,
the maximum anchorage capacity was verified of CFRP PC CarboComp glued to
concrete. The CFRP laminate had a thickness of 1.5 mm and a width 100 mm. For this
specimen a bond length of 250 mm was provided, calculated(4) as the anchorage length
above which no significant increase in the bond strength of the laminate will occur. A
second specimen was identical to the first one, however anchored with a steel bolt
(diameter 12 mm) according to the configuration shown in Fig. 11. In addition to the
unidirectional fibres, fibres were also available in the plus and minus 45° direction in the
anchorage zones (PC CarboComp Plus, thickness in the anchor zone 1.8 mm). In between
the bolt and the laminate, a glued steel plate (130 mm x 100 mm) was provided. The bolt
was tensioned by applying a moment of 100 Nm.
From these bond tests it appeared that the use of a single bolt increased the anchorage
capacity from 47.5 kNto 68.5 kN, a 44% increase. The failure of the bolted specimen is
first characterized by laminate debonding and at a later stage by shear failure of the bolt
at the level of the concrete-laminate interface. Over almost the total bonded surface, a
thin layer of concrete remained attached to the laminate. The slip recorded during the test
is shown in Fig. 12 (in this figure the total load of the actuator is given, corresponding to
twice the load per CFRP laminate, see Fig. 11). At about 0.4 mm, the slip suddenly
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
increased corresponding to initiation of partial bond shear failure of the laminate and
after which the load could be further increased until shear failure of the bolt.
Four-point bending tests were conducted on 4 specimens (2 beams and 2 slabs) with a
span of 1.8 m. One reference specimen and one strengthened specimen of each type were
tested. The dimensions of the specimens, test set-up and configuration details are shown
in Figs. 13 and 14.
The unstrengthened beam (FIR) failed at a load of 2 x 32 kN, by yielding of the internal
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
steel followed by concrete crushing. The strengthened beam (Fl) failed at a load of 2 x
190 kN (strength increase by a factor 6), by a combination of shear failure in the
strengthened beam and shear failure between the epoxy end block and the wooden beam
part. The recorded load-midspan deflection curves of the beams are shown in Fig. 15.
Because of the low amount of steel reinforcement in the prefab slabs, the unstrengthened
slab (F2R) failed at first cracking, corresponding to a load of 2 x 13.4 kN. The
strengthened slab (F2) failed in shear at a load of 2 x 162 kN (strength increase by a
factor 12). This shear failure initiated as a vertical shear crack in the balsa wood and
extended as an inclined crack in the concrete slab (Fig. 16).
With respect to the reference specimens, for both the strengthened beam and slab an
increase of the strength and stiffness is obtained which are a multiple of the
corresponding reference values. The ultimate deflection of the strengthened beam is
about half that of the unstrengthened beam, while for the slab the ultimate deflection
increased to a large extent with respect to the reference slab. The latter aspect is due to
the small ultimate deflection of the reference slab (failure at first cracking). Though the
tendency of a reduced deformability at ultimate (as typically observed for concrete
members strengthened with FRP EBR), the internal steel is yielding when reaching
ultimate, so that a minimum required rotation capacity(?) is exceeded.
In an extensive test programme, the effectiveness of the use of bolts and hence the
advantage of the new type of multidirectional CFRP laminate was verified. Details of this
test programme are reported elsewhere<9 ), in the following a summary is provided. Four-
point bending tests have been executed on RC T-beams with a span of 4 m and a total
depth of 350 mm (beam width of 150 mm, respectively 350 mm at the flange, flange
depth 100 mm). In total, 5 beams have been tested, including one reference
(unstrengthened) beam, and 4 beams strengthened with multi-directional CFRP laminate,
with or without the use of additional mechanical fixings by means of bolts (type Ml2,
class 4.6 and Ml6, class 8.8, see Fig. 17).
For all the strengthened beams, one layer of multidirectional PC CarboComp Plus was
used, with a length of 3.66 m. The laminates are 1.8 mm thick and 100 mm wide, with
fibres over the full length in the longitudinal as well as in the plus and minus 45°
directions. In Beams 3 and 4, an additional multi-directional CFRP was glued at each end
of the laminate, and two bolts (threaded bars) were installed with epoxy adhesive. The
bolts were pretensioned with a force of about 10 kN. In Beam 4, also four bolts with the
An overview of the cracking load, yielding load, maximum load, strengthening ratio and
failure aspect of the tested beams is given in Table 1. The load vs. midspan deflection
curves are given in Figure 18. The strengthened beams, due to the presence of the
external reinforcement, are significantly stiffer than the reference beam after cracking.
Yielding occurred at a load level about 40 % higher than the reference beam, and even
after yielding of the steel reinforcement the load considerably increased. The maximum
attainable load in this test programme was limited by debonding of the laminate.
In Beam 2 the so-called concrete rip-off mechanism was activated, where a shear crack,
starting at the end of the external reinforcement, developed along the internal
reinforcement. In Beam 3 concrete rip-off was prevented. Indeed, when the first shear
cracks were observed at the end of the laminate, the tensile force in the bolts started to
increase, indicating the activation of the bolt anchorage. The load could be increased up
to a debonding failure away from the anchorage zone due to vertical crack displacement
(in regions with a combination of high shear forces and moments). After debonding, as
the bolts started working in shear, providing anchorage of the unbonded laminate, the
external reinforcement further acted as an unbonded tension member until anchorage
failure occurred (though at a lower load level than at the initial debonding level).
The test results of Beam 4 indicate a similar behaviour as Beam 3, preventing the rip-off
mechanism. However, due to the extra bolts in the shear span, initiation of peeling due to
vertical crack displacement did not lead immediately to complete debonding and allowed
for a higher strengthening ratio (15% higher than Beam 3). Similar to Beam 3, in a
secondary stage the external reinforcement acted as an unbonded tension member until
failure of the bolted anchorage. The anchorage failure was initiated with large slip,
damage at the bolt location and pulling of the laminate through the bolts.
The behaviour of Beam 5 is similar to the other strengthened beams at the first stage,
though a higher load than Beam 4 could be reached before debonding. After this, the
external reinforcement started working as an unbonded tendon, with anchorage mainly
provided by the outer bolts and with a certain amount of load transfer between the
laminate and the concrete due to friction and mechanical interlock (inner bolts kept the
laminate in contact with the concrete). Given the detailing of the anchorage configuration
of this beam, it appeared possible in this secondary stage to attain the same load level at
which the initial debonding failure occurred.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CONCLUSIONS
ACKNOWLEDGMENTS
The KMDA (Royal Society of Zoology Antwerp), the Ministry of the Flemish
Community and the European TMR Network ConFibreCrete are acknowledged for their
support.
REFERENCES
Technical report on the Design and use of externally bonded fibre reinforced
polymer reinforcement (FRP EBR) for reinforced concrete structures, Working Party
EBR of Task Group 9.3 'FRP reinforcement for concrete structures', International
Federation for Structural Concrete, Lausanne, 138 pp.
8. Magnel Laboratory for Concrete Research (2000), "Proeven in verband met een
versterkingssysteem voor betonelementen" (in Dutch), Test report 2000/030, Part 1
and 2, Department of Structural Engineering, Ghent University, 44 & 18 pp.
9. A. Nurchi, S. Matthys, L. Taerwe, M. Scarpa (2003), "Tests on RC T-beams
strengthened in flexure with a glued and bolted CFRP laminate", Paper submitted,
6th. International Symposium on FRP Reinforcement for Concrete Structures
(FRPRCS-6), Signapore.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
.. ··~
..
epoxy mortar
Fig. 6- Wrapping with AFRP of masonry Fig. 7 - Transverse links through the
columns wrapped column
6 7 8 9
12 3 4
4 1 2 3
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--LVDT1
--LVDT2
0.0~------------------------------------------~
rn
relab ·oea.m
xy mo,.Lar
~I
10
3 • 10 mm
Strengthened beam F1
= i!t:
to* tiJO:nm .,. to
2 • 8 mm
Q
pre~b ~:;~r
11 wrapping
i einforcement
200
180
/ - ....
--
fi140
160
7
17 --Reference beam F1 R f--
a 12o f--
1100
7 --Strengthened beam F1 f--
i 80
7
l 60
7
40
7
20 IT
0
~
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
Deflection at midspan (mm)
Fig. 15- Load-midspan deflection
Beam3
Beam4
BeamS
s1rengthened
FRPend
FRP 81eelplate
230x100x1.5
Beam2:
100
Z" Beam 1: concrete
I!!. crushing
G80
"CCI
..;f
....:!1
r:
60
-Beam 1(reference)
~ t-Jf~~======================~-Beam2~~m~
~ 1a 1a -Beam 3 (CarboComp + 2 bolts)
20 ~rn------1 U 71
ZS~--1------.L.----:zs" -Beam 4 (CarboComp + 6 bolts)
'------------------------~-Beam 5 (CarboComp + (2+4) bolts+
steel
0
0 20 60 80 100 120 1~ 160 180 200
Deflection at midspan [mm)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Synopsis: In 1997, a precast, prestressed T-beam in the Ala Moana Shopping Center
parking garage, in Honolulu, Hawaii, was strengthened in flexure using carbon fiber
reinforced polymer (CFRP) strips epoxy bonded to the soffit of the beam. When the
parking garage was demolished in June 2000, this beam and two control beams were
salvaged and brought to the University of Hawaii for testing. This paper presents the
retrofit procedures used during field application of the CFRP strips. It also describes the
beam recovery and preparation for laboratory testing. The test program and results of the
flexural testing of both unstrengthened and strengthened beams under four-point loading
are presented in detail. The CFRP retrofit significantly increased the flexural capacity of
the beam while also increasing its flexural ductility. The failure moment was well in
excess of the nominal moment capacity predicted using the strain-compatibility
procedure described in the ACI 440R-02 report.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
335
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
336 Robertson et al.
Ian N. Robertson, teaches structural engineering design courses and performs
experimental research investigating the performance of concrete and steel structures
under dynamic, cyclic and long-term loading conditions. He is a registered professional
structural engineer in the State of Hawaii and has been involved in structural engineering
design and research for over 20 years.
INTRODUCTION
During a routine structural inspection of the Ala Moana Shopping Center Parking Garage
in Honolulu, Hawaii, it was noted that one of the prestressed T-beams had substantial
spalling damage to the beam ledges and a large flexural crack across the bottom flange of
the beam. The beam is a precast prestressed inverted T-beam supporting joists and a
slab, which acts as the top flange for the T-beam (Figure 1). The beam was repaired in
1997 using CFRP pre-cured strips bonded to the beam soffit to augment the flexural
capacity (Figure 2). This was the first use of FRP materials for structural retrofit in the
State of Hawaii. CFRP wet lay-up wrap and epoxy mortar were used to repair the
damaged beam ledges. CFRP wraps were also provided at the ends of the span to restrain
the end of the CFRP tension strips.
The repairs were performed by Concrete Coring of Hawaii using Sika Carbodur pre-
cured CFRP strips epoxied to the bottom flange of the beam. The repairs were performed
following standard manufacturer's instructions, with no thought that the beam would be
tested at some future date. In additional to the retrofit beam, an identical unstrengthened
beam was tested as a control specimen. This is one of the first tests of a field installed
FRP retrofit after an extended field service period. The Hawaii Department of
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Transportation (HDOT) funded this research program in order to evaluate the potential
for FRP retrofit of deficient bridge structures in the state.
RESEARCH SIGNIFICANCE
Field application of FRP materials for strengthening of concrete members has increased
significantly subsequent to extensive laboratory testing of retrofit techniques (GangoRao
and Vijay, 1998; Fanning and Kelly, 2001; Shahawy et al, 2001; Spadea et al, 2002 and
many others). Very few of these field applications are available for testing after exposure
to service conditions. This paper presents the results of flexural testing of a prestressed
The CFRP retrofit was designed in 1997 by Martin and Bravo Structural Engineers,
Honolulu, Hawaii, following design procedures presented in the literature at the time. No
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
standard was available at that time fQr surface application ofFRP to concrete members.
The surface of the concrete was prepared by grinding to remove paint and weak surface
paste. This resulted in a relatively smooth surface preparation similar to ICRI-CSP
surface profile 2 (ICRI, 2002). Current FRP application specifications would generally
require a slightly rougher surface profile such as ICRI-CSP 3-4.
The CFRP strips were cleaned prior to installation and a layer of Sikadur 30 Hi-Mod Gel
epoxy was placed on the soffit of the beam. The strips were then pressed onto the epoxy
using a roller as shown in Figure 3. 150 mm wide strips ofSikaWrap Hex 103C uni-
direction carbon fabric were saturated with Sikadur Hex 300 epoxy and applied to the
ends ofthe beam to restrain the ends of the flexural strips (Figure 4). Additional CFRP
fabric sheets were used to wrap the epoxy mortar patches at the ledge spalls at third
points along the span. Since these additional wraps are not typical of flexural
strengthening, they were removed prior to testing the beam. The end anchorage wraps
were however left in place as they are commonly installed as part of the flexural
strengthening.
