Citation
An experimental investigation into the behavior of concrete elements retrofitted with NSM composite strips at elevated temperatures

Material Information

Title:
An experimental investigation into the behavior of concrete elements retrofitted with NSM composite strips at elevated temperatures
Creator:
Namrou, Abdul Rahman ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (183 pages). : ;

Subjects

Subjects / Keywords:
Fiber cement -- Testing ( lcsh )
Fiber-reinforced concrete ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Near-surface-mounted (NSM) fiber reinforced polymer (FRP) is another strengthening alternative of externally bonded fiber reinforced polymers. NSM FRP is a promising alternative technology that has emerged for enhancing the strength capacity of concrete structures. Most laboratory researchers have focused mainly on the overall member performance and/or the bonding performance of the NSM bars or strips. Limited research has focused on the effect of temperature exposure on NSM FRP performance. The results of an experimental program performed on forty-eight (48) concrete block specimen with NSM carbon-fiber reinforced polymer (CFRP) strengthening systems at elevated temperatures that reaches to 200 degrees Centigrade [392 degrees Fahrenheit] to investigate flexural performance. The effect of using two different adhesive systems (epoxy anchoring system) with manufacturer recommendation at ordinary and high temperature exposures is also studied. The adhesive was injected in a NSM groove size (25 mm [1 in] deep x 13 mm [0.5 in] wide) the width and depth of the groove were greater than 3 and 1.5 times the CFRP thickness and width, respectively. Test results show that the interfacial strength of the specimens bonded with the ordinary epoxy is maintained until 75 degrees Centigrade [167 degrees Fahrenheit] is reached, while the strength noticeably decreases with an increasing temperature above this limit. The specimens with the high-temperature epoxy preserve interfacial capacity up to 200 degrees Centigrade [392 degrees Fahrenheit] despite a trend of strength-decrease being observed. The failure of the test specimens is brittle irrespective of adhesive type. Interfacial damage is localized along the bond-line with the presence of hairline cracks that further develop when interfacial failure is imminent. This thesis also presents an experimental result concerning the bond performance of concrete-adhesive at elevated temperatures that reaches to 200 degrees Centigrade [392 degrees Fahrenheit] applied for three hours. Then, the concrete prisms were tested under three point flexural loading. The experimental program is comprised of seventy-two (72) specimens bonded with low viscosity, high viscosity adhesives and high-temperature adhesive and their comparative performance is of interest in the present investigation. Emphasis is placed on the residual capacity of the conditioned bond-concrete interface and corresponding failure mode. For high temperature exposure, it is shown that the high temperature laminated adhesive outperforms the high and low viscosity adhesives by remaining fairly consistent and allowing the strengthening system to remain effective for up; to three house of 200 degrees Centigrade [392 degrees Fahrenheit]
Thesis:
Thesis (M.S.)--University of Colorado Denver. Civil engineering
Bibliography:
Includes bibliographic references.
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Abdul Rahman Namrou.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
892343932 ( OCLC )
ocn892343932

Downloads

This item has the following downloads:


Full Text
AN EXPERIMENTAL INVESTIGATION INTO THE BEHAVIOR OF CONCRETE
ELEMENTS REROFITTED WITH NSM COMPOSITE STRIPS AT ELEVATED
TEMPERATURES
By
ABDUL RAHMAN NAMROU
B.S., North Dakota State University, 2011
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
2013


2013
ABDUL RAHMAN NAMROU
ALL RIGHTS RESERVED


This thesis for the Master of Science degree by
Abdul Rahman Namrou
has been approved for the
Civil Engineering Program
By
Yail Jimmy Kim, Chair
Kevin Rens
Frederick Rutz
November 12, 2013


Namrou, Abdul Rahman. (M.S., Civil Engineering)
An Experimental Investigation into the Behavior of Concrete Elements Retrofitted with
NSM Composite Strips at Elevated Temperature
Thesis directed by Associate Professor Dr. Yail Jimmy Kim
ABSTRACT
Near-surface-mounted (NSM) fiber reinforced polymer (FRP) is another strengthening
alternative of externally bonded fiber reinforced polymers. NSM FRP is a promising
alternative technology that has emerged for enhancing the strength capacity of concrete
structures. Most laboratory researches have focused mainly on the overall member
performance and/or the bonding performance of the NSM bars or strips. Limited research
has focused on the effect of temperature exposure on NSM FRP performance. The results
of an experimental program performed on forty-eight (48) concrete block specimen with
NSM carbon-fiber reinforced polymer (CFRP) strengthening systems at elevated
temperatures that reaches to 200C [392F] to investigate flexural performance. The
effect of using two different adhesive systems (epoxy anchoring system) with
manufacturer recommendation at ordinary and high temperature exposures is also studied.
The adhesive was injected in a NSM groove size (25 mm [1 in] deep x 13 mm [0.5 in]
wide) the width and depth of the groove were greater than 3 and 1.5 times the CFRP
thickness and width, respectively. Test results show that the interfacial strength of the
specimens bonded with the ordinary epoxy is maintained until 75C [167F] is reached,
while the strength noticeably decreases with an increasing temperature above this limit.
The specimens with the high-temperature epoxy preserve interfacial capacity up to 200C


[392F] despite a trend of strength-decrease being observed. The failure of the test
specimens is brittle irrespective of adhesive type. Interfacial damage is localized along
the bond-line with the presence of hairline cracks that further develop when interfacial
failure is imminent.
This thesis also presents an experimental result concerning the bond performance of
concrete-adhesive at elevated temperatures that reaches to 200C [392F] applied for
three hours. Then, the concrete prisms were tested under three point flexural loading. The
experimental program is comprised of seventy-two (72) specimens bonded with low
viscosity, high viscosity adhesives and high-temperature adhesive and their comparative
performance is of interest in the present investigation. Emphasis is placed on the residual
capacity of the conditioned bond-concrete interface and corresponding failure mode. For
high temperature exposure, it is shown that the high temperature laminated adhesive
outperforms the high and low viscosity adhesives by remaining fairly consistent and
allowing the strengthening system to remain effective for up to three hours of 200C
[392F],
The form and content of this abstract are approved. I recommend its publication.
Approved: Jimmy Kim
IV


ACKNOWLEDGEMENTS
This thesis would not have been possible without the support of so many people in many
forms. It was the product of a large measure of serendipity with people who have changes
the course of my life. I truly believe that every person I come across with have some
effect on my life, be it the most apparently insignificant way. On behalf of this, I must
also acknowledge all those of whom I have come across with and have led me to the path
that I chose. Some have had much larger roles than others.
First, I must take to opportunity to acknowledge and thank my parents for their endless
love and support. They have implanted in me high work ethics and confidence to prosper.
I would like to thank my sisters for standing by my parents when I am gone.
With this believe in mind, It is with immense gratitude that I acknowledge the deepest
appreciation to my advisor, Associate Professor Yail Jimmy Kim, for his patience, vision
and guidance to pursue my goals. Without his encouragement, I would never have
considered a Masters degree in the first place. Also, I would like to acknowledge the
department chair and committee chair of civil engineering Dr. Kevin Rens for generous
support and making the transition from North Dakota State University easy. With much
respect, I would like to thank Dr. Khalil Jarrar, J.D. for his moral support; I learned from
him that there is no limit for educational content.
v


This thesis work was supported by the United States Department of Transportations
through the Mountain Plains Consortium Program and the University of Colorado.
vi