During demolition of the parking structure in June 2000, the precast prestressed beam
with CFRP strengthening was salvaged along with two nominally identical beams (Figure
5). In the recovery process, the top slab forming the beam flange was removed to
facilitate shipping. In addition, portions of the beam web suffered minor damage in the
form of concrete spalls and negative bending cracks. This damage was repaired by
personnel from PlasTech Inc., Hawaii, using Sika epoxy injection and epoxy mortar
patch materials (Figure 6). It is assumed that none of these repairs affected the flexural
performance of the beams during testing.
The top flange of the beam was reinstated.in the Structural Laboratory at UH with the
same reinforcement layout as the original slab (Figure 7). The dimensions of the four-
post test frame limited the flange width to 1500 mm, which corresponds to one fifth of
the beam span and a slab overhang of six times the slab thickness on either side of the
web. The ACI 318 building code effective flange width for this beam would be 1800 mm
(ACI 2002a). Beam failures were initiated by tension reinforcement failure and not
compression failure of the top flange. It is therefore assumed that the reduced flange
width did not significantly affect the beam flexural performance. The cross-sectional
dimensions of the beams are shown in Figure 8.
Table 1 lists the concrete material properties for the two T-beams tested in this study.
After beam testing, concrete cores were taken from the web and anchorage blocks of each
beam to determine compressive strength. Core recovery and testing were performed
according to ASTM C42-99 (ASTM C42, 1999). The cores were all 100 mm diameter
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
by 140 mm long. The compressive strength determined from each core was adjusted
because the length to diameter ratio was below 1.75 (ASTM C42, 1999). The resulting
concrete compressive strengths are listed in Table 1..
Standard 150 mm diameter by 300 mm concrete cylinders were made while pouring the
top slabs of each T-beam. These cylinders were tested in compression on the same day
as the corresponding T-beam test producing the average compressive strengths listed in
Table 1.
Table 2 lists the material properties of the reinforcing and prestressing steel used in the
test beams. Web shear reinforcing bars were recovered during beam demolition and
tested in tension. The average tensile strength for shear reinforcing in each T-beam is
listed in Table 2. Coupons of the 10 mm nominal diameter prestress strands were also
recovered from the test beams and tested in tension producing the ultimate tensile
strengths listed in Table 2.
Table 3 lists the material properties for the FRP materials used in the flexural
strengthening ofT-beam 2. These properties are based on the manufacturer's test
information since no material coupons were available from the original repair work.
Because of damage to the FRP materials during destructive testing of the T-beam, it was
not possible to recover representative samples for testing. Pull-off tests on the CFRP
strips would also not be representative of the installed condition because of damage
caused to the epoxy bond during the beam test.
Both control and strengthened T-beams were tested under four-point loading as shown
schematically in Figure 9. T-beam 1, the control specimen, was tested over a span of
7.24 meters with a pinned support at one end and roller support at the other. The load
was applied through two line loads each 610 mm from midspan. ForT-beam 2, the
strengthened beam, the support locations used for T-beam 1 would have been directly
under the ends of the CFRP strips. These reactions would therefore have enhanced the
restraint provided by the CFRP wrap at the ends of the strips. To avoid this additional
restraint, steel support brackets were bolted to the ends of the beam so that the supports
could be located beyond the ends of the beam. The span forT-beam 2 was 7.76 meters as
shown in Figure 9. Figure 10 shows T-beam 2 in the test frame prior to testing.
Because of the enhanced flexural capacity provided by the CFRP strengthening, the
flexural capacity ofT-beam 2 now exceeded the theoretical shear capacity. In order to
prevent a premature shear failure, shear reinforcement was installed on both beams in the
T -beam 2 response
T-beam 2 (strengthened beam) was tested under the same loading conditions as the
control T-beam 1. However, in order to prevent the support condition from providing
additional restraint to the end of the Carbodur strips on the beam soffit, steel extensions
were fabricated and bolted to the ends of the beam. This resulted in a longer span for T-
beam 2 compared with T-beam 1 (Figure 9).
During the flexural test ofT-beam 2, the response was similar to that for the control
specimen until flexural cracking of the beam. The post-cracking stiffness for T-beam 2
was greater than that forT-beam 1, and did not degrade as rapidly. Figure 13 shows the
moment-deflection response ofT-beam 2 compared with that for the control beam.
ACI committee 440 recently published a report on the strengthening of concrete members
using externally bonded FRP (ACI, 2002b). This ACI440R-02 report was used to predict
the failure bending moment for T-beam 2. The anticipated nominal moment capacity of
846 kN-m was easily exceeded by the strengthened beam, which supported a maximum
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The maximum midspan deflection forT-beam 2 was 100 mm compared with the 75 mm
deflection for the control specimen. The addition of CFRP flexural strengthening
increased the ductility of the beam. Failure occurred when the CFRP strips delaminated
from the bottom of the beam. This delamination appeared to initiate at the base of a
flexure-shear crack that had formed just outside the left load point. Vertical offset in the
soffit of the beam on either side of this crack may have contributed to the initiation of
delamination. In addition, large strain differential between the CFRP strips and the
flexurally cracked concrete may also have contributed to deterioration of the bond
between CFRP and concrete. If the CFRP shear reinforcement had extended around the
soffit of the beam, it may have provided additional restraint to the CFRP strips and
delayed the delamination, thus further increasing the flexural strength.
For the first 500 mm from the delamination initiation point, the failure occurred in the
surface concrete, with a thin layer of concrete remaining attached to the CFRP strips.
Beyond this point, the CFRP strips separated from the epoxy, likely because of the
increased angle of peeling as the CFRP stripped away from the beam soffit. The
delamination occurred rapidly and extended from the shear-flexure crack to the end of the
CFRP strips, which pulled part way out of the CFRP fabric wrap anchor. The anchor was
not sufficient to prevent pull-out once delamination had occurred, but there was no
tendency for delamination to initiate at the end of the strips as had been reported in some
laboratory studies.
Two precast prestressed concrete T-beams were recovered from a Honolulu shopping
center parking garage and tested in flexure in the University of Hawaii Structural
Engineering Laboratory. One of the beams had been strengthened in 1997 using CFRP
carbodur strips epoxy bonded to the soffit of the beam. The other was used as a control
specimen. The following conclusions were made based on the results of these tests.
• Sika Carbodur CFRP pre-cured strips epoxy bonded to the soffit of the
strengthened beam significantly increased the flexural strength over that of the
control beam without reducing the beam ductility.
• There was no visually noticeable degradation of the CFRP strips, CFRP fabric
wraps or epoxy bonding agents during the 5 years of field exposure between
application in 1997 and testing in 2002.
• The ACI 440R-02 strain-compatibility procedure for estimating the flexural
strength of concrete beams with externally bonded CFRP appears to be
conservative for the condition tested here. The failure bending strength of the
ACKNOWLEDGEMENTS
The authors are extremely grateful to Adriano "A. B." Bortolin ofSika Products, USA,
for providing valuable information concerning the original FRP application, at which he
was the Sika representative. Sika Products also donated all additional FRP and epoxy
materials required to repair the recovered beams and retrofit them in shear so as to avoid
a premature shear failure. The authors are also indebted to Brian Ide, the structural
engineer responsible for the original FRP strengthening design. Brian provided
construction drawings, design calculations and photographic records of the original
retrofit. Chandler Rowe and his colleagues at PlasTech Inc., Honolulu, Hawaii, are
thanked for donating their labor and expertise in the repair of the recovered beams and
for installation of the shear retrofit materials at the UH Structural Engineering laboratory.
This project was funded through research grant No. 46507 from the Hawaii Department
of Transportation Research Board. This financial support is gratefully acknowledged.
The opinions and observations made in this paper are those of the authors and do not
necessarily reflect the opinion of any of the project sponsors.
REFERENCES
ACI 2002a, "ACI 318-02/318R-02, Building Code Requirements for Structural Concrete
and Commentary", American Concrete Institute, Farmington Hills, Michigan, 443 pp.
ACI 2002b, "ACI 440R-02, Guide for the Design and Construction of Externally
Bonded FRP Systems for Strengthening Concrete Structures", American Concrete
Institute, Farmington Hills, Michigan, 45 pp.
ASTM C42, 1999, "Standard Test Method for Obtaining and Testing Drilled Cores and
Sawed Beams of Concrete", American Society for Testing and Materials, West
Conshohocken, PA, 4.
Fanning, P. J., and Kelly, 0., 2001, "Ultimate Response ofRC Beams Strengthened with
CFRP Plates," Journal of Composites for Construction, Vol. 5, No.2, pp. 122-127.
GangaRao, V. S., and Vijay, P. V., 1998, "Bending Behavior of Concrete Beams
Wrapped With Carbon Fabric," Journal of Structural Engineering, Vol. 124, No.1, pp. 3-
10.
ICRI 2002, "Selecting and Specifying Concrete Surface Preparation for Coatings,
Sealers, and Polymer Overlays", International Concrete Repair Institute Technical
Guideline No. 03732.
Shahawy, M., Chaallal, 0., Beitelman, T. E., and El-Saad, A., 2001, "Flexural
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Spadea, G., Mencardino, F., and Swamy, R.N., 2002, "Strength and Ductility of
Reinforced Concrete Beams Externally Reinforced with Carbon Fiber Fabric," ACI
Structural Journal, Vol. 99, No.2, pp. 163-171.
With CFRP Top Slab 62 (9025 psi) Average of3 cylinder tests
Figure 1: Precast Prestressed T -beam repaired using CFRP strips and wrap
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 2: Beam cross-section showing three CFRP strips for flexural strengthening
/ 11 '
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
------1~0mm•-------
Slab Reinforcement
2-leg 1Omm 0 Stirrups~~-----<H
at 300mm o.c. 140mm
Moment Capacity
1000
900
800
e 700.
:2:
-
~
c
Cll
E
0
RRO I
500
n1
:i
"C 400
.!!!
c.
Q. 300
<
200
100
0
0 10 20 30 40 50 60 70 80 90 100
Midspan Vertical Displacement (mm)
Moment Capacity
ezI
~
-
c
Cll
E
0
,
:!!
.!!!
·a.
c.
<
0 20 40 60 80 100 120
Midspan Vertical Displacement (mm)
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
349
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
structures. has been involved in numerous projects involving structural upgrade and
rehabilitation of concrete structures using conventional methods as well as FRP in
addition to full-scale in-situ load testing. Mr. Alkhrdaji is currently an active member of
ACI 440 (FRP) and ACI 437 (Strength Evaluation) and a member of ASCE and ICRI.
INTRODUCTION
Many water and sewer utilities and power generation facilities are increasingly
challenged with problems associated with deterioration of underground pipelines
constructed using prestressed concrete cylindrical pipes (PCCP). Of concern is the
sudden failure of these PCCP pipelines that form the backbone of main w.ater systems in
many US cities. PCCP pipelines are typically buried 4.5 to 7.5 m (15-25 ft) and extend
through heavily populated areas, which makes inspection and repair inherently difficult.
A large number of PCCP water mains in the US have been in service for 25 to 50 years.
For many years, these water mains were functioning well and minimum maintenance was
needed. As a result, water agencies were reactive to PCCP problems. Repair or
rehabilitation took place only after a break had occurred. However, in the past few years,
a number of pipeline blowouts have occurred in US cities that disturbed the service they
provide and resulted in costly repairs and collateral property damage. This raised the
concern of city officials, and led to a new focus on the development of inspection and
maintenance programs as necessary measures to prevent future blowouts.
Maintenance programs have been proven very effective for many agencies as they helped
prevent major breaks from occurring. To make the most of these programs, an adequate
repair and rehabilitation program needs to be established upon detection of deterioration
problems. This placed professionals in a search mode for the most effective and
economical methods for the repair and rehabilitation of PCCP pipelines. One method
that was proven effective consists of installing a new liner inside an existing PCCP. The
liner could be made of steel or FRP composite.
Upgrading with steel liner is achieved by installing a large diameter, thin, steel tube
inside a deficient PCCP. The steel liner is typically installed in sections and welded in
place. A low-strength grout is then injected into the annular space between the new liner
and the existing PCCP through ports manufactured in the steel liner. Cement mortar may
be placed around the inside of the liner as a final coating.
Another way may be achieved using a fiber reinforced polymer (FRP) composite liner.
FRP composites that have been extensively used by industries such as aerospace,
automotive, and sport equipment, are now a mainstream technology for upgrading
structures. For more than 15 years now, FRPs have been successfully used for the
structural upgrade and strengthening of buildings and bridges.