TABLE OF CONTENTS
Chapter
1. Introduction................................................................15
1.1 General...................................................................15
1.2 Research Significance.....................................................19
1.3 Scope.....................................................................20
1.4 Thesis Organization.......................................................22
2. Literature Review...........................................................24
2.1 Near-Surface-Mounted FRP Strengthening....................................24
2.1.1 FRP Techniques...........................................................24
2.1.2 FRP Applications.........................................................25
2.2 Background of Near-Surface-Mounted Techniques............................25
2.2.1 FRP Failure Modes........................................................27
2.2.2 Bonding Failure..........................................................27
2.2.3 Strengthened Capacity of the Element.....................................28
2.2.4 NSM Groove Spacing and Depth Recommendation..............................29
2.2.5 Interfacial Behavior of NSM CFRP Bonded to a Concrete Substrate..........30
2.2.6 Failure Modes of NSM Bonded Specimens....................................31
2.3 High Temperature Effect...................................................32
2.3.1 Effect of Elevated Temperatures on CFRP Materials........................32
2.3.2 Behavior of Concrete Members Strengthened with NSM CFRP at High
Temperature....................................................................33
2.4 Performance-Based Fire Safety Design......................................34
2.4.1 History of Performance-Based Codes.......................................34
vii


2.4.2 Advantages and Disadvantages of Prescriptive-Based Design................36
2.4.3 Advantages and Disadvantages of Performance-Based Design.................37
2.4.4 Development and Evaluation of Design Fire................................39
2.4.5 Importance Factor Approach...............................................39
2.5 Conclusion...............................................................40
3. Behavior of Near-Surface-Mounted CFRP-Concrete Interface in Thermal Distress ....48
3.1 Fabrication and Specimen Design............................................48
3.1.1 General Overview.........................................................48
3.1.2 Preparation..............................................................49
3.1.3 Materials................................................................50
3.1.4 Test Specimen............................................................51
3.1.5 Thermal Exposure.........................................................52
3.1.6 Test Setup and Instrumentation...........................................53
3.2 Material-Level Testing....................................................53
3.2.1 Dynamic Mechanical Analysis..............................................53
3.2.2 Concrete Testing under Thermal Distress..................................54
3.2.3 CFRP Strip Testing under Thermal Distress................................54
3.2.4 Ordinary Epoxy and High-Temperature Coupons..............................55
3.3 Experimental Results.......................................................55
3.3.1 Analysis of Variance for Concrete at Elevated Temperatures...............55
3.3.2 Interfacial Strength of NSM CFRP.........................................56
3.3.3 Load-Displacement Response...............................................58
viii
3.3.4 Failure Mode
59


3.4 Summary and Conclusion......................................................60
4. Bond Performance of Concrete-Adhesive Interface at Elevated Temperatures.....77
4.1 Fabrication and Specimen Design.............................................77
4.1.1 General Overview...........................................................77
4.1.2 Concrete and Adhesive Material Properties..................................78
4.1.3 Testing Specimen and Thermal Exposure......................................79
4.1.4 Instrumentation............................................................80
4.2 Material-Level Testing......................................................81
4.2.1 Stress-strain Response of Adhesive Resin...................................81
4.2.2 Storage Modulus of Adhesive Resin..........................................82
4.2.3 Frequency Test.............................................................82
4.3 Experimental Results.......................................................83
4.3.1 Dynamic Elastic Modulus of Concrete Prisms................................83
4.3.2 Load Displacement Response of the Bond Test Concrete Prisms...............84
4.3.3 Failure Mode of Bonded Concrete Prisms.....................................86
4.4 Summary and Conclusion......................................................87
5. Structural Concrete Performance under Fire...................................105
5.1 Introduction...............................................................105
5.2 Importance Factor Evaluation Approach......................................106
5.2.1 Importance Factor Calculation.............................................107
5.2.2 Different Type Contribution to the Overall Influence Factor...............109
5.3 Importance Factor Analysis for NSM CFRP Concrete Interface Interaction......113
5.4 Importance Factor Analysis for the Bond Performance Concrete-adhesive Interfacel 14
IX


5.5 Utilization of the Importance Factor into Performance-based Approach..116
5.6 Conclusion............................................................118
6. Conclusion and Recommendation..........................................123
6.1 Recommendation for Future Work........................................127
References................................................................128
Appendix
A.........................................................................128
B.........................................................................152
C ........................................................................170
D.........................................................................172
E ........................................................................173
F.........................................................................181
x


LIST OF TABLES
Table
2.1. FRP Tensile Strength and Youngs Modulus (Adapted from ACI 440.2R-08)........42
2.2. Bridge features weightage factors (Kodur and Naser 2013).....................43
2.3. Risk grades w.r.t the overall class coefficient and the importance factor (Kodur and
Naser 2013).......................................................................44
3.1. Properties of CFRP strip from the manufacturer (Hughes Brothers 2008)........62
3.2. Test Specimens load & failure mode of ordinary epoxy & high temperature epoxy .63
3.3. Compressive concrete strength depending upon temperature (Kim et al. 2012)...64
4.1. Properties of low-viscosity adhesive from the manufacturer (MBrace, 2007)....88
4.2. Temperature dependent stress response of adhesive coupons: (HV) high viscosity
adhesive coupons; (LV) low viscosity adhesive coupons; (HT) high temperature adhesive
coupons...........................................................................88
4.3. Dynamic elastic modulus response over the specified temperature..............89
4.4a. Temperature dependent adhesive-concrete interfacial capacity of high viscosity
adhesive..........................................................................92
4.4b. Temperature dependent adhesive-concrete interfacial capacity of low viscosity
adhesive..........................................................................93
4.4c. Temperature dependent adhesive-concrete interfacial capacity of high temperature
adhesive..........................................................................94
5.1. Parameters used to predict the overall class coefficient for NSM CFRP concrete
blocks...........................................................................120
5.2. Parameters used to predict the overall class coefficient for bond performance
concrete prisms..................................................................121
5.3. Importance factor table criteria based on Kodur and Naser (2013) propose....121
xi


LIST OF FIGURES
Figure
2.1. Pre-cured system aramid-type FRP installation (Carmichael and Barnes 2005).44
2.2. Groove dimension limitations (ACI 440.2R-08, 2008).........................44
2.3. Development of Design Fires (Custer 1995)..................................45
2.4. Evaluation of Alternative Designs (Custer 1995)............................46
2.5. The influencing characteristics of fire hazard in bridges (Kodur and Naser 2013)....46
2.6. The contribution to overall influence factors from different classes (Kodur and Naser
2013)...........................................................................47
3.1. Wooden mold in which concrete blocks were formed...........................65
3.2. NSM CFRP-concrete interface test specimen dimension (not to scale).........65
3.3. Interface test specimens after applying an epoxy adhesive..................65
3.4. Test specimen: (a) air-blasting before CFRP-bonding; (b) CFRP installation.66
3.5. Test setup:(a) high temperature exposure;(b) conditioned specimen;(c) tension test.66
3.6. Thermocouple temperature readings over a three hour range acquired from the data
logger: (a) at 50C [122F]; (b) at 75C [167F]; (c) at 100C [212F]; (d) at 125C
[257F]; (e) at 150C [302F]; (f) at 175C [347F]; (g) at200C [392F]........68
3.7. Dynamic Mechanical Analysis (DMA); (a) testing machine; (b) DMA clamps; (c)
DMA specimens...................................................................68
3.8. Dynamic Mechanical Analysis (DMA) results: (a) CFRP strip; (b) high temperature;
(c) ordinary epoxy..............................................................69
3.9. Temperature-dependent strength measured: (a) concrete in compression; (b) CFRP
strip in tension; (c) ordinary epoxy in tension; (d) high-temperature epoxy in tension; (e)
normalized strength of OE; (f) normalized strength of HE .......................70
3.10 Failure mode: (a) HE epoxy coupons at 200C [392F]; (b) OE epoxy coupons at
200C |392F|; (c) CFRP strips at 200C [392F]................................71
Xll