For pipeline ·applications, FRP systems are typically paper-thin carbon fabric sheets that
are bonded to the inside of the pipes with epoxy adhesive to create a multi-layer
composite lining system. Due to the high tensile strength of carbon fibers (8 to 10 times
that of steel), lining with carbon FRP composites can significantly increase the load-
carrying capacity of PCCP pipes. Their lightweight and corrosion-resistant properties
make them efficient and economical for the structural upgrade of these pipes. Cost
savings associated with reduced use of labor and heavy machinery and minimum shut-
down time giving FRP systems a major advantage to steel lining.
On November 17,1996, a 2.6-m (102-in). PCCP water main in Cranston, Rhode Island,
blew out, releasing 57 to 76 million liters (15 to 20 million gallon) of water through
residential streets (Figure 1). The pipeline serves two-thirds of the state and runs parallel
to a smaller, riveted steel water main. The pipe ruptured due to corroded prestressing
wires (the primary reinforcement that contained the 0.83 MPa (120 psi) water pressure),
soil overburden and traffic load. The $2-million emergency repair included stabilizing
the structural support of the smaller pipeline so that water could continue to flow to
customers. The repair also included replacing two 4.9 m (16-ft) long sections of the
pipeline. The possibility that other sections of the pipe may not be able to resist the
design forces prompted a condition evaluation and strengthening program.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The Providence Water Supply Board could not shut down its pipeline for extended period
to reline the deteriorated pipe sections. With these constrains, a carbon fiber in-situ lining
appeared to be the fastest, least disruptive, and most economical solution. Providence
Water Supply Board was the first to use carbon fiber to line the interior of a public
service water main.
Rehabilitation of the pipeline was performed in two phases. Phase 1 took place during
the winter of 1998-1999 and included installing 4 access manholes, performing 11
external repairs, and lining 15 pipe sections with carbon fiber composites. Phase 2 took
place during the winter of 1999-2000 and included installing 3 additional manholes,
repair 6 external coatings, and lining 11 pipe sections with carbon fiber composites. This
article discusses the design, application, and full-scale load testing of the FRP liner for
Phase 2.
The FRP upgrading system selected for this application was a proprietary FRP system
consisting of dry fibers and resins combined in a specific manner and installed using a
wet-lay-up method. The selected carbon fabric has an ultimate strength of 3,790 MPa
(550 ksi), ultimate strain at failure of 0.0167 m/m (in/in), and a tensile stiffness modulus
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Conceptually, the upgrade aimed at creating an FRP pipe within the deteriorated pipe
sections capable alone of resisting all internal and external loads. Because carbon FRP
fabric is flexible, it can be applied to the inside walls of the pipe using a hand lay-up
procedure thus eliminating the need for excavation. The FRP liner is made of a multi-ply
composite system consisting of longitudinal and circumferencial plies. To provide
continuity at the joints, the plies were to be extended approximately 0.3 m (12 in.) into
the adjacent non-repaired sections. This initial detail is referred to as Detail 1 and is
shown in Figure 2. Prior to finalizing the design, the contractor consulted a professional
engineer specialized in the design and testing of PCCP pipes as well as a national PCCP
manufacturer. A concern was raised by both that the transition detail at the joints may
not be adequate. According to their experiences, at higher water pressure, cracks could
form on the interior mortar lining that allows water to infiltrate and flow between the
steel cylinder and the concrete lining. Typically, this is not a concern since infiltrating
water will still be encapsulated by the main structural system. However, when using FRP
liner bonded directly to the mortar lining, water infiltrating through the cracks or joints
would flow behind the FRP liner thus bypassing the strengthening system and rendering
it ineffective. To avoid this behavior, a new waterstop detail (Detail 2) was developed.
A schematic for Detail 2 waterstop is shown in Figure 3. To verify the behavior of the
The FRP liner was designed to carry the 0.55-MPa (80-psi) service water pressure plus a
0.28-MPa (40-psi) surcharge. A load magnification factor of 1.6 was used to calculate a
factored design pressure of approximately 1.31-MPa (190-psi). The design live load
consisted of two HS-20 trucks passing parallel or perpendicular to the aqueduct length.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
External forces considered for the design of FRP liner consisted of hoop force due to
internal pressure, hoop force due to breakage of external post-tensioning strands, and
bending forces caused by the topping soil and HS-20 traffic. The stress in the FRP liner
due to internal pressure ffi is determined as follows:
PD tsFy
!ji =2!--t- (1)
f f
in which Pis the factored water pressure, Dis the pipe diameter, ts is the thickness of the
steel cylinder, tr is the thickness of the FRP liner, and Fy is the yield strength of the steel
cylinder. To calculate FRP stress due to prestressing steel breakage, an approximate
value for the initial concrete hoop strain due to prestressing &cb is first calculated using
the following expression:
Apfpe
~=
Sp tc Ec
m
in which Ap is the area of one prestressing strand,/pe is the effective stress in the strand, sP
is the spacing of the wires, tc is the thickness of concrete excluding the concrete cover for
prestressing wires, and Ec is the modulus of concrete.
At breakage of the prestressing wires, the prestressing force will redistribute to the FRP,
steel cylinder, and the remaining concrete shell. Considering strain compatibility, change
in strain for the three constituents at equilibrium is equal, say c. From force equilibrium,
the following can be derived:
Ac Ec (&cb -&)+As Es (&cb -&)-Af Ef &=0 (3)
From which the value of c is determined and the FRP liner stress due to prestressing
breakagejfi, is calculated as E1 c. To determine the external forces due to soil and HS-20
loading, the pipe section was analyzed using a finite element modeling program. The
circumference of the pipe section was divided to 20 equal segments, as shown in Figure
4. Hinges were assigned at mid-height of the section, at nodes 6 and 16. At these
FRP design was achieved using a trial and error approach in which a different number of
FRP plies was assumed until the sum offft, jjp, and ffom was less than the ultimate strength
of the FRP fabric using a strength reduction factor of 4> = 0.85. Accordingly, three plies
were to be applied circumferentially. In addition, one ply was to be placed longitudinally
(along the pipe axis). The final thickness of the FRP liner was 3/ 16 in., and the FRP stress
at maximum loading was 3.16 GPa (458 ksi). Although the inside diameter of the
pipeline is slightly reduced due to FRP lining, the capacity of the pipeline was not
affected. The increase in capacity caused by reduced friction on the pipe walls (due to
smooth final coating of FRP lining) offsets the effect of the smaller pipe section.
FULL-SCALE TESTING
To validate the FRP design and the proposed waterstop details, the specialty concrete
repair contractor conducted two full-scale tests. The first test consisted of one pipe
section. The carbon fiber was installed on the inside wall and terminated at the spigot
and bell joints. The objective was to determine the capacity of the assembly if water was
able to flow through cracks on the concrete liner or at the joints. After the carbon fiber
liner was installed, the prestressing strands were cut; and the pipe assembly was sealed
with large steel bulkheads. At the bell end of the pipe the bulkhead had a spigot detail
while at the spigot end of the pipe the bulkhead had a bell detail. Both ends were sealed
using a standard 0-ring system to simulate the actual joint construction detail. The two
bulkheads were tied together using six, 75 mm (3-in) diameter, high strength steel bars.
The bars were tightened enough to hold the system without introducing any axial stresses
in the walls of the pipe.
After assembling the setup, the test pipes were filled with water and were progressivety
pressurized until failure was achieved. As expected, the pipe section failed at
approximately 63% of the maximum service pressure, a value corresponding to the
capacity of the steel cylinder alone and is well below the maximum design pressure.
Examination of the failed pipe section revealed that the FRP liner was intact and that
failure was governed by yielding of the steel cylinder. This failure mode proved the
concern that Detail 1 waterstop was not adequate.
The second test consisted of three pipe sections jointed together using a standard bell and
spigot detail (Figure 5). Testing of multiple sections (versus a single pipe section)
provided the opportunity to test a complete waterstop termination detail at both the spigot
and bell ends of a pipe section. The middle section was strengthened using the same FRP
layout; however, Detail 2 waterstop was used at the joints.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
This test demonstrated that the theoretical approach used to calculate the capacity of the
strengthened section is valid and that a significant capacity can be achieved if adequate
waterstop detailing was used, in this case Detail 2. Based on the full-scale test results a
decision was made by the engineer to proceed with FRP liner construction using the
alternate waterstop detail (Detail 2).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The project team planned to rehabilitate the deteriorated pipe segments during the winter,
when water demand ·was low and the pipe could be dewatered for several weeks without
disturbing the water demand.
Access points to the pipeline were limited and, predictably, not always near the areas to
be repaired. Ladders were placed in the 9.1-m (30-inch) diameter service holes to
provide access to the repair areas, and a tripod and winch pulley system was used to raise
and lower materials into the work area. Because of the confined work environment,
special safety measures needed to be implemented. All technicians completed confined
space training. Personal protective gear required by each individual entering the pipe
included a protective suit, rubber boots, latex gloves, hardhat, safety glasses, full-body
harness, personal lighting, and escape air pack. Respirators, dust masks, and other
supplemental safety gear were also used depending on the work being performed. Air
monitors measured levels of oxygen, carbon monoxide, lower explosive limits, and
hydrogen sulfide. A ventilation system provided 20 to 60 air changes per hour depending
on the size of the work area.
The project started in typical, January, New England weather in a pipe that had not been
dehumidified for many years. This created high levels of humidity and temperatures that
required constant monitoring to ensure ambient conditions for the carbon fiber
installation. Blowers, dehumidifiers, and heaters were used to maintain the relative
humidity level at 40- 50 percent and temperatures at 13-l6°C (55-60°F). Air circulation
and conditions of this level created challenges considering that service hole spacing could
be up to 610 m (2000 ft).
As with any surface·adhered system, bond of the FRP system to the existing concrete is
very critical and surface preparation of the concrete surface is very important. To achieve
this, the pipe's interior surface was water-blasted to remove loose materials and create a
concrete surface profile CSP 3, based on the classification of the International Concrete
Repair Institute (ICRI) and the manufacture's recommendations. Scaffolding was erected
to allow the workforce to reach the top of the pipe and to prevent walking on the
completed bottom areas (Figure 7). To prepare the surface, an epoxy primer was applied,
followed by a trowel-applied epoxy putty material to fill voids and level imperfections.
The surfaces were now ready for FRP installation.
Prior to installation, the FRP fabric sheets were cut to a predetermined length in an
above-ground staging area. Once the sheets were lowered into the work area, they were
saturated with an epoxy saturant. Each layer of FRP was then applied to the pipe's
circumference in three sections with a 100-mm (4-in.) overlap. Metal rib rollers were
used to push out any air bubbles, and to press the FRP sheets into the saturant and the
concrete liner to achieve "intimate contact." A second layer of the epoxy saturant was
then applied to form a complete fiber/laminate matrix.
This process was repeated for the subsequent layers of carbon fiber. Lastly, an epoxy
topcoat approved for potable water applications was applied to protect the FRP and to
provide a safe, sealed passageway for the water.
Using lightweight, flexible carbon fiber material for strengthening the Providence
Aqueduct turned out to be an innovative, cost-effective solution. In fact, the total project
cost came in on time and budget; and its success earned the International Concrete Repair
Institute's ICRI Award of Excellence in the Water Systems Category that year.
The Providence Aqueduct upgrade project now serves as a model for water utilities
facing similar repair situations, where a proactive condition survey and preventative
repairs avoided disastrous failures or where excavation of pipes is neither cost-effective
nor desirable. This project represents one of the first-time-ever on-site installations of
FRP liner in potable water pipes for strengthening purposes. But based on the success of
this project, FRP is likely to be adopted as a choice repair technique for future municipal
and utility plants waterwork projects throughout the United States involving isolated pipe
section repairs.
made with minimal disruption to the public; and except for some locations where pipe
sections had to be excavated for external repairs, the community hardly noticed this
pipeline repair project. The supporting full-scale tests proved the integrity of the
upgraded pipe beyond theoretical performance and assured the owner that strength was
added to the pipe sections.
NOTATION
REFERENCES
Fortner, Brian, (1999), "Main Line Mending," Civil Engineering Magazine, American
Society of Civil Engineers, July 1999, pp. 42-45.
Civil Engineering Magazine, "Carbon Fiber Locks Out Pipeline Corrosion," American
Society of Civil Engineers, December 1997, p. 10.
Lewis, Richard A., (1995), "A Managed Approach to Prestressed Concrete Cylinder Pipe
Infrastructure Utilization," 1995 Conference on Corrosion and Infrastructure- Practical
Applications and Case Histories - Final Program, Sponsored by NACE International, US
Department of Transportation, and American Public Works Association, November 28-
30, 1995, 164 PP.
PP.
Mechanical Anchor
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
361
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
362 Abdelrahman et al.
Dr. Gamil Tadros, MAC!, is Technical Application Consultant in the Canadian Network
of Centers of Excellence on Intelligent Sensing for Innovative Structures "ISIS Canada",
Winnipeg, Manitoba, Canada. He is an advanced engineer with Strait Crossing Inc. with
responsibilities in bridge design and construction.