3.11. Temperature-dependent interfacial strength: (a) specimens bonded with ordinary
epoxy; (b) specimens bonded with high-temperature epoxy; (c) temperature-dependent
average interfacial strength.................................................71
3.12. Load-displacement of specimens bonded with ordinary epoxy: (a) at 25C [77F];
(b) at 50C [122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C [257F]; (f)
at 150C [302F]; (g) at 175C |347I'|: (h) at 200C [392F]......................73
3.13. Load-displacement of specimens bonded with high-temperature epoxy: (a) at 25C
[77F]; (b) at 50C [122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C
[257F]; (f) at 150C [302F]; (g) at 175C [347F]; (h) at200C [392F]...........75
3.14. Load-displacement comparison of exposure temperatures: (a) bonded with ordinary
epoxy; (b) bonded with high temperature epoxy...........................................75
3.15. Failure mode: (a) interface failure of ordinary epoxy; (b) interfacial failure of
ordinary epoxy after test; (c) interfacial failure of high-temperature epoxy after test.75
3.16. Failure mode: (a) ordinary epoxy at 50C; (b) ordinary epoxy at 200C; (c) high-
temperature epoxy at 50C; (d) high-temperature epoxy at 200C..........................76
4.1. Concrete beam mold being prepared for concrete casting.............................95
4.2. Adhesive coupon preparation and casting: (a) mold shape; (b) high-viscosity epoxy;
(c) low-viscosity epoxy.........................................................95
4.3. Adhesive coupon: (a) during testing; (b) before and after thermal exposure.95
4.4. Temperature dependent residual capacity of adhesive coupons: (a) high viscosity
adhesive coupons; (b) low viscosity adhesive coupons; (c) high temperature adhesive
coupons.........................................................................96
4.5. Test set-up: (a) specimen details (not to scale); (b) test set-up configuration using the
MTS machine......................................................................96
4.6. Temperature data-logger.....................................................97
4.7. Temperature recording acquired from the data logger: (a) at 50C [122F]; (b) at
75C [167F]; (c) at 100C [212F]; (d) at 125C [257F]; (e) at 150C [302F]; (f) at
175C [347F]; (g) at 200C [392F]; (h) combined temperatures..................99
4.8. Electric stationary diamond saw.............................................99
4.9. DMA results of adhesive: (a) low viscosity; (b) high viscosity; (c) high temperature
100
xm


4.10. Frequency test: (a) frequency test set up; (b) frequency test reading........100
4.11. Residual impact resonant frequency of interface test specimen (dot = individual;
line = average): (a) frequency response; (b) temperature-dependent dynamic elastic
modulus.............................................................................101
4.12. High viscosity adhesive concrete prism specimens: (line = individual; dot =
average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity.........101
4.13. Low viscosity adhesive concrete prism specimens: (line = individual; dot =
average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity.........102
4.14. High temperature adhesive concrete prism specimens: (line = individual; dot =
average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity.........103
4.15 Normalized bond performance concrete prism specimens capacity..................103
4.16 Failure mode of concrete prisms bonded with different types of adhesive: left
specimen (LV); central specimen (HV); right specimen (HT): (a) at 50C [122F]; (b) at
200C 1392 I' |....................................................................104
5.1. Overall class coefficient to assign fire risk grade of NSM CFRP concrete blocks ..122
5.2. Importance factor based on the overall class coefficient of NSM CFRP concrete
interface: (a) for HE epoxy; (b) for OE epoxy.......................................122
5.3. Overall class coefficient to assign fire risk grade of bond performance concrete-
adhesive interface...................................................................123
5.4. Importance factor based on the overall class coefficient of bond performance
adhesive concrete interface: (a) for HT epoxy; (b) for LV epoxy; (c) for HV epoxy...123
xiv


1. Introduction
1.1 General
Structural strengthening is frequently required for upgrading the performance of
deteriorated concrete structures. Externally-bonded carbon fiber reinforced polymer
(CFRP) sheets have been broadly used, while an alternative technique called the near-
surface mounted (NSM) method is emerging (De Lorenzis and Teng 2007). The NSM
method employs CFRP strips or rods inserted into the narrow precut groove of a concrete
member and permanently bonded with an adhesive. As is for the case of the externally-
bonded CFRP sheet application, epoxy adhesives are dominantly used for bonding NSM
CFRP reinforcement. NSM strengthening techniques may require more initial costs than
conventional externally-bonded CFRP sheet application (Aidoo et al. 2006); however,
their superior bond and durability can provide better long-term performance. A number
of laboratory-scale projects have been carried out to examine the effect of NSM CFRP on
enhancing structural behavior in flexure (El-Hacha et al. 2004; Hassan and Rizkalla
2004) and in shear (Chaallal et al. 2011). Field demonstration has been undertaken as
well. Alkhrdaji et al. (1999) performed a strengthening project using a decommissioned
reinforced concrete slab bridge with multiple spans. CFRP-strengthening was intended to
increase the flexural capacity of the strengthened span by 30% in comparison to that of
an unstrengthened span. Test results revealed that both strength and stiffness of the bridge
were improved due to the presence of NSM CFRP. Stone et al. (2002) strengthened a
simply supported solid slab bridge. Prior to installing NSM CFRP strips, a non-
15


destructive test was done to examine the distribution of internal reinforcing steel. Load
tests using a truck were conducted before and after the strengthening work to assess the
efficacy of the NSM CFRP.
Fire is a latent problem for all constructed structural members because thermal stress can
significantly degrade their functionality. Such an extreme event has been an issue for
residential or office buildings over several decades. Given infrastructure systems are
concerned with transporting hazard materials (e.g., flammable or combustible
commodities), bridge structures are exposed to potential fire hazards (Garlock et al. 2012).
Fire resistance of existing members needs to be ensured so that catastrophic events can be
precluded. This crucial requirement is valid for CFRP-strengthened structures. According
to a literature search (to be discussed in the Background section), the rehabilitation
community is aware of the significance of thermally-induced detrimental effects
associated with CFRP application. Extensive effort has been done to investigate the
behavior of CFRP-strengthened concrete structures subjected to high temperature. The
majority of research is dedicated to concrete members with externally bonded CFRP
sheets (Foster and Bisby 2008). The application of NSM CFRP in this circumstance is
limitedly studied (Palmieri et al. 2012). All existing research programs are based on
CFRP composites in conjunction with epoxy adhesives that are fundamentally susceptible
to the degree of thermal exposure. The interaction between the NSM CFRP-concrete and
adhesive-concrete in a thermal environment is another area to be explored, provided no
research has been reported earlier. This thesis addresses these identified technical
challenges and presents an experimental program concerning the interfacial behavior of
16


NSM CFRP strips embedded in a concrete substrate and bonding adhesive to concrete
subjected to elevated temperatures. Foci of the study are bond deterioration, residual
capacity, and failure characteristics. A pioneering endeavor is made to examine the
thermal performance of NSM CFRP-concrete interface bonded with a high-temperature
epoxy that can overcome the limitation of ordinary epoxy adhesives. The results
discussed here are taken from a test program that is devoted to understanding the
fundamental thermomechanical behavior of NSM CFRP for strengthening concrete
structures in thermally-induced stress states. Based on the 2013 ASCE Report Card on
Americas Infrastructure (ASCE, 2013), the overall infrastructure grade of Americas
GPA is D+. Being more concise, one-ninth of North America bridges are classified as
structurally deficient with a C+ grade with an average age of nations 607,380 bridges is
currently 42 years and that an estimate investment of $20.5 billion per year for the next
15 years.
Over the past 20 years, the civil engineering applications has developed the use of fiber
reinforced polymers (FRPs), and FRPs are now providing a number of innovative
approaches for both new construction, and predominantly for rehabilitation and
strengthening of existing structures (Bakis C.E., 2002). The carbon FRP (CFRP)
composite is widely preferred for retrofit applications over the glass or aramid FRP
(AFRP GFRP). The advantages of CFRP composites consists of resistance to corrosion,
high strength and stiffness, easy and prompt installation, reduced service costs, low
density, and sustainable performance (Teng J.G., ICE 2003) (Kim YJ a. H., 2008). The
achievement of the stress transfer is provided by the CFRP load-carrying capacity and the
17


matrix transfer the load to the fibers. CFRP composites may be bonded externally to
improve the load-carrying capacity of the structure. According to Malek AM et al. and
Kim YJ. et al., premature debonding failure due to mechanical or environmental loading
may occur on the externally bonded CFRP due to stress concentration at the cut-of point
(Malek AM, 1998) (Kim YJ F. A.-FL, 2005). Debonding of CFRP is influenced by the
normal and shear stresses along the bondline (Smith ST, 2001). Environmental loading
such as wet-drying cycles, freezing and thawing, and low temperature can deteriorate the
concrete bonding interface (Green M.F., 2000).
With the above-mentioned drawbacks of externally-bonded FRP as a strengthening and
retrofit of deteriorated structural concrete, near-surface-mounted (NSM) FRP
reinforcement has emerged as a promising alternative technology in both shear and
flexure. Along the tensile soffit of the concrete structure, a small groove (slot) is cut and a
CFRP composite may be strips, laminates, and bars are inserted and bonded with an
adhesive resin that may be epoxy or other cementitious grouts. The NSM CFRP has
improved bond characteristics and enhanced durability compared to externally bonded
CFRP composites. A greater portion of full strength bonded CFRP is often able to utilize
through this method because of the greater bonding characteristics that prevent premature
debonding failures. Moreover, because the CFRP strengthening system is located within
the member itself, protects it from the environment and fire damage.
18


1.2 Research Significance
A large research effort with the current emergence strengthening technique of NSM
CFRP for reinforced concrete members has focused on strength gain with use of
composite strips, failure mode, interfacial stresses between concrete and CFRP at room
temperature in both shear and flexural strengthening applications. Minimal research has
been reported over the durability issues of the NSM FRP strengthening techniques for
concrete. Due to the potential hazard of fire, a challenge in CFRP strengthening
applications may take place. The performance of the adhesives resins used to bond the
CFRP to the structure may be vulnerable to high temperature due to a characteristics
change. (Calberger T., 2009). The critical temperature of an adhesive is called glass
transient temperature, Tg. The effectiveness of CFRP-strengthening is dependent upon the
temperature exposure level. Although limited research regarding the fire safety of CFRP
composite strip concrete strengthening have been recently published (Bisby L.A., 2005),
there still exists inadequate information in this area. The promising strengthening method
requires more experimental and analytical investigations. High temperature provides
detrimental stress to CFRP-strengthened concrete members because the strengthening
system includes polymeric materials that are radically susceptible to thermal exposure.
Although NSM CFRP strips are a strong alternative to traditional externally-bonded
CFRP sheets when upgrading constructed concrete structures, their thermal response is
not sufficiently elucidated in the research community. This research program emphasizes
the interfacial behavior of the NSM CFRP embedded in a concrete substrate subjected to
elevated temperatures, including a comparative study as to the performance of ordinary
and high-temperature epoxy adhesives.
19


This thesis presents an on-going research program concerning the performance of NSM
CFRP concrete members performance subjected to high temperature. A brief review on
literature relevant to the research is also provided. The experimental study reported here
examines the material characteristics and bond performance of NSM CFRP-concrete
interface and the adhesive bond-concrete interface at elevated temperatures.
The objectives of this research were:
To investigate experimentally the performance of NSM CFRP strengthening
systems bonded to concrete with different type of adhesive subjected to high
temperature exposure for three hours up to 200C [392F], As might realistically
bridges and existing structures allow for evacuation to take place and to evaluate
the potential behavior of these strengthening techniques under the effect of fire.
To investigate experimentally the adhesive bond-concrete interface subjected to
elevated temperatures.
To investigate experimentally the mechanical properties and characterization of
different type of adhesives for the use of bonding NSM FRP strengthening
applications under the same condition of high temperature exposure.
To propose a performance-based design equations for structures under the effect
of fire.
1.3 Scope
The performed work within this thesis involved experimental testing of forty-eight (48)
plain concrete blocks made with the same concrete mix design to eliminate strength
20


discrepancies each dimensioned as following (200 mm [8 in] long x 100 mm [4 in]
deepx 100 mm [4 in] wide) having a groove of size (25 mm [1 in] deepx 13 mm [0.5 in]
wide) along the strengthening direction molded where the CFRP strip is inserted and
bonded with the different type of adhesives designated as ordinary epoxy known for its
industrial name as Power Fasteners T308+ (Power Fastener, 2011), a two components
epoxy adhesive anchoring system; and the other as high temperature laminated resin
(PTM&W Industries 2013), a two-part system epoxy. The concrete blocks then oven
cured over a temperature range from 25C [77F] to 200C [392F] for three hours.
The other study presented in the thesis is the bond performance of adhesive-concrete
interface. The performed work was done by molding seventy-two (72) concrete prisms
each dimensioned as following (400 mm [16 in] longx 100 mm [4 in] widex 80 mm
[3.15 in] deep). The beams were split in half during the molding process by a wooden
sheet plate. After proper concrete curing, the specimens were bonded together with the
same types of high viscosity adhesive known for its industrial name as Power Fasteners
T308+ (Power Fastener, 2011), a two components epoxy adhesive anchoring system; low
viscosity adhesive known as BASF Mbrace (Mbrace 2007); and high temperature
adhesive laminated resin (PTM&W Industries 2013), a two-part system epoxy. The
concrete prisms were cured as the same conditions up to 200C [392F] for three hours.
A material property of the CFRP strips, adhesive epoxies, and concrete were determined
over the temperature range mentioned earlier.
21