INTRODUCTION
The Middle East is challenged by the rapid deterioration of concrete structures due to
corrosion of steel reinforcement. The long coasts on the Mediterranean, the Red Sea and
the Arab Gulf cause an adverse environment that accelerates corrosion of steel. Use of
advanced composite materials, (ACM), proved to be a competitive solution to extend the
service life of structures. This is not only due to the non-corrosive properties of the
material but also for the large improvement in mechanical properties that allows the
structural engineer to choose the best material for application.
ACM have been used in civil engineering applications in Egypt for over a decade.
Research started in the use of Fiber Reinforced Polymers (FRP) bars as internal
reinforcement, followed by externally bonded FRP laminates for strengthening concrete
structures< 1l. As experience was gained in the design and application of FRP, several
rehabilitation projects using FRP laminates were successfully completed.
The Panorama building in one of the hotels in the city of Sharm-El-Shaikh, Egypt, is 11.0
m wide and 50.0 m long(2J with a triangular shape, as shown in Figure 1. The one story
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
building is located close to the edge of a hill. Villas are located at the ground level, while
the roof was not used. It was the owner's desire to use the roof as an open-air cafe
having a view of the city from the top of the hill. The building consists of reinforced
concrete slabs supported on rectangular beams and reinforced concrete columns.
After two years of construction, both the slabs and beams were severely cracked, as
shown in Figure 1. The crack width was up to 1.0 mm in some locations. The cracks
were observed on the entire width of the slab, at the fourth bay away from the hill,
crossing all beams and walls. Flexural and shear cracks were also observed at some of
the beams. The Parapet of the roof, which is made of brick, was also cracked. The cracks
were wide at the top of the parapet, reducing in width towards the bottom. The location
of these cracks coincided with the cracks in the concrete slab. No cracks were observed
in the columns.
were made crossing the existing cracks in the slabs and beams. No further development
of the cracks was observed.
The second phase of the strengthening scheme was to restore the building and increase
the structural capacity of both the slabs and beams to resist higher live loads. It was the
owner's demand to minimize the time and the working space during this phase of work.
Two alternatives were studied to strengthen the superstructure; enlarging the concrete
section of beams and slabs, and use of CFRP laminates. Based on cost benefit, it was
decided to use the CFRP laminates to strengthen the slabs in flexure and the beams in
flexure and shear.
Material Selection
CFRP strips of 1.2 mm thickness and 50 mm width were used to strengthen the slabs and
beams in flexure, as shown in Figure 2. The strips have a fiber volume content of 68%
and epoxy resin(4l. The tensile strength and modulus of the strips were 2800 MPa and
165 GPa, respectively. The strips were bonded to the concrete surface using epoxy-based
two-component adhesive mortar. The adhesive strength of the mortar to the concrete
surface was 4 MPa. CFRP laminates with a thickness of 0.13 mm were also used to
strengthen the beams in shear as shown in Figures 3 and 4. The tensile strength and
modulus of the laminates were 3500 MPa and 230 GPa, respectively.
Design Criterion
It was important to estimate the stress in the steel reinforcement caused by the additional
straining actions resulting from the excessive settlement of the foundations. The
measured crack width of the slabs and beams were used to estimate the tensile stress in
the steel reinforcement. The required area of CFRP was calculated to allow for double
the live load, accounting for the increase in the stress of the steel reinforcement. CFRP
strips, 11.0 meters long and spaced every 500 mm were used on top of the slabs, as
shown in Figure 4. CFRP strips were used on the bottom surface of two cracked beams
to increase its flexural capacity. One layer of CFRP laminates was applied on the sides
of the beams in a "U" shape to strengthen the beams in shear, as shown in Figure 4. The
laminates were 300 mm wide and spaced every 50 mm.
Application of CFRP
The tensile strength of the concrete was measured and found to have an average value of
2.0 MPa. Before application of the laminates, the humidity of the concrete was measured
The second case study is a reinforced concrete multi-story building constructed ten years
ago using a flat slab system in Heliopolis, Cairo. The building is 30 m wide and 80 m
long divided into three parts with two expansion joints across the entire width of the
building. In addition to the basement, the first and second parts of the building are 12
stories high, while the third part is only seven stories high. The thickness of the solid
slab is 220 mm in the first and second parts and 240 mm in the third part. The flat slab is
supported on rectangular columns spaced every 4 to 6 m. The columns are supported on
1.0-meter deep mat foundation. A general layout of parts I and 2 of the building showing
the expansion joint and distribution of columns is shown in Figure 6.
The different floors of the residential building were designed to carry a live load of 2.0
kN/m2 • The owner of the building requested to use the ground and first floors in the first
and second parts for commercial purpose. This required retrofit for both slabs to carry a
live load of 5.0 kN/m2 • It was also the request of the owner to finish the strengthening
procedure in three weeks period including design and construction.
Finite element analysis was carried out to calculate the induced bending moments in the
flat slab due to self-weight, superimposed dead, and live loads independently. The
maximum calculated service moment was 88 kN-meter, while the resistance of the
critical section of the slab was only 65 kN-meter. This indicated that an increase in the
flexural capacity of the slab by 35 percent was required. The punching shear capacity of
the slab was checked and found to be sufficient without strengthening the slab in shear.
Material Properties
CFRP plate bonding system requires use of unidirectional carbon fibers attached with
resin to the surface of the concrete. The system utilized high strength CFRP plates with
an ultimate tensile strength of 2800 MPa and elastic modulus of 165 GPa. The plates
were 50-mm wide and 1.2-mm thick with a fiber volume content of 68 percent. The
plates were supplied in 50-m length and rolled with 1.5-m diameter as shown in Figure 7.
The resin was a two-component structural epoxy paste adhesive with tensile strength and
elastic modulus of 25 MPa and 4.5 GPa, respectively. Epoxy mortar with higher tensile
strength and modulus was used for leveling the concrete surface before applying the
plates.
Pull-off tests were carried out on the concrete slab, as shown in Figure 8 to ensure
enough tensile strength before proceeding with the application of CFRP. The tensile
strength of concrete was found to be 2.5 MPa, which is higher than the minimum value
recommended by the manufacturing company of 2.0 MPa. The high tensile strength of
concrete ensured good bond between the CFRP plates and concrete.
Design Philosophy
Design of CFRP plates was carried out using the requirements of both the ACT Code
318(5) and the recommendations of the ACI Committee 440(6) for design of FRP systems
used for strengthening concrete structures. The design criteria covered the strength,
serviceability, and ductility requirements as well as the risk of loss of CFRP plates in the
event of fire.
Strength Requirements
The nominal strength of the slab was calculated based on the flexural, shear, and bond
capacity of the section. The design criteria of the section included each of the following:
Strength reduction factors,~. was taken equal to 0.7 for both flexure and shear, and equal
to 0.6 for delamination of plates. Mechanical anchorage was used in case that anchorage
length of CFRP plates was not satisfied.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Ductility
The strain in the steel at the onset of maximum capacity of the concrete sections was
maintained to be at least 0.005. This was achieved by limiting the maximum percentage
ofCFRP plates in RC sections so that the strain in the steel reinforcement exceeds 0.005.
This strain value ensures yield ofthe steel and enough deformation to satisfy the ductility
requirement. The ACI Committee 440( 6) recommends a minimum of 0.005 tensile strain
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
to design tension-controlled sections. At this strain level, high deflection and large crack
width will guarantee enough warning signs before failure.
Fire
Risk of loss of CFRP plates in the unlikely event of fire was taken into design
consideration. High temperatures due to fire will cause a plastic flow of the epoxy
resulting in a loss of the load transfer to the CFRP plates. Typically, the critical
temperatures for epoxy are in the range of 50°C to 90°C<6l. In recognition of the
temperature risks, the unstrengthened slab was checked to ensure an ultimate capacity
that provides a positive factor of safety against collapse. The unstrengthened capacity
was capable of resisting the service loads without yielding of steel reinforcement. The
ultimate strength of the unstrengthened system exceeded the service loads by a factor, ~T
of 1.2 according to equation 1.
(1)
Where Sn, SD, and SL are the nominal strength of the unstrengthened section and induced
straining actions due to dead and live loads, respectively.
Construction Details
Application of CFRP plates started with preparation of the concrete surface, followed by
applying the epoxy mortar and finally attachment of the plates. The purpose of surface
preparation was to remove the outer, weak and potentially contaminated concrete skin
together with poorly bound material, in order to expose small- to medium-sized pieces of
aggregate. Mechanical steel hammers were used to prepare the surface. This was
achieved without causing micro-cracks or other damage that may reduce bond. The
concrete surface was leveled and epoxy mortar was used prior to the application of the
plates to fill large voids and blowholes, as shown in Figure 7. Extra care was taken
during the application of epoxy mortar to ensure that it produced full bond between
CFRP and concrete and that composite action was developed by the transfer of shear
stress across the thickness of the adhesive layer.
CFRP plates bonded to the bottom surface of the slab had enough development length;
while the plates bonded perpendicular to the edges of the top surface of the slab did not
have enough length to transfer forces to the concrete. Steel angles anchored to the
concrete slab with steel bolts were used as mechanical anchorage for the CFRP plates, as
shown in Figure 11. It was reported that the mechanical anchorage can transfer 60
percent of the maximum force of the plates< 8l. The maximum tensile force in the
mechanically-anchored CFRP plates was designed to be 30 percent of the maximum
force to allow for a factor of safety of "2" against anchorage failure.
Another detailing problem was encountered during placing the CFRP plates on the top
surface of the slab. This was due to the fact that the plates did not intersect with the
columns, but were placed around the columns, as shown in Figure 12. The dimensions of
the columns varied from 0.6 x 0.6 m to 0.30 x 1.50 m, which means that there were no
bonded CFRP plates on a 1.5-m segment of the maximum negative bending moment
zone. In order to evaluate the forces in the slab around the columns, finite element
analysis was conducted. Both steel reinforcement and the CFRP plates were modeled as
frame elements, while the concrete was modeled as shell elements. Distribution of the
stresses in the steel bars perpendicular to the long side of the columns is given in Figure
12. It can be seen that the maximum stresses occur in the steel bars located on the sides
of the column, reducing towards the centerline of the column. The results of the FEA
showed that adding CFRP plates only in the perpendicular direction to the long side of
the column reduced the stresses in the steel bars by 5 to 28 percent. Bonding CFRP
plates in the perpendicular and transverse directions to the long side of the column
reduced the stresses in the steel bars by 9 to 28 percent. Therefore, it was decided to
bond CFRP plates around the columns in both directions to reduce the stresses in the steel
bars. The calculated maximum stresses in the steel bars with and without strengthening
were 194 and 214 MPa respectively, as shown in Figure 12.
A total of 2200 linear meters of CFRP plates were used in strengthening the concrete
slab. Strengthening procedure did not stop the other civil or the electrical work in the
building. The Application of the plates was successfully completed in only ten days.
Loading Test
After application of the CFRP plates, loading tests were performed on the largest two
bays of the slab, according to the Egyptian Code of Practice for Design and Construction
of Concrete Structures<9l. Sand packages were flaced uniformly on the top surface of the
slab to produce an equivalent load of9.0 kN/m, which is equal to one and half the value
of expected live load in addition to the weight of the superimposed dead load on the slab.
The maximum deflection was found to be 6 mm at the mid-span of the slab after 24 hours
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CONCLUSIONS
This p~per h!trcd!!ces de~ign ccncept ~nd ccnt:tr-~cticn det2.i!e cf ~.vc success~~!
applications of CFRP laminates in strengthening RC structures in Egypt. Analysis of the
strengthened concrete slab with CFRP strips using finite element analysis is also
introduced. CFRP strips were attached to the concrete surface by epoxy paste as well as
mechanical anchorages using steel angles. Application of CFRP was completed in short
time compared to other conventional techniques in strengthening.
REFERENCES
be a
in torsion Plan
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
r
23
• •
Part 1
I • • •
•
Part 2
• •
I
25
• • • • •
l I I I
40.0
I I I I
l
Figure 6: Layout of parts (1) and (2) of the building
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 9: CFRP strips attached to the Figure 10: CFRP strips attached to
bottom surface of concrete the top surface of concrete
220
200
180
Stress in steel
bars (MPa) 160
140
120
-1.5 -1.0 -0.5 0 0.5 1.0 1.5
Distance from the middle of the column (m)
Longitudinal
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
CFRP strips
373
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
374 Basler et al.
Miklos Basler is the Key Account & Business Development Manager for Structural
Strengthening for Sika Services AG, Zurich, Switzerland. Miklos is a member of ACI
and the Swiss Society of Engineers and Architects, fib Task Group 9.3- FRP
reinforcement for concrete structures and of the IABSE Working Commission 8:
Maintenance and Repair of Structures. He holds a B.S.C.E. from the Faculty of Civil
Engineering, University in Belgrade.