1.4 Thesis Organization
Literature review on NSM FRP behavior in strengthening applications is presented in
chapter 2. The properties, description of some existing structures, and proposed
application models in addition to some design guidelines are discussed in this chapter.
An overview of the experimental program of concrete blocks with NSM CFRP is
discussed in chapter 3. Also, a detailed behavior of NSM CFRP-concrete interface in
thermal distress along with the test scheme and material level. Chapter 4 presents the
bond performance of concrete-adhesive Interface at elevated temperatures. Additional
mechanical properties testing of different types of adhesives are also tested at high
temperatures.
Chapter 5 discusses the design recommendation of the fire performance-based
methodology structures along with the typical objectives of the performance-based
specifications. Chapter 6 provides a summary from this experimental study and some
recommendations for further experiments in this research area based on the evaluation of
the results from the testing program.
Appendix A presents additional detailed pictures of the process for the NSM CFRP
concrete block testing that have been done in which you can see the failure mode.
Appendix B presents the bond performance of the concrete-adhesive prisms in which you
can see the failure mode from the pictures.
22


Appendix C shows the dynamic mechanical analysis test specimens and set up.
Appendix D shows the frequency tests performed on concrete prisms and the dynamic
elastic modulus in pictures.
Appendix E shows the three types of adhesive epoxy coupon testing in pictures along
with the failure mode and stress-strain figures.
Appendix F shows the CFRP testing in pictures.
23


2. Literature Review
2.1 Near-Surface-Mounted FRP Strengthening
2.1.1 FRP Techniques
The necessity of structural upgrade has been required to increase the capacity of many
existing structures. Because FRP is a strong and light material, it has been extensively
used in structural rehabilitation industry. FRP is a strengthening techniques of concrete
structures that started with externally-bonded systems, where sheets that could be fibers
in one-direction or multi-directions are bonded to the externally to the surface of the
concrete. Typical, there are two methods of strengthening techniques where 1) FRP
sheets are saturated on site along with the resin bonded to the concrete surface and called
wet-overlaid, 2) FRP sheets are saturated and cured before site application and then
applied to concrete surface with adhesive, this method is called pre-cured system. Figure
2.1 shows the procured FRP installation of aramid fibers type along the soffit of the
structure. With the externally-bonded application of FRP, premature debonding failure
has often occurred due to concentration of stress at the cut-of point (Kim YJ F. A.-H.,
2005). Researchers have been focused on NSM strengthening technique as an improved
method of strengthening existing structures.
24


2.1.2 FRP Applications
The application of FRP to existing structures can be divided into two main categories:
externally bonded FRP systems and near-surface-mounted NSM systems. The composite
of FRP does not vary and typically involves one fiber type regardless of the application
type. The types of FRPs available in the market are Carbon (C), glass (G), or aramid (A)
FRPs (ACI440 2008) with carbon is the most commonly used.
Typically, FRP materials have higher ultimate tensile strength relatively to steel yield, the
strength of 420 MPa [60 ksi] and ultimate strength of 520 MPa [75 ksi], however, steel is
more elastic that FRPs with a typical value of 200 GPa [29,000 ksi] (Bertolotti, Eric A.
2012). As a result, the FRP failure rapture is considered to be brittle failure in tension
with an approximate strain of 1-3%; by comparison, whereas steel ruptures at
approximately 30% (Vasquez 2008). The typical values of tensile strength and Youngs
modulus of different types of FRP are shown in table 2.1.
2.2 Background of Near-Surface-Mounted Techniques
Structural strengthening with NSM reinforcement is one of the most promising
techniques, where the original method was found in the literature dating to the mid of
1940s (De Lorenzis et al. 2000). Steel rods or bars were used as reinforcement and
cement mortar was used as an adhesive (Burke 2008). Nowadays, FRP bars or strips take
place of the steel rods or bars strengthening techniques and adhesive resin replaces the
25


cementitious bonding. The concept of bonding FRPs to a concrete element is simply by
inserting FRP rod or strip into a pre-cut groove concrete opening and bonding it with
adhesive resin. The revised technique of NSM FRP has been used in several applications
in the late 1990s that includes reinforced concrete silos strengthening in Boston, Myriad
Convention Center in Oklahoma City, and deck strengthening of the Naval Pier in San
Diego (De Lorenzis et al. 2000). The results have shown higher level of strengthening
efficiency, less mechanical damages and aging effects, and less susceptibility for fire
damage (Taljsten et al. 2003). The full testing experiment was done on a full scale bridge
in Missouri, that took over the FRP external bonded sheets system by 10% with respect to
strength gain (Alkhrdaji et al. 1999) (Bruke 2008). A lot of researchers have focused ever
since on the strength gain, groove characteristics and depth, FRP shape, adhesive types
and bonding characteristics. In addition to that, researchers consider the overall
performance of the structural element and the performance-based bond.
To increase the load carrying capacity NSM strips or bars can be installed in a parallel
grooves that are cut at a specified distance. The FRP bars are available as deformed steel
with US corss-sectional customary size area as number 3 or 4, while the FRP strips are
available in two typical dimensions of 16 mm [0.63 in] wide and 2 to 4.5 mm [0.079 to
0.177 in] thick (Hughes Brothers 2010).
26


2.2.1 FRP Failure Modes
At the meantime, there is no building code requirement that addresses the FRP
strengthening systems specifically, however, there is design recommendation for the
applications of NSM FRP concrete reinforcing systems. The available literature of the
NSM FRP failure modes has focused on the three failure categories, debonding resistance
failure, the strengthened capacity of the element, and the depth of the groove within the
concrete. If adequate spacing of NSM FRP strips or bars was used, it will result in
individual failure planes. If not adequate spacing distance of NSM FRP strip or bars was
used, it will result in a singular failure plane and of course decreases the overall load
carrying capacity of the structure.
2.2.2 Bonding Failure
To be able to create composite action, the adhesive resin comes into play to transfer the
stress between the FRP bar or strip reinforcement and the concrete susbstrate (De
Lorenzis et al. 2001). Most researchers represent the failure mode of the tensile testing by
the beam pull-out. El-Hacha and Rizkalla (2004) found from eight T-beam testing that
the FRP has little impact on the deflection due to loading before cracking (elastic range)
but improved significantly after cracking. The authors also found that the NSM CFRP
strips resulted in higher capacity than NSM CFRP bars where he explains that this
behavior is due to premature debonding of the NSM CFRP bars.
27


Taljsten and Nordin (2007) also found from their eight six-meter-long beam testing
experiments that the NSM reinforcement did not contribute to the stiffness of the overall
beam in the elastic range but after cracking the beam shows stiffer behavior and
significant increase in the load-carrying capacity.
Nine T-beams were tested by Hassan et al. (2004) and found that premature debonding
between the concrete-adhesive interface failure happened in the specimens with bond
lengths of 150 and 250 mm [5.9 and 9.8 in] with no improvement in strength in
comparison to the control specimen. On the other hand, strength and stiffness gains were
observed with bonding length of 500 to 700 mm [19.7 to 27.6 in] (Bruke 2008).
Seracino et al (2007) conducted thirty-six (36) push pull tests of concrete blocks bonded
to NSM FRP strips restrained by adhesive inside a middle groove. From this analysis, the
researchers conducted a non-linear regression analysis to derive some equations needed
to determine the intermediate cracking debonding forces that according to the authors, it
is controlled by the dimensions of the NSM FRP strip and the concrete compressive
strength. Seracino et al. (2007) recommended 200 mm [7.9 in] as a minimum bonded
length for the FRP.
2.2.3 Strengthened Capacity of the Element
Rectangular beams were tested by Barros et al. (2005) in four-point bending with
different NSM FRP and steel reinforcement. In most samples, failure occurred in the
28


concrete cover spalling that includes the tension steel reinforcement and the NSM FRP
strip. In some cases, concrete was pulled away from the internal steel. From this research,
the authors notices strain increase from 62% to 91% of the ultimate tensile strain and a
strength increase of 78% to 96% in comparison to the control beams. They also noted that
the stiffness and cracking moment increased in the strengthened beams.
Teng et al. (2006) conducted similar testing scheme of concrete beams in four-point
bending like Barros et al. (2005) but with different NSM CFRP bonding lengths. Stiffer
behavior and enhanced strength gains of about 30% and 90% to beams with bonded
lengths of 1200 mm to 1800 mm [47.2 to 70.9 in] respectively.
2.2.4 NSM Groove Spacing and Depth Recommendation
According to Blaschko (2003) spalling of concrete corner may occur if the FRP
reinforcement was placed close to the edge of the concrete element. The ACI Committee
440 (2008) stated the clear spacing of NSM groove strips should be two times the groove
depth, and four times the grove depth from the side cover. Hassan and Rizkalla (2004)
stated that the clear spacing of the NSM groove strips is two times the FRP bar diameter,
and four times the FRP bar diameter from the side cover. Kang et al. (2005) specified that
40 mm [1.6 in] is the clear spacing of the NSM groove strips and the side cover distance.
Rashid, Oehlers and Seracino (2008) detailed that 53 mm [2.1 in] is the clear spacing of
NSM groove strips and 3.5 times the FRP depth with respect to the side cover distance.
29


ACI 440.2R-08 (2008) section 13.3 sets a limit for NSM minimum groove dimension.
The depth of the groove should be greater or equal to 1.5 times the diameter for FRP bars
or maximum length of the FRP strip. The width of the groove is also greater or equal to
1.5 times diameter of the FRP bars. For FRP strips the groove width is greater or equal of
3 times the thickness of the strip. More bonded surface will develop between the adhesive
resin and the concrete when the depth of the FRP embedment increases in the concrete
element as shown in Fig. 2.2. In other words higher debonding strains and greater
ductility will be developed (Oehlers et al. 2008).
2.2.5 Interfacial Behavior of NSM CFRP Bonded to a Concrete Substrate
The interfacial behavior of NSM CFRP embedded in a concrete substrate has been of
interest for over a decade. Laboratory investigations were conducted using beam
specimens (El-Hacha et al. 2004; Hassan and Rizkalla 2004; Sena Cruz et al. 2004;
Barros et al. 2005) and isolated block elements for a pull-out test (Teng et al. 2006;
Seracino et al. 2007). CFRP-debonding frequently occurs at geometric discontinuities of
a strengthened beam due to stress concentrations (e.g., flexural cracks and CFRP cut-off
points). Several failure modes are available for NSM CFRP systems: FRP rupture,
adhesive-concrete interfacial fracture, and adhesive splitting (El-Hacha et al. 2004; Teng
et al. 2006; Seracino et al. 2007). Bond strength of CFRP strips is generally greater than
that of CFRP bars because of their geometric configuration and corresponding stress
distribution (Blaschko 2003). Bond length of NSM CFRP is an important factor
controlling the efficacy of a strengthening system. An insufficient bond length can cause
30


premature debonding of the CFRP (Hassan and Rizkalla 2004). Other parameters can
also affect the behavior of strengthened beams such as concrete strength, CFRP aspect
ratio, and groove size (Teng et al. 2006; Seracino et al. 2007). The embedment depth of
CFRP strips appears to be related to a pseudo-confining effect that may delay splitting
failure of the strips (Seracino et al. 2007). Long-term performance of NSM CFRP-
concrete interface has been studied in some experimental programs. Sena Cruz et al.
(2006) tested a midspan-hinged reinforced concrete specimen strengthened with NSM
CFRP in monotonic and cyclic loads. A decrease in stiffness was noticed when the
specimens were subjected to cyclic loading in comparison to the case under monotonic
loading. The peak pull-out force of NSM CFRP, however, was not influenced by cyclic
load. Badawi and Soudki (2009) tested reinforced concrete beams strengthened with
NSM CFRP in high-cycle fatigue over one-million sinusoidal loading. Bond of NSM
CFRP was satisfactory, thereby preserving ductility of the strengthened beams. The
endurance limit of these fatigue beams was improved up to 24% when relative to that of
an unstrengthened control beam.
2.2.6 Failure Modes of NSM Bonded Specimens
Based on the conducted literature review, the failure modes of NSM bonded specimens
are limited to concrete crushing; adhesive splitting due to high concentrated stresses
between the FRP and the adhesive resin interface; concrete splitting due to relatively no
enough restrained on the concrete element or low tensile strength of concrete; and FRP
31


rapture which is a rare case and only happen if there is sufficient restrained that
delay/prevent the debonding or concrete splitting or crushing to occurs.
2.3 High Temperature Effect
2.3.1 Effect of Elevated Temperatures on CFRP Materials
CFRP composites demonstrate different thermomechanical properties compared to
conventional structural materials such as steel reinforcement. For unidirectional CFRP,
the coefficient of thermal expansion in the transverse direction is higher than that in the
longitudinal direction because the resin matrix expands more than the fibers do when
heated. CFRP can exhibit a negative coefficient of thermal expansion (MBrace 2007).
Thermally-induced distress in a concrete structure having CFRP materials may thus be of
technical concern (Gentry and Hussain 1999). Glass transition temperature is one of the
important parameters for the constituents of a CFRP composite. This critical temperature
is defined as a temperature beyond which morphological changes in the polymeric resin
of CFRP take place. It is worth noting that carbon fibers can withstand over 1,000C
[1,832F] (Rostasy 1992), while most commercially available resins have a glass
transition temperature of below 100C [212F], Thermomechanical properties of a
polymeric resin rapidly decrease when the temperature applied exceeds its glass
transition temperature (Dimitrienko 1999). The effect of elevated temperatures on the
mechanical behavior of carbon fibers is not conclusive. Dimitrienko (1999) reported that
32


the strength and stiffness of carbon fibers tended to increase with temperature, while the
opposite was presented by Sumida et al. (2001).
2.3.2 Behavior of Concrete Members Strengthened with NSM CFRP at High
Temperature
Although elevated temperatures are one of the critical parameters for constructed
reinforced concrete members (e.g., fire), limited research has been done on the thermal
resistance of NSM CFRP. Palmieri et al. (2012) tested NSM CFRP-strengthened
reinforced concrete beams in fire associated with a service load. The range of thermal
exposure varied from room temperature to 900C [1,652F] for 2 hours. Fire protection
systems were added to the strengthened beams using glass fiber cement and calcium
silicate boards. Some beams exhibited bond failure of the CFRP during the fire test,
evidenced by a sudden increase in deflection of the strengthened beams. Residual
strength tests showed that the strengthening system maintained sufficient interfacial bond
as long as the temperature applied was lower than 200C [392F], Burke et al. (2013)
showed test data concerning one-way slabs strengthened with NSM CFRP exposed to
elevated temperatures. The strengthened slabs showed a noticeable decrease in load-
carrying capacity at a temperature of 100C [212F], including failure of the adhesive-
concrete interface. Kodur and Yu (2013) developed a modeling approach to predict the
behavior of reinforced concrete beams strengthened with NSM CFRP subjected to a fire.
Temperature-dependent material properties were taken into consideration, including
bond-deterioration of the CFRP caused by thermal stress. The effect of temperature
33