David White, P.E. is the Director of Marketing, Repair and Protection for Sika
Corporation, Lyndhurst, NJ, USA. He is a voting member of ACI Committee 440, as
well as Committees 503 and 548. David is also active in the International Concrete
Repair Institute Strengthening Committee and the American Segmental Bridge Institute.
He holds a B.S.C.E. from Columbia University in New York and an M.S.C.E. from
Polytechnic University in Brooklyn, N.Y.
Mario Desroches is the Area Manager for Sika Canada in Quebec, Canada. He is the
Director of the Quebec Chapter for ICRI. Mario holds a B.S.C.E. from Montreal
University, Ecole Polytechnique and has worked for Sika since 1991. He personally has
supervised over 60 FRP strengthening projects and has also collaborated with ISIS
Canada for FRP Research applications.
1. INTRODUCTION
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Simple solutions for shear strengthening are much in demand. As with flexural
strengthening, a number of different systems are available in the market. These included
steel plates, FRP Fabrics and FRP Plate Systems. Steel plates require cutting, drilling and
corrosion protection, followed by anchorage and bonding onto the structure. Fabrics have
the disadvantage of being difficult to anchor, especially in the peel zones.
The proposed CFRP L-shaped plates combine easy installation on site with the unique
advantages of FRP materials. Positive features are their low weight, corrosion resistance,
high strength, consistent manufacturing and predictable mechanical properties. The system
does require compatibility between the L-shaped plates and the adhesive.
The CFRP L-shaped plates differ from the straight CFRP plates in both appearance and
production technology [3]. The matrix for both is a durable epoxy-based thermoset resin.
The L-shaped brackets have a 90° bend with an internal radius of 25 mm and are
approximately 1.4 mm thick (Figure 1). All of the tests were carried out with the 40 mm
standard plate width. The surfaces of the L-shaped plates are protected on both sides with
peel-ply fabrics without any release agent. They have to be removed immediately before
installation. The use of solvents on site for cleaning the surface is not necessary and should
be avoided. To achieve improved anchorage, the ends of the plate legs to be anchored have
to be covered with a thin layer of adhesive. This can be done with the help of a serrated
trowel one day before installation on site (Figure 2).
Three preliminary tests [3] and [6] on the CFRP L-shaped plates are described below.
An alternate method of drilling the anchorage holes has been developed by a specialised
Swiss manufacturer. The equipment is based on a comer saw with a diamond tipped
chainsaw that is capable of drilling rectangular, vertical openings into the compression slab,
perfectly in line with the surface of the web (Figure 4).
After the adhesive had cured for seven days, the concrete blocks were clamped in the test
machine and tensioned from the end of the long leg. Strain gauges were installed next to the
bend of the L-shaped plates and also to the bonded part of the short leg. The results showed
that the bonded length of the short leg had little influence over the failure load. The failure
mode was always shearing of the concrete along the length of the whole plate, and never a
failure in or around the bend zone of the plate.
So far, there does not exist a satisfactory formula, which takes all these influencing
factors into consideration. The usual standards, therefore, give simplified models that are
conservative. The truss model, with shear transfer in the compression zone (shear
cracking model), is widely used. In this model, the shear strength of a section with
conventional reinforcement is calculated as the sum of the shear loads, each taken by the
concrete and the reinforcement.
The concrete quality corresponds to that of concrete normally used for building
construction (compressive strength after 28 days: > 45 MPa). After hardening of the
concrete, anchorage holes were drilled using a core drill into the flange from the top of
the compression slab. The anchorage lengths therefore corresponded to the full
compression slab depth. The CFRP L-shaped plates were then bonded to the prepared
concrete with a structural adhesive having an overlap at the underside of the full web
width.
For comparison purposes, the load I deflection curves of two previous tests are also
shown. Cross-section and span of both correspond to beam Tl. The measurements show
that beam Tl had approximately the same load behaviour as the EMP A beam from 1992
with typical steel stirrups 0 10 mm, s=225 mm. In 1981 a test beam with external flexural
and shear strengthening with bonded steel plates, was tested and showed nearly the same
load I deflection behaviour as beam T 1.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Test beams specially designed for high shear stresses were statically tested to failure with
different shear reinforcements (test beams S1 to S4). A fifth beam was tested to investigate
the load-bearing behaviour of a pre1oaded and subsequently strengthened beam (test beam
S5). Lastly, a test beam was subjected to 5 million load cycles at a high load level and a
subsequent failure test (test beam S6). A range of significant data was collected through
extensive measurement. The results were used to improve the design concept for the shear
strengthening of reinforced concrete (RC) structures using CFRP L-shaped plates.
Based on these six combinations, together with an extensive measurement concept, a lot
of useful information was collected (Figure 11 ). Besides the loading force, the steel
strains, the concrete compressive strains and the strains in the L-shaped CFRP plates
were registered. Furthermore, crack measurements were carried out on the beams.
The main findings and proposals for the design can be summed up a follows:
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
"Anchorage pull out of CFRP L-shaped plates" have to be considered.
6. CASE STUDIES
Seven parallel girders of the end span, each 900 mm high and 400 mm wide required
strengthening at one end only. Six pairs of L-shaped plates, spaced 140 mm, have been
applied. The 84 CFRP L-shaped have been anchored each in a l 00 mm deep oblong holes
formed by drilling 3 0 26 mm overlapped boreholes. Just prior to the application of the
L-plates, these openings were filled with the same structural adhesive that is used for the
bonding typical CFRP plates to the concrete.
After full curing of the structural adhesive, the two round openings were cut out
(Figure 16). The strains were measured before and after the installation of the CFRP and
checked with the structural design made initially.
Systematic tests and the first applications of CFRP L-shaped plates show the potential of
this new strengthening composite system. Their contribution to increase the load-bearing
capacity of the reinforced structure is clearly shown. The factory-made brackets ensure that
the material properties are well defined and consistent. The system is user-friendly for
installation. Anchorage is a possibility, something that is more difficult to achieve with
standard CFRP fabrics.
8. ACKNOWLEDGEMENTS
The authors acknowledge the support of the whole involved EMPA staff. A special thank
is addressed to Mr. Christoph Czaderski for providing photos and diagrams.
REFERENCES
[1] Meier U., Deuring M., Meier H., Schwegler G.: Strengthening of structures with
CFRP Laminates: Research and applications in Switzerland. Advanced composite
materials in bridges and structures, 1st International Conference. Sherbrooke,
Quebec, Canada, 1992, pp 243-251.
[2] Steiner W., 1997: Strengthening of structures with carbon fiber plates. Composite in
construction - conventional and innovative. International Conference. Innsbruck,
Austria, September 16-18, 1997, pp 723-728.
[4] EMPA Test Reports 169219E/1 and 169219E/2: Testing of CFRP shear plates on
reinforced concrete T-beams, Diibendorf, 1998.
[5] EMPA Report No. 116/7, August 2002, "Shear strengthening with prefabricated
CFRP L-shaped plates. Test beams S1 to S6.
[7] ACI 440.2R-02, "Guide for the Design and Construction of Externally Bonded FRP
Systems for Strengthening of Concrete Structures", American Concrete Institute.
[8] fib, "Externally bonded FRP reinforcement for RC structures "Technical report,
Bulletin 14.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
225 69 55
300 74 59
100
90
......
..... 80
/""
~ 70 /
~ 60
v . . . . . .·•· . ......
c: 50
·~ 40
5: 30
w 20
10
0
50 100 150 200 250 300 350
Anchorage I Bend Zone length [mm]
.1 __.,_Anchorage[%] --• - ·Bend Zone [%] I
]00,-----,------,-----,------.-~--~
liD
~~·~···~--~···~--~~~~~~
,.~/~ ........... - ,_ ................ .... ..... I
~~ ij I
J4ll
-o 3D
//.
:1
--.. I
J
!3 : 1-+----+-+------
~3D •
1 ~
• •• _ _ ,...,..........
-""121'713
10)1-1.1~1--------j·----·~- ..............
I
oL----+----+---~--~~-L--~
0 100
Md-$Jin ll!&dirl [rmj .
700 _,.
I
•
'•
.. .. . .
--
-- . ..............
.:;-r-:· ·············:. .....................
!
.
...........
, .....
~
/II I
:·II I elediiodfdelemtled dE6!d011<2
-
,
1\h:DI
-
••• e.Rt.-...bean~BIIRI.tesl
I ~ -
l!pllflh fZl773
-l!erlnT1 &Ed:I'L1097
100 -
0 I I
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0 100 100
Crou.Sectlon
FW2 F._.t2
~1 :1 r ··um·-~~
1.25m
t 1.00m t 1.25m
I..1. 3.50m
.1 •. I
0.25m 0.25m bw=0.1s~l 1.
Figure 10: Key dimensions of test beams Sl to S6
_,\ i
Synopsis: A detailed evaluation, combined with a new snow and wind study, was
carried out on the Olympic Saddledome to confirm its ability to support heavier
suspended loads. The results of the study confirmed that the main roof cables have
sufficient capacity to resist the increased load. However a new snow loading condition,
identified in the snow study, had the potential to overstress specific roof panel elements.
A finite element analysis of the roof panels supported the conclusion that under the
newly identified snow/wind load combination, several roof panels of the Saddledome
will face demands that are above the ultimate roof capacity according to the National
Building Code of Canada (NBCC 1995). Therefore, it was decided to strengthen selected
roof panels of the Saddledome using FRP plates.
Special consideration was given for analyzing the effect of creep on the strengthened
concrete panels. This was due to the relatively high creep stress expected in the roof
panels as a result of using lightweight concrete. It was feared that such high stress might
result in creep of the epoxy resins and system debonding under loads lower than that
predicted by the analysis. In this paper, Design considerations for the effect of creep are
discussed and the results of full scale tests are presented. Performance design for
serviceability and deformability of the roof panels was considered. Special provisions in
the project specifications were used to ensure satisfactory surface preparation to achieve
adequate bond between the FRP laminates and the roof panels. Performance
specifications were also developed for the required mechanical and durability
characteristics of the FRP strengthening system rather than specifying the type of FRP
material.
385
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
ACI member Gamil Tadros, Ph.D., P.Eng. is ISIS Canada Technical Consultant. He
obtained his PhD from The University of Calgary in 1970. He is the recipient of the A.B.
Sanderson Award in 1999 from the CSCE. He was elected Fellow of the Canadian
Academy of Engineering in 2001. His research interests include bridge engineering and
the use ofFRP.
ACI member Aftab Mufti, Ph.D., P. Eng., is Professor of Civil and Geological
Engineering at the University of Manitoba, Winnipeg, Manitoba. He is also the Program
Leader and President ofiSIS Canada, a Network of Centres of Excellence. He is a Fellow
of the Canadian Society for Civil Engineering, the Engineering Institute of Canada, the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
American Society of Civil Engineers and the Royal Society for Arts, Manufacture and
Commerce and Fellow of the Canadian Academy of Engineering. His research interests
include finite element modelling, bridge engineering and the use ofFRP.
INTRODUCTION
The Olympic stadium "The Saddledome" shown in Figure 1 was built in 1984 as
a part of Calgary's preparation to host the Winter Olympic Games in 1988, The roof of
the circular plan stadium structure has a saddle shaped surface that could be described as
a hyperbolic paraboloid. The design notion of the Saddledome geometry is based on the
interaction of a sphere and a hyperbolic paraboloid. Architectural advantages of the 128
m diameter structure, include the ability to have the required number of seats within a
stadium that is 50 percent smaller in volume than the average stadium in North America.
Besides being aesthetically pleasing, the design also minimized energy consumption for
heating the structure during wintertime, which is a major concern in Canada.
The roof structure is made up of two groups of cables; sagging cables and
hogging cables having opposite curvature to form the hyperbolic paraboloid surface. The
roof cables are pretensioned and were laid to form a grid of 6.0 x 6.0 m panels in which a
drop-in lightweight precast concrete panels were used as shown in Figure 2. The use of
lightweight concrete reduced the total roof weight significantly and allowed the use of
The roof is anchored to a ring beam at the border following the meridian of the
sphere. The ring beam is supported on thirty-two columns. The roof is also constrained
at two ends by four stability A-shaped frames, which are designed to resist all horizontal
loads. The frames and the columns are anchored to the bedrock. The design was based on
allowing a ring beam movement of up to ± 150 mm with a rotation angle of 0.04 radians.
Thus, the outer columns do not restrain the ring beam movement horizontally and the
roof "floats" on the columns. The floating roof was made possible by utilising bearings
that permit both sliding and rotation at the same time. The bearings also have a
considerably large uplift capacity to resist wind suction and uplifts due to temperature
changes. The ring beam is the major structural element of the roof and is designed as a
pre-tensioned precast concrete element that is constrained only by the A-shape frames
(I).