altered the failure mode of these beams from concrete crushing to CFRP rupture. The
location of CFRP was found to be a contributing factor to fire resistance of the
strengthened beams. The need for adequate insulation was discussed.
2.4 Performance-Based Fire Safety Design
Prevention of heat, smoke, evacuation, and rescue, etc. in addition to the structural
stability and integrity both locally and globally are the recent approaches for fire safety
performance-based design. The intent, structural use, studying of fire scenarios over the
structural life-time, ensuring the soundness of the design, and durability defines the level
of performance-based fire design. Even though the exposure of heat, high temperatures
and gases are non-structural, they play an important factor to ensure the robustness and
the functionality of the structure to allow for adequate rescue time. With this entire in
mind, advanced simulation and computational methods are used in quantitative
assessment on structural fire resistance (Liew 2002).
2.4.1 History of Performance-Based Codes
Researchers and scholars have been publishing many documents on performance-based
design approach. This drive was initiated by the fire protection advancement in research
and technology, prediction of fire risks and the complex restriction with the prescriptive
based design (Hdjisophocieous et al. 1998). During the 1970s, Nelson (1972) applied
goal oriented designs to fire safety on some federal buildings. Wehrili et. al (1972),
34


introduced new terms closely related to performance-based codes. The author also
standardized checklists to evaluate the fire safety designs for hospital bedrooms.
Haviland (1978) discussed with the fire safety community the unconformity fire safety
levels, the confusion between the codes and the applications, and the less focused on the
residences with known low levels on fire safety (Hdjisophocieous et al. 1998). Havilands
(1978) performance-based objectives were to protect the public welfare from the fire.
Boring et al. (1981) discussed the need to update the codes to take advantage of the new
emerging technology and include the performance measures. He was worried about the
inapplicable of the restrictive codes. Boring et al. (1981) matched the safety of public
welfare that Haviland discussed but added the ability of the structure to preserve its
functionality. The Conseil International du Batiment (CIB, 1982) described the
performance guidelines and criteria approach in a report issued in 1982. The report
detailed that the prescriptive requirements are not economical and costly despite the
simplicity of working with this approach. The International Standard ISO 6241 (1984)
provides the guidelines for preparing performance standard approach in buildings. A
research was conducted in Japan by Wakamatsu (1988) where he outlines the design
method for fire safety building evaluation. In his outline, he stated that performance-
based codes substitute the existing prescriptive codes though establishing equivalency.
The author motivation was the problems behind implementing the prescriptive codes that
are less efficient, difficult to execute with the available technology, and difficult to
understand the level of fire safety (Hdjisophocieous et al. 1998).
35


Ferguson (1993) considered the prescriptive codes requirements may be simple to apply
with less flexibility to allow innovation. The author states that the rewritten building
regulations in 1985 offered more guidance and alternative solutions using fire safety
based on simple qualitative requirements (Hdjisophocieous et al. 1998). In 2001, the
Canadian Commission on Building and Fire Codes (CCBFC, 1994) unveil a draft that
developed a fully objective based code. Meacham and Custer (1995) provided a summary
based on the up-to-date performance-based approaches for fire safety design. The authors
state that there is a need for more guides related to fire engineering design
(Hdjisophocieous et al. 1998). On the other hand, Meacham and Custer (1995) stated that
the performance-based design was based on specific cases with respect to the entire
fire/building interaction (Hdjisophocieous et al. 1998). The main objectives of fire safety
performance-based design is to protect the building occupants from fire, retard its growth
with minimum fire impacts that support the fire-fighting operations, and to reduce the
injuries if any structural loss occurs (Bukowski and Babrauskas 1994) (New Zealand
Building Codes 1994).
2.4.2 Advantages and Disadvantages of Prescriptive-Based Design
Fire safety design codes can be prescriptive type or performance-based type. Currently,
most codes use prescriptive codes or combination of both prescriptive and performance-
based codes. The disadvantages of prescriptive-based method are that it does not clearly
identify the factor against failure, because there is no consideration of the strength
interaction and stability between the elements and the overall structural system (Kim et
36


al.). Despite the popularity and limitation of the prescriptive based design methods when
dealing with less complex requirements, many countries around the world are making the
shift between the prescriptive based to performance-based design (Hdjisophocieous et al.).
Almost all researchers agreed on the simplified evaluation of requirement for fire safety
engineering with minimum engineering expertise for typical structures (Hdjisophocieous
et al.). For special structures with non-typical design for example large spaces, high
ceiling may challenge the engineer to apply the prescriptive based design. The
inapplicability of the accumulating restrictive regulations from code to code had caused
higher cost design that may not consider the heat transfer for example in large volume
spaces or the use of the intended spaces such as electronic of chemical cleaning rooms
that determine the fire resistance degree needed, limited innovative approaches, and the
belief that the only way to achieve fire safety design.
2.4.3 Advantages and Disadvantages of Performance-Based Design
Fire safety engineering in performance-based design is an engineering approach to fire
protection design based on (1) agreed upon fire safety goals, loss objectives and
performance objectives, (2) deterministic and probabilistic evaluation of fire initiation,
growth and development, (3) the physical and chemical properties of fire and fire
effluents, and (4) quantitative assessment of design alternatives against loss and
performance objectives (Meacham and Custer 1995). Recently, performance-based
codes are getting more and more recognized by allowing the design engineer to use more
means to evaluate the safety of the structure under fire. It is mainly caused by the
37


imposing new requirements of the excessively restrictive prescriptive design methods.
Yet, evaluating the overall structural performance is not easy. The degree of fire
resistance is the key feature of implementing the performance-based fire design codes.
Performance-based codes permits the use of the latest fire research, analysis, and models
that will meet the code safety and needs of both the user and the design engineering
(Hdjisophocieous et al. 1998). It establishes clear safety goals that the engineer has
control on. Performance-based approach helps harmonization of international guideline
and the use of new technical applications at lower costs and more flexible designs
(Hdjisophocieous et al. 1998). Other than expertise judgment of evaluating the risk
assessment of fire, performance-based design approach can be applied either probabilistic
or deterministic. Typically, the end results are measured in terms of the level of risks that
are imposed on the attendants and the structure if probabilistic methods were used, while
the fire growth calculations, spreading of smoke and behavior of structure are the criteria
if deterministic methods were used. The choice of the method used can be determined on
the complexity of the problem. However, it is difficult to define the level of safety
quantitatively and to evaluate compliance with the current requirements
(Hdjisophocieous et al. 1998). Usually, performance-based approach needs of computer
models to evaluate performance of the structure from the fire starts to the fire decay over
the components of the fire safety systems to check determine the level of success of the
design.
In summary, Fire safety engineering proposes a method to quantify and assess the
performance of the structure with respect to the fire growth, and evacuation. With the of
38


moving towards to the performance-based design codes that predicts the structural system
strength over the element strength itself (Kim et al. 2003).
2.4.4 Development and Evaluation of Design Fire
The development of design fire was clearly illustrated through a flow chart figure [Fig.
2.3] made by Custer (1995). Custer started with evaluating the situation that defines the
client loss objective(s). Then he developed the performance criteria that suites the
situation and then developed fire scenarios where fire is the main variable component. If
the performance exceeds the performance objectives then the design fire is selected. If the
performance undermines the performance objectives then another objective(s) for the
client loss has to be defined. The process continues until all the performance objectives
are met (Meacham and Custer 1995). The author also proposes a chart for evaluating
alternative designs that meet the optimum requirement for fire safety as shown in Fig. 2.4
(Custer 1995).
2.4.5 Importance Factor Approach
Kodur and Naser (2013) proposed a method to evaluate bridges subjected to fire. The
authors suggested an importance factor to evaluate the risks of fire over bridges. The
importance factor is based on the susceptibility of bridges to fire. Many factors are
involved in the calculations of the importance factor, some are related to size and
39