STRENGTHENING REQUIREMENTS
During the last eighteen years, the Saddledome has performed very welL Stantec
Consulting Ltd was authorized to conduct a detailed analysis of the roof of the Canadian
Airlines Saddledome. This analysis was intended to assess the effects of adding a new
heavier clock and a press box extension to the building and to assess the remaining
capacity available for concert loads. The results of this study indicated that the roof could
accept the additional loads conditional upon the original designers' assumptions.
However, a detailed snow load investigation was recommended to confirm this
assessment. The results of this load investigation and a follow up reliability analysis
defined the revised loading conditions that needed to be considered when analysing the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Saddledome (2).
A finite element (FE) model of the roof system including the new loading
conditions showed that, under one specific snow/wind load combination, the roof panels
of the Saddledome would face demands that would be above the nominal roof capacity
according to the National Building Code of Canada (NBCC I995) (3). The model
incorporated the effects of the geometrical non-linearity of the roof structure using P-
delta analysis. Based on this analysis, it was determined that some roof panels required
strengthening. Areas of excessive loading are marked on the modelled floor plan
presented in Figure 3.
The flexural capacity of the middle ribs in the panels of concern needed to be
increased by up to 45 percent of its current capacity depending on the panel location.
The typical roof panel of the Saddledome is shown in Figure 4. The middle ribs are
marked in Figure 4 as elements that need strengthening. A summary of the strengthening
requirements of the roof panels is presented in Table I.
concrete, balanced failure of the section due to steel yielding while concrete crushing,
rupture of the FRP after steel yielding and debonding of the FRP at the interface were
considered. Table 2 presents the predicted increase in the flexural capacity of the roof
panel with the four design alternatives.
The required length of the FRP strengthening element was governed by the need
to provide adequate anchorage length to satisfy the interface shear demand at the end of
the beams (5, 6, 7 and 8). Many investigators (e.g. 9 and 10) have reported the
complexity of the shear transfer mechanism between the concrete beam and the FRP
composite. A proper model shall incorporate all factors affecting the longitudinal shear
stresses at the interface including the stress redistribution of the FRP plate over concrete
cracks and the shear step effect (11 and 12). Thus, the design guidelines as provided by
ISIS Canada (6), the ACI (7) and The CEB-FIP (8) for adequate end anchorage
requirements when using externally bonded FRP for strengthening of reinforced concrete
structures were carefully considered. Moreover, U shape FRP sheets were wrapped at the
ends and at one third points around the rib to enhance the anchorage of the FRP
sheets/plates in both design alternatives band c. Poulsen (13) reported that providing end
anchors can significantly enhance the structural performance of the flexure strengthened
element using FRP. Garden et al. (10) also showed the effectiveness of using U-shaped
(1)
where 'l'sE and 'l'u are the curvatures of the concrete elements at the service and ultimate
moments MsE and Mu respectively. The service moment is determined at a specific strain
level of the FRP reinforcement usually taken as 2000 x 10-6 strains (16 and l 7). The
Canadian Highway Bridge Design Code (CHBDC, 2000) (17 and 18) requires sections
reinforced with FRP to have a ductility performance factor higher than 4.0 (J > 4.0). As
no similar limits have yet been established by the building codes (e.g. Ref.l9 and 20), the
above limit was used to ensure ductile performance of the strengthened roof panels.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(2)
where Mn is the nominal strength of the roof slab, Mn is the moment due to self-weight,
Ms is the moment due to snow loads and Mw is the moment due to wind loads. a 0 , as
and aw are the load factors for the dead, snow and wind loads respectively. Values for
the load factors considered for nominal strength analysis are presented in Table 3. If
damage occurs in the FRP material in the event of fire, the overall factor of safety will be
reduced while it became evident that the roof panels will not collapse.
Very little research examined the effect of. creep on the behaviour of FRP
strengthened concrete elements (22 and 23). Garden and Hollaway (24) showed that the
creep in concrete beam strengthened with FRP plates is governed by the creep properties
of the concrete itself more than by the creep properties of the FRP material or the epoxy
adhesive. However, It has also been shown by Howard (24) that geometric: discontinuity
of polymer composites such as fibres ends and voids and the brittle nature of the epoxy
resin can induce microdamage in the FRP composite. This damage can result in non-
linear stress-dependent viscoelastic behaviour of the material even when lightly loaded
(24 and 25).
Thus, it was obvious that potential creep of the epoxy could disrupt the assumed
mechanism of stress transfer which can reduce the flexural capacity of the strengthened
roof panels. As it was evident that considerably high compressive stresses might exist due
to creep of the lightweight concrete panels, it was decided to experimentally examine the
effect of these compressive stresses on the performance of the FRP strengthening system .
under ultimate loading conditions.
Experimental Investigation
To investigate the creep effect on the strengthened beams, three full-scale T-
beams similar to that of the Saddledome roof panels were tested. The purpose of the
experiments was to examine the effect of the high level of compressive stresses on the
behviour of the FRP strengthened concrete beams. 30 MPa lightweight concrete similar
to that used in the Saddledome was used. The first beam was tested with no FRP
strengthening; the other two beams were tested after being strengthened with one CFRP
plate (alternative b) as shown in Figure 5-b). The second and third beams were subjected
to an axial force of 450 kN resulting in a average compressive stresses of (2.5 MPa)
before the FRP strengthening was applied. The test set-up is shown in Figure 6. The axial
compression load is applied at centroid of the concrete section to produce pure
compressive stresses. This was achieved by using a set of steel plates at each end and
four DYWIDAG bars. The load was sustained on the beam to provide the long-term
creep effect. The beams were tested in flexure loading by applying one concentrated load
at mid span. The first and the second beams were tested at 28 days of age while the third
beam was kept under sustained loads for six months. The results of the first two tests only
are presented here.
Figure 7 shows the load-deflection curves of the concrete beams with and
without the FRP. The results of the first and second tests showed that the use of the CFRP
plate increased the ultimate flexural load (Puu) by 29.5%, reduced the maximum mid-span
deflection (8max) by 9.9% and increased the ultimate moment capacity Mu 11 by 46.4%. The
experimental results are summarized in Table 4. The difference in the increase of the
ultimate flexural load from the increase of the ultimate moment is attributed to the non-
linear effect of the reduction of maximum deflection on the ultimate moment capacity of
the beams as shown in equation (3)
pull .S s:
Mult =--+N °max (3)
4
S is the loaded span and N is the 450 kN axial load representing the creep effect. The
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
ultimate moment capacity of both beams exceeded the loading requirements of the roof
panels in the Saddledome. Both beams with and without FRP failed in flexure due to
concrete crushing as the beam deflection reach its maximum value. Delamination of the
FRP of the second beam started after the beam reached its ultimate load. The first and the
second beams failed with a maximum deflection of 100 and 105 mm respectively.
Therefore, adding FRP did not affect the ductility of the beam while increasing its failure
load. The tests confirmed that the existence of relatively large compressive stresses in
Performance specifications
The use of FRP materials in strengthening necessitates the use of performance
specifications that concentrate on the required performance of the composite system
rather than specific materials or mixing content of the FRP materials. The performance
specifications required the FRP materials to achieve specific mechanical and durability
criteria. Moreover, the specifications emphasized quality control and quality assurance
requirements including staff qualifications, guidelines for preparation and performance
limits for the concrete substrate and for FRP system installation. Further emphasis was
directed to expected performance of the final product rather than construction procedures.
SUMMARY
This paper describes the strengthening of the roof panels of Calgary Saddledome
using CFRP materials. The special geometry and structural system of the Saddledome
created special constraints that were considered during design. Design alternatives of the
strengthening system were discussed. The structural design of the strengthening system
incorporated other factors to ensure adequate performance of the strengthened roof
panels. An important constraint is the creation of residual compressive stresses in the roof
panels due to creep of the ring beam. Full scale tests were carried out to confirm the
behaviour of the FRP strengthening system under residual compressive stresses. The
study proved those FRP materials provides an innovative and efficient alternative for
strengthening the roof panels of Calgary saddledome.
ACKNOWLEDGEMENTS
The authors would like to extend their thanks to Dr. Kenneth Neale, The
University of Sherbrooke for his sincere advice throughout the project. Special thanks for
the City of Calgary and the Saddledome foundation for their support.
NOTATION
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
REFERENCES
2 Maes, M. A., Risk-Consistent Design Checks for the Calgary Saddledome Roof
Tendons involving Live Load, Snow and Wind. Special Report to The City of
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Calgary, 2000.
5 Chajes, M. J., Finch, W. W., Januszka, T. F., Theodore, A. T. Jr., Bond and
Force Transfer of Composite-Material Plates Bonded to Concrete, ACI
Structural Journal, Vol. 93, No.2, 1996, pp. 208-217.
7 ACI Committee 440, Guide for the Design and Construction of Externally
Bonded FRP Systems for Strengthening Concrete Structures. ACI 440.2R-02,
2002, 45p.
10 Garden, H., Quantril, R. J., Hollaway, L. C., Thome, A.M. and Parke, G. A. R.,
An experimental study of the anchorage length of carbon fibre composite plates
used to strengthen reinforced concrete beams. Construction Building Materials.
Vol. 12, 1998, pp. 203-219.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Concrete Research, Vol. 54, No. 1, February 2002, pp. 47-59.
15 Jaeger, L. G., Mufti, A. A., Tadros, G. 1997. The concept of the overall
performance factor in rectangular-section reinforced concrete members.
Proceedings of the 3rd International Symposium on Non-Metallic (FRP)
Reinforcement for Concrete Structures, Sapporo, Japan, Vol. 2, pp. 551-559.
16 Newhook, J., Ghali, A., and Tadros, G, Concrete flexural members reinforced
with fiber reinforced polymer: design for cracking and deformability. Canadian
Journal ofCivil Engineering, Vol. 29, No.1, 2002, pp. 125-134. (17)
19 ACI Committee 318, Building Code Requirements for Structural Concrete (318-
99) and commentary (318R-99), American Concrete Institute, Farmington Hills,
MI, 1999.
21 Neville, A., Dilger, W. and Brooks, J. J., Creep of Plain and Structural Concrete,
First Edition, Construction Press, London, UK, 1983.
23 Teng, J. G., Chen, J. F., Smith, S. T. and Lam, L. Behaviour and Strength of
FRP-Strengthened RC structures: a state-of-the-art review. Proceedings of the
Institution ofCivi1 Engineers, Structures and Buildings, Vol. 156, Issue I, 2003,
pp: 51-62,
I .
T a bl e 3 L oa d f ac t ors use dfior nomma strength analysis
ao as aw
Case 1 1.0 1.0 0.6
Case 2 l.O 0.4 l.O
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Figure 2: Roof panels of Calgary saddledome during construction showing the sagging
and hogging cables
Figure 3: Finite element analysis of Calgary saddledome showing areas of excessive stresses
=
I
I
D
D 1r1
NEEDS j ND
•
STRENGTHENING i STRENGlll(NING
_. REOUIRtD.
Figure 4: Typical roof panels of the saddledome showing the structural elements that
needs strengthening
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
~r----c----~--------~---c----~---.
50
. --·---~f-
40 -------:---------rua=•tr----
~~~~·-
_____ j ____________ _
10
Figure 7: Load-deflection curves for the mid-rib of the roof panel with and without FRP
including the effect of creep
CHAPTERS
Synopsis: The earthquake of June 23, 2001, that affected most of the southern part of
Peru, put in evidence the seismic vulnerability of icons of the cultural heritage of the
country. The historical downtown of the city of Arequipa (located at 1000 km to the
South ofLima) was heavily affected by the earthquake, with forty percent of its
representative buildings suffering damage ranging from moderate to severe with partial
collapse.
The towers of the cathedral of Arequipa, built integrally with a volcanic stone called
sillar, suffered extensive damage. As a consequence, the left tower partially collapsed,
whereas, the right tower remained standing but in an unstable condition. This paper
describes the reinforcing strategy of the right tower with Carbon Fiber Reinforced
Polymer (CFRP) laminates, which were used to provide tensile strength and confinement
to the central stone core of the tower. After completing the CFRP installation, carved
stones were placed on top of the laminates to keep the original appearance.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
401
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
402 Tumialan et al.
Gustavo Tumialan, ACI member, is Senior Engineer at Simpson Gumpertz & Heger
Inc. in Boston, Massachusetts. He is active in the field of rehabilitation of masonry and
reinforced concrete structures. He is member of ACI Committee 440 and the Existing
Masonry Committee of The Masonry Society (TMS).
Antonio Nanni, FACI, is the V&M Jones Professor of Civil Engineering and Director of
the Center for Infrastructure Engineering Studies (CIES) at the University of Missouri-
Rolla. He is interested in construction materials, their structural performance, and field
application. He is an active member in the technical committees of ACI, ASCE, AS TM
and TMS.
INTRODUCTION
The cathedral church of Arequipa was first built in 1629, and it is the building
that symbolizes the character of the city and its inhabitants. Several reasons make the
cathedral the most representative building: it is the highest building in the urban
downtown, it is located in the heart of the city, it is built with sillar, and it is the silent
witness of important historical events. Finally, but not least, it is one of the most
important neo classical monuments of Peru.
The cathedral has a 100 meter long far,;ade and two towers at the top of the
building having a height of 28 m. The towers are the main feature of the cathedral,
representing an important visual reference in the urban and rural surroundings. Since the
construction of the cathedral, the towers have suffered damage and partial collapses due
to seismic events. The towers' slenderness made them vulnerable during earth quakes, so
common in the southern region of South America due to existence of the Nasca plate.
Thus, the 1868 earthquake caused the partial collapse of both towers which were then
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Due to their light weight, high tensile strength and ease of installation, fiber
reinforced polymer (FRP) composites were selected for the rehabilitation project of the
cathedral of Arequipa. Specifically, this paper describes the reinforcing strategy of the
masonry right tower of the cathedral with Carbon FRP (CFRP) laminates, which were
used to provide tensile strength and confinement to the central core of the tower. After
completing the CFRP installation, carved stones were placed on top of the laminates to
keep the original appearance.
Each tower is divided into three bodies: the first body comprises the pillars and
arches above the roof of the church, until the first friso. This body has a height of 6.5 m,
with each pillar having a cross section of 2.5 x 2.5 m, making this first body very rigid.
The second body comprises the transition zone from the first comisa until the base of the
upper arches, with an approximate height of 6. 7 m; the footprint in this region diminishes
from 7 x 7 m to 4.5 x 4.5 m The third body comprises the upper arches until the last
comisa, right before the light metal structure at the top. Due to its small weight and
stiffness, the metal structure at the top is not relevant. The approximate height of the third
body was 5.1 m. Figure 1 illustrates a tower scheme.
The bottom pillars at the base of each tower were the elements that took the
overturning moment generated by the seismic event. During the earthquake, the ir
behavior was acceptable; only some small cracks were observed at their base. Due to
their slenderness ratio and location at the top of the cathedral's main body, the two towers
are structurally considered as appendices of the main structure. This impl ies that their
dynamic behavior is almost independent of that of the main structure. Figure 2 shows the
shaking of the towers during the earthquake of 2001.
Due to the change of cross section and stiffness the portion of the tower starting
from the upper half of the second body behaved also, to some extent, as an appendage of
the lower tower part. This caused a high concentration of stresses at the base of the
second body as well as at the base of the third body because of the change of stiffness.
The overcoming of the tensile and compressive strengths of the unreinforced masonry led
to the dislodging and crushing of the sillar blocks in the second body causing ultimately
the collapse of the third body of the left tower because of the decrease of the sup port area
provided by the second body (see Figure 3a). This tower portion fell as a rigid body on
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
From the post-event field observations and structural analysis, it was evident that
the weakest part of the tower was the second body or transition zone. In addition the
proneness to failure was worsened because of cumulative damage from past earthquakes,
which was not properly repaired. Prior to the 2001 earthquake, the towers went through
reinforcing interventions in different years. Details of these interventions are presented
elsewhere (1 ).
STRUCTURAL INTERVENTION
Due to the different damage, the scope of the structural intervention had to be
different for each tower. The left tower suffered the collapse of the third and part of the
second body; thus, the complete reconstruction of these two bodies was required.
The right tower suffered extensive damage in the second body, losing part of its
mass, and finally staying in an unstable equilibrium condition. For this reason, an internal
steel structure was placed in order to stabilize it for possible aftershocks. Details are
presented elsewhere (1). It was decided to leave the internal frame in place so the
dismantling of the second and third body of the tower was not necessary. The proposed
structural intervention consisted of placing additional reinforced concrete (RC) columns
at the bottom of the tower, next to the pillars, and the rebuilding of the missing part in the
second body. The second body was reinforced with CFRP laminates, which were used to
provide tensile strength and confinement to that section of the tower.
Structural Analysis
The geometry of the tower was modeled with a commercially available CAD
program and the frequencies and vibration modes were computed using a commercially
available finite element program. As a reference, Figure 4 shows the finite element
model. The sillar block, due to its volcanic origin is a lightweight stone, with a unit
weight in the order of 12.5 kN/m 3 . Thus, the total weight of each tower was estimated to
be approximately 4.25 MN.
No records were available from the earthquake in the city of Arequipa itself, but
its duration was estimated as more than one minute. The only record available was one
located at 150 km to the west of the city which showed a peak horizontal acceleration of
0.14g. The dynamic behavior of the tower was considered to be almost independent of
that of the main structure but with a larger dynamic amplification. Thereby, the dynamic
analysis of the towers was performed independently of the main structure and with a
fixed base as a boundary condition. A static analysis was also performed with horizontal
forces equivalent to 30, 40 and 50 percent of the weight. The results of the analyses
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
By inspection, the axial demand was determined to be much lower than the
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
nominal factored axial load. Thus, it was decided to analyze the member as a tubular
member having an outer diameter of 4.20 m and an inner diameter of 2.60 m., subj ected
only to a bending moment equal to 575 KN -m. Due to the nature of seismic forces,
CFRP reinforcement was required to be placed vertically along the circumference of the
second body. Alternatively, interaction diagrams for different configurations of CFRP
reinforcement could have been developed.
Tests were not performed to determine the material characteristics of the existing
sillar masonry. Instead, a compressive strength, f~ , equal to 4.8 MPa was considered,
which was reported by an investigation on a similar masonry typology (5).
To provide continuity in the section of the second body where a change of cross
section and stiffness existed, it was required to anchor the CFRP reinforcement. As part
of the overall rehabilitation program an RC ring beam was planned to be built in that
region to provide confinement. Thus, it was decided to anchor the FRP reinforcement
To restrain the lateral expansion of the second body under axial loads caused by
an earthquake, it was determined to place CFRP laminates in the hoop direction. The
splice length was specified to be 150 mm which was staggered along the height to
prevent having a weak area prone to debonding of the laminates.
Reinforcing Process
The cross section of the second body was restored using sillar blocks (see
Figures 6.a and b). This activity provided with a unique opportunity to train young
artisans in the practice of carving and laying sillar blocks. This new acquired ability will
be useful for future interventions in the historical downtown of the city. In the process of
gathering information for the project, it was determined that the mortar had undergone
changes from the early to present times. Originally, only sand and lime were used in the
mix, together with milk and egg shells, compared to present times when sand, lime and
cement mix were used. For the construction of the masonry walls, a mortar with
volumetric proportion of 1:5 (cement : sand) was used.
After completing the new circular wall whythe, the gap between this wall and
the existing masonry was filled with a mix of crushed sillar, sand and cement. Prior to
installing the CFRP laminates, the excess mortar in the joints was eliminated using a
grinder, and the uneven surface was leveled with a putty material (see Figures 7.a and .b).
Given the masonry surface characteristics, there was no need for sandblasting. In fact, in
sillar masonry, the blocks typically exhibit a rough and porous surface.
The CFRP laminates were installed using a manual lay -up technique following
the procedure recommended by the manufacturer (3). Due to the porosity of the masonry
surface, the higher viscosity saturant was applied directly to the masonry surface instead
of the primer conventionally used in concrete members. The high absorption
characteristics of sillar masonry would have required a large amount of the low-viscosity
primer.
The longitudinal CFRP strips were anchored around a ring beam in the region of
transition zone. The laminates were installed from top to bottom (see Figure 8.a) and
adhered to the surface of a small ditch where the ring beam was going to be cast (see
Figure 8.b). After the installation of the vertical reinforcement was completed, CFRP
laminates were placed in the hoop direction (see Figures 9.a and .b). Finally, a steel cage
was placed in the ditch and concrete was poured (see Figure 10). A new masonry wall
was built with carved sillar blocks on top of the laminates to keep the original appearance
of the tower. The gap between the two walls was also filled with a mix of crushed sillar,
sand and cement.
CONCLUDING REMARKS
This project provided a remarkable opportunity for combining traditional materials with
new technology. The use of carbon FRP reinforcement, externally bonded to the surface
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
ACKNOWLEDGEMENTS
The support of Master Builders Technologies - Latin America, Cleveland, Ohio, and
Union de Concreteras (UNICON), Lima, Peru is gratefully acknowledged.
REFERENCES
(1) Torrealva, D., Blanco, A., Tumialan, G., and Nanni, A., (2003), "Seismic
Rehabilitation of Cathedral Towers in Peru," Eighth International Conference on
Structural Studies, Repairs and Maintenance of Heritage Architecture 2003,
Halkidiki, Greece, 7 - 9 May 2003, 10 pp.
(4) American Concrete Institute - Committee 440, (2002), "Guide for the Design and
Construction of Externally Bonded FRP Systems. for Strengthening Concrete
Structures," ACI 440.2R-02, September, 2002.
(5) Lara Galindo, J.C. (1988), "Ensayos de Albaiiileria de Sillar" Bachelor of Science
Thesis, Department of Civil Engineering, Pontific ia Universidad Cat61ica del Peru,
Lima, Peru, 1988.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Third
Body
Second
Body
First
Body
Church
Body
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
·1 152
~?lA'
•3.11.1
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
"'"'
~"111
--tJQZ
~7(19~
~0.1
...... ·t-Gtl
~HIV1
~11 (S
I
I I / .
I / /
I/ /
I;~ / - -
_._._-- -
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
by J. G. Tumialan, N. Galati, A. Nanni, and D. Tyler
Synopsis: Fiber reinforced polymer (FRP) composites in the form oflaminates or bars
have been proven to be effective for the strengthening of unreinforced masonry (URM)
walls subjected to overstresses. Two installation techniques have been proposed:
externally-bonded FRP laminates (i.e. manual lay-up or adhesion of pre-cured laminates)
and near surface mounted (NSM) FRP bars. The latter technique consists of placing a
bar in a groove cut into the surface of the member being strengthened. This paper
presents a field application on flexural strengthening with NSM FRP bars of two cracked
URM walls in an educational facility. Design considerations and the results of an
experimental program conducted to validate the strengthening are described.
413
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com
l
414 Tumialanetal.
Gustavo Tumialan, ACI member, is Senior Engineer at Simpson Gumpertz & Heger Inc.
in Boston, Massachusetts. He is active in the field of rehabilitation of masonry and
reinforced concrete structures. He is member of ACI - Committee 440 and the Existing
Masonry Committee of The Masonry Society (TMS).
Nestore Galati, a doctoral student in Composite Materials for Civil Engineering at the
University of Leece, Italy where he received his B.Sc. in Materials Engineering. He
obtained his M.Sc. degree in Engineering Mechanics at the University of Missouri-Rolla.
His research interests include repair of masonry and reinforced concrete structures. He is
an EIT in the United States.
Antonio Nanni, FACI, is the V&M Jones Professor of Civil Engineering and Director of
the Center for Infrastructure Engineering Studies (CIES) at the University of Missouri -
Rolla. He is interested in construction materials, their structural performance, and field
application. He is an active member in the technical committees of ACI, ASCE, ASTM
and TMS.
Dan Tyler, is the Branch Manager for Western Waterproofing Co., Inc. in Anaheim,
California. He is a graduate of Bethany College in Lindsborg, Kansas and has a B.A. in
Economics. He has been a part of the Western Group Companies for 10 years and in
three branches; Houston, Kansas City, and Anaheim. The Western Group is member of
ACI, ASTM, and ICRI.
INTRODUCTION
Fiber reinforced polymer (FRP) composites in the form of laminates or bars can
provide viable solutions for the strengthening of unreinforced masonry (URM) walls
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Installation of FRP externally -bonded laminates applied by wet manual lay -up
or adhesion of precured laminates has been the most commonly used method. More
recently, the technique of near surface mounted (NSM) FRP bars has been adopted. This
technique consists of placing a bar in a groove cut into the surface of the member being
strengthened. The FRP bar is embedded in either an epoxy or cementitious -based paste
which transfers stresses between the substrate and the bar. The successful use of NSM
FRP bars for the strengthening of concrete members (2) has been extended to URM walls,
one of the building components most prone to failure due to overstressing. Thus, FRP
bars have been also successfully used for increasing the in -plane and out-of-plane
capacity, and provide pseudo-ductility to URM walls (3), (4), (5).
Two URM walls at the Grandview High School in Kansas City, Missouri
exhibited cracking at the bed joints at the mid-height region (see Figures la and lb). The
cracks had an approximate width of 6.00 mm for which they could be classified as
moderate (8). The masonry walls were built with standard concrete masonry units with
nominal dimensions 200 x 200 x 400 mm. The height of the walls was 3.00 m. and the
length was 10 m. The problem was caused by the inward movement of an unstable
exterior wall, which displaced the open web steel joists resting on it. As a consequence,
the joists pushed against the URM wall (Figure lc) leading to its cracking. Before
proceeding with the strengthening, the exterior wall was stabilized by steel tiebacks and
the cracks were injected. The use of NSM FRP bars was proposed to reinstate the
integrity of the cracked masonry walls. The design approach consisted of restoring the
flexural capacity of the cracked walls to that of the original.uncracked walls.
The design protocol followed an allowable stress design (ASD) method. The
protocol was in compliance with the recommendations provided by ACI 440.1R-Ol (6)
and MSJC (7), even though these two documents do not address the adopted
strengthening technique. The following considerations can be taken into account to
conduct the design of the strengthening:
• The allowable compressive stress due to flexure only in masonry is estimated
as F;, = J3' f~ (7). f~ is the compressive strength of masonry.
According to the ASD approach, the capacity of the strengthened wall can be
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
ff = e!Ef (2c)
n=E/em (2d)
where:
A! area ofFRP
bm wall strip width
df effective depth of the FRP reinforcement
Ef modulus of elasticity of FRP
Jb calculated compressive stress in masonry due to flexure only
jj calculated tensile stress in FRP
n ratio of the modulus of elasticity of FRP to the modulus of elasticity of
masonry
strain in FRP
strain in masonry
The coefficient 'k' can be obtained from the well-known relationship, where PJ is the
amount of FRP reinforcement:
(3)
(4)
EXPERIMENTAL VALIDATION
Five masonry specimens were built with concrete masonry blocks, and
strengthened with glass FRP (GFRP) bars to prove the efficiency of the FRP
reinforcement for improving the flexural capacity of URM walls as well as to ve rify some
of the design considerations. Their dimensions were 610 mm wide by 1220 mm high. The
wall thickness was about 90 mm. In that manner the specimens had a slenderness ratio
height-thickness, hj t equal to 14, similar to the hj t ratio in the walls at the Grandview
High School. The compressive strength of masonry, f~, was 10.5 MPa. The specimens
were strengthened with 6.25 rum-diameter (#2 size in US customary) and 9.5 mm-
The strengthening layout represented URM wall strips with GFRP bars placed at
different spacings. The parameters investigated were the bar size and the amount of
reinforcement. Table 1 presents the test matrix with specimen designation that includes
two numbers denoting the size of the bar and the number of bars used for the
strengthening. Thus, W2-3 corresponds to an URM wall, strengthened with three 9.5
rum-diameter GFRP bars.
In general, two types of failure were observed: FRP de bonding and shear failure
in the masonry near the support (see Figure 3), initial cracking was delayed and the crack
widths were thinner as the amount of FRP reinforcement increased. The results, in terms
of ultimate load, maximum bending moment and type of failure are summarized in Table
1. Details of the experimental procedure and results are presented elsewhere (5).
Figure 4 shows the moment vs. mid-span deflection diagrams for the five
strengthened walls. The flexural strength and stiffness of the FRP strengthened walls
increased as the amount of reinforcement increased. Based on recommendations by
MSJC, the nominal moment for the URM concrete specimen was estimated as 0.23 kN-
m. Although none of the presented specimens was strengthened with GFRP bars in the
vertical joints (this is only possible in a stack bond pattern), the flexural capacity of a
strengthened wall, when this is the case, does not decrease as compared to a wall built in
a running bond pattern (11).
Figure 5 illustrates the experimental moment vs. strain diagrams for only two of
the wall specimens. Despite this limitation, it was possible to observe in W3 -1 that for a
strain of 0.20 Efu• associated to F1 = 0.2Q.ffu, the experimental allowable moment for this
FRP-controlled section was approximately equal to 0. 70 kN -m which is very close to the
theoretical allowable moment of 0.74 kN-m, calculated using Equation 4. In the
specimen W3-3, masonry-controlled section, for a theoretical allowable moment of 1.27
kN-m, the strain in the GFRP bar would be approximately 0.10% which represents a
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
stress of0.10.!fu·
Figure 6 illustrates the plot of the experimental service moment vs. analytical
allowable moment. The experimental service moments for each specimen were
quantified from Figure 4 for wall deflections equal to 0.007 h (12), being h the effective
height of the wall (i.e. test span). The allowable moments were estimated by following
the protocol previously described considering the governing condition, either FRP -
controlled (i.e. F1 = 0.20./fu is reached first) or masonry-controlled (i.e. Fb = }j' f~ is
FIELD APPLICATION
Material Properties
A compressive strength, f~, equal to 12.4 MPa was considered for concrete
masonry based on common practice. A mortar cement type M or S was assumed to carry
out the calculations. This type of mortar would provide the largest flexural demand to be
met with the strengthening. GFRP bars with a diameter of 6.25 mm were selected for the
strengthening of the URM walls. These bars have a guaranteed tensile strength, J; ,of
827 MPa and modulus of elasticity, E 1 , of 40.68 MPa (9).
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Design Process
The design approach consisted of restoring the flexural capacity of the cracked
walls to that of an uncracked wall. Thus, the cracking moment, Mer. in the URM wall
was computed considering the MSJC provisions (7). The amount of FRP reinforcement
was designed based on this moment. The first trial consisted of the analysis of a masonry
section having a width, bm, of 100 mm. (or bar spacing of 100 mm.)
• Computation of Mer:
For a mortar cement type M or Sand hollow ungrouted units the flexural strength, J,.,
is 0.431 MPa (7). These conditions provide the flexural demand to be met with the
strengthening. Considering the nominal thickness of the wall to be equal to 200 mm,
and that the wall was built with face shell mortar bedding (moment of inertia, ), the
Mer was estimated as:
p1 =A1 { d =0.00165
/bm f
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
The modulus of elasticity of masonry was computed as:
Em =900/~ ==11.16GPa (1618ksi)
The calculated compressive stress in masonry was calculated from Equation 2a:
r
Jb = 2Aff%
kd b = 09
. 5MPa < F, = 4.13MPa
f m
STRENGTHENING PROCEDURE
The strengthening procedure used to install the GFRP bars can be summarized
as:
(1) Slots having a width of approximately one and a half times the bar diameter were
grooved in the vertical mortar joint, then the surface was cleaned to remove dust.
(2) An epoxy-based embedding paste was applied in the groove (see Figure 8a). Strips of
masking tape were attached at each edge of the groove to avoid staining of the
masonry.
(3) The GFRP bars were encapsulated in the groove (see Figures 8b and 8c), and
(4) Surface of the joint was finished and tooled (see Figure 8d).
The strengthening of the walls was completed in two days with minimal
disruption to the building occupants. In addition, the aesthetics of the classrooms was not
altered since the reinforcement was placed in the vertical joints.
CONCLUSIONS
Laboratory tests of URM walls showed that the strength and pseudo-ductility
can be substantially increased by using NSM FRP bars. The field application confirmed
the ease of installation observed in the laboratory environment and showed low
disturbance and minimal aesthetic impact ofNSM FRP bars showing promising potential
for the retrofitting of existing masonry structures.
ACKNOWLEDGEMENTS
REFERENCES
(1) Nanni, A. and Tumialan J.G., (2002), "Use ofFRP Composites for the Strengthening
of Masonry Structures," Submitted to the Structural Engineering International
Journal - International Association for Bridge and Structural Engineering (IABSE),
January, 2003.
(2) De Lorenzis, L., Nanni, A., and La Tegola, A., (2000), "Flexural and Shear
Strengthening of Reinforced Concrete Structures with Near Surface Mounted FRP
Rods," 3rd Int. Conf. on Advanced Composite Mat. in Bridges and Structures,
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
(3) Tumialan, J. G., Morbin, A., Nanni, A. and Modena, C., (2001), "Shear
Strengthening of Masonry Walls with FRP Composites," COMPOSITES 2001
Convention and Trade Show, Composites Fabricators Assoc., Tampa, FL, October 3-
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
6, 6 pp. CD-ROM.
(4) Tumialan, J. G., San Bartolome, A., and Nanni, A., (2003), "Strengthening ofURM
Infill Walls by FRP Structural Repointing," 9 1h N. Amer. Masonry Conf., Clemson,
South Carolina, June 1-4, 2003, 12 pp. CD-ROM.
(5) Galati, N., Tumialan, J.G., La Tegola, A., and Nanni, A., (2003), "Out-of-Plane
Behavior of URM Walls Strengthened with FRP Systems," Advancing with
Composites 2003, "Plast 2003", Milan, Italy, May 6-10,2003.
(6) American Concrete Institute - Committee 440, (2001), "Guide for the Design and
Construction of Concrete Reinforced with FRP Bars", ACI 440.1R-Ol, American
Concrete Institute, Farmington Hills, MI, 2001.
(7) Masonry Standards Joint Committee, (2002), "Building Code Requirements for
Masonry Structures," ACI-530-02/ASCE 5-02/TMS 402-02, Detroit, MI, New York,
NY, and Boulder, CO.
(8) Rainer, P., (1983), "Movement Control in the Fabric of Buildings," Nichols
Publishing Co., New York, NY.
(9) Hughes Brothers, Inc., (2002), "Technical Specifications for As Ian GFRP Rebar,"
Seward, NE.
(11) Secondin, S., (2003), "Masonry Reinforced with FRP systems," Tesi di Laurea,
Universita degli Studi di Padova, Facolta di Ingegneria, Padova, Italy, March 2003.
(12) International Building Code- IBC (2000). BOCA, ICBO, SBCCI, Country Club
Hills, IL, Whittier, CA, and Birmingham, AL.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Notes:
FRP-controlled
!Fl
IMlMasonl)'-controlled
'Based on Deflection of0.007h
D: Debonding
S-S: Sliding Shear
F-S: Flexural Shear
bm Ms Pt n Pt k kdt It /b
(mm) (kN-mm) (mm) (MPa) (MPa)
CLASSROOM
MUSIC ROOM
·· 1~;-----o· ~~--zr---r.-
. e,
..
A,
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
5.
OM-------L---~~--~---+----~~--~~~--~~
0 5 10 15 20 25 30
Deflection (mm)
Figure 4. Moment vs. Deflection Curves
0~-----+~ --~~~~-+~~--~--~--~----~
0.00 0.20 0.40 0.60 0.80 1.00 1.20
Design Strain, £ru (%)
Figure 5. Experimental Moment vs. Strain
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
0 I 2 3 4
Analytical Allowable Moment (kN-m)
Figure 6. Service Moment vs. Allowable Moment
Dabardlam.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Length
inch millimeter (mm) 25.4Et
foot meter (m) 0.3048E
yard meter(m) 0.9144E
mile (statute) kilometer (km) 1.609
Area
square inch square centimeter (cm 2) 6.451
Volume (capacity)
ounce cubic centimeter (cm 3 ) 29.57
3 0.003785
gallon cubic meter (m )t
cubic inch cubic centimeter (cm 3 ) 16.4
Force
kilogram-force newton (N) 9.807
kip-force newton (N) 4448
pound-force newton (N) 4.448
Pressure or stress
(force per area)
kilogram-force/square meter pascal (Pa) 9.807
kip-force/square inch (ksi) megapascal (MPa) 6.895
newton/square meter (N/m 2) pascal (Pa) I.OOOE
pound-force/square foot pascal (Pa) 47.88
pound-force/square inch (psi) kilopascal (kPa) 6.895
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Temperature§
deg Fahrenheit (F) deg Celsius (C) tc = (IF - 32)/1 .8
deg Celsius (C) deg Fahrenheit (F) tp = 1.8tc + 32
* This selected list gives practical conversion factors of units found in concrete technology. The reference
source for information on SI units and more exact conversion factors is "Standard for Metric Practice" ASTM E
380. Symbols of metric units are given in parentheses.
t E indicates that the factor given is exact.
:j: One liter (cubic decimeter) equals 0.001 m 3 or 1000 cm 3 .
§ These equations convert one temperature reading to another and include the necessary scale corrections. To
convert a difference in temperature from Fahrenheit to Celsius degrees, divide by 1.8 only, i.e., a change from 70
to 88 F represents a change of 18 F or 18/1 .8 = I0 C.
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
Index
L R
Lees, J. M., 55 Reda Taha, M., 385
Lesko, J., 181 rehabilitation, 3, 361
lighting columns, 55 reinforced concrete, 121, 203, 219, 279
lightweight concrete, 91 reinforced polymer (CFRP) laminates, 401
load testing, 121,219,349 reinforcement, 139, 401
repair, 279
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
--`,```,,``,,,`,,,``````,,,,```,-`-`,,`,,`,`,,`---
®
international
SP-215
Copyright American Concrete Institute
Provided by IHS under license with ACI Licensee=UNI OF NEW SOUTH WALES/9996758001
No reproduction or networking permitted without license from IHS Not for Resale, 08/10/2015 01:23:41 MDT
Daneshlink.com