materials used in the bridge, others are the chances of fire occurrence. Kudor and Naser
(2013) proposes a weighted averages based on parameters and sub-parameters
determined from engineering judgment and design recommendations from previous
studies of Garlock et al. (2012); Elhag and Wang (2007); Dwaikat (2011); and Scheer
(2010). Figure 2.5 and Table 2.2 shows the characteristics and the weightage factors
based on bridge features that influenced the of fire hazard in a bridge according to Kodur
and Naser (2013).
According to Kodur and Naser (2013) there is about 5% of the nationwide bridge
population is considered to have a critical risk grades to fire hazard. The importance
factor classification listed in table 2.3 was based on the parameters and the sub-
parameters of the weightage factors shown in table 2.2. Using some equations stated in
chapter 5, Kodur and Naser (2013) determine the contribution to overall influence factors
from different classes as shown in Figure 2.6.
2.5 Conclusion
Based on the Author knowledge, it is very difficult to draw a conclusion from the
literature reviews. Failure modes of those literature review tests were either debonding
failure or concrete failure and they dependent on the certain conditions such as test setup,
data analysis, strengthening method, etc. It is clear that the NSM FRP strengthening
system shows better performance that the externally bonded FRP. Because of the
discrepancy between the NSM strip and NSM bar there is no clear conclusion, even
40


though the performance FRP in the NSM strips are better from the NSM FRP bars and
the FRP strips have the tendency rapture at high strength loads rather than debonding.
With all these studies conducted, very minimum research has focused on the performance
of NSM FRP at high temperature. The research presented in the following chapters is a
one step further in developing the NSM FRP emerging technology and ultimately
developing a design method or recommendations for the NSM FRP strengthening
applications.
41


Table 2.1. FRP Tensile Strength and Youngs Modulus (Adapted from ACI 440.2R-08)
FRP Fiber Type Ultimate Tensile Strength_______Youngs Modulus
Carbon 1,020-2,080 MPa 100-140 GPa
[150-350 ksi] 15,000-21,000 ksi
Glass 520-1,400 MPa 20-40 Gpa
[75-200 ksi] [3,000-6,000 ksi]
48-68 Gpa
Aramid 700-1,720 MPa [7,000-10,000
[100-250 ksil ksi]
42


Table 2.2. Bridge features weightage factors (Kodur and Naser 2013)
Parameter Sub-parameters Weightage factor (i Max. weightage factor (tp^max))
Class 1: Ceomerrica/ features, material properties and design characteristics O, *-0.44)
Structural system Truss jArch 1 5
Girder continuous 2
Girder amply supported 3
Cable-stayed 4
Suspension 5
Material type Reinforced concrete bridge 1 5
High strength/{prestressed) concrete bridge 2
Steel-concrete composite bridge 3
Concrete bridge strengthened with external FRP 4
Steel and timber brides 5
Span (m) <50 1 4
50-200 2
200-500 3
>500 4
No. of lanes 2 1 3
2-4 2
>4 3
Age (years) <15 1 4
15-20 2
30-50 3
>50 4
Current rating 100 1 5
60-80 2
40-60 3
20-40 4
<20 5
Additional service features 1 deck 1 5
2 decks pedestrians 2
Accommodates railroad 3
Multi-level 4
Above water 5
Weightage factor ((pbJd Max. weightage factor (tp^mox))
Class il: Hoard (fire) likelihood ($b- 023)
Response time (min) <5 1 5
5-10 2
10-20 3
20-30 4
>30 5
Historkal/architectural significance Conventional 1 3
Landmark 2
Prestigious 3
Threat perception None (low) 1 3
Not available (medium) 2
Frequent (high) 3
Fire scenario A small whicle fire abovejunder the bridge 1 5
A large truck collision and fire with other vehicles 2
A fuel tanker collision and fire with bridge sub-structure 3
Major fuel tanker collision and fire with multiple vehicles 4
and against bridge sub-structure
Fire due to fuel freight ship collision with a bridge pier 5
Weightage factor (ip**) Max. weightage factor (rpuAmax))
Class III: Traffic demand (i/rt-0.nj
ADT (vehicles/day) <1000 1 5
1000-5000 2
5000-15,000 3
15,000-50,000 4
>50,000 5
Facility location Rural 1 3
Suburban 2
Urban 3
Weightage factor (tpejc) Max weightage factor (tpfJf(max))
Class JV: Economic impact 3)
Closeness to alL routes (km) <10 1 3
10-20 2
>20 3
Time expected for repair (month) <3 1 3
3-9 2
>9 3
Cost expected for repair <1 million 1 3
1-3 million 2
>3 million 3
Goss V: Expected fire losses d09)
Lifejproperty losses Minimum to no injuries 1 3
Minimum casualties 2
Many casualties 3
Env. damage Minor damage 1 3
Significant damage 2
Unacceptable damage 3
43


Table 2.3. Risk grades w.r.t the overall class coefficient and the importance factor (Kodur and Naser 2013)
Risk gride Overall class coefficient (x) Importance factor (IF)
Critical 5*0.95 1.5
High 0.51 -0.94 1.2
Medium 0.20-030 1.0
Low <0.20 as
Figure 2.1. Pre-cured system aramid-type FRP installation (Carmichael and Barnes 2005).
Figure 2.2. Groove dimension limitations (ACI 440.2R-08, 2008)
44


Develop fire scenarios (possible design fires)

Select design fire(s)
Figure 2.3. Development of Design Fires (Custer 1995)
45


Far each design fire
develop trial designs
Select trial FP design
Evaluate selected design v. design fire
Select next design
alternative
Yes
Design acceptable
-------j----------
Modify selected design

Evaluate modified design v. design fire
Yes
Evaluate next design
alternative
Figure 2.4. Evaluation of Alternative Designs (Custer 1995)
I-----------
Vulnerability to fire
Fire hazard in a bridge
__________I___________
Classes
Geometrical features, material
properties and design
____ characteristics
-----------------1
Critical nature (of bridge)
Hazard (fire) like ihood
Classes
Traffic demand
Economic impact
Expected fire losses ard environmental
damage
Figure 2.5. The influencing characteristics of fire hazard in bridges (Kodur and Naser 2013)
46


Exacctcd fire
losses
9%
Figure 2.6. The contribution to overall influence factors from different classes (Kodur and Naser 2013)
47


3. Behavior of Near-Surface-Mounted CFRP-Concrete Interface in Thermal
Distress
3.1 Fabrication and Specimen Design
3.1.1 General Overview
Forty-eight (48) non-reinforced concrete blocks were made with compressive strength, fc,
of 40 MPa [5,800 psi] with each dimensioned as following (200 mm [8 in] longx 100 mm
[4 in] deepx 100 mm [4 in] wide) having a groove of size (25 mm [1 in] deepx 13 mm
[0.5 in] wide) along the strengthening direction. After concrete curing, the NSM Aslan
500 CFRP strip is inserted and bonded. Each twenty-four (24) concrete blocks were
bonded with the different types of adhesives, Power Fasteners T308+ (Power Fastener,
2011), a two components epoxy adhesive anchoring system which will be labeled as
ordinary epoxy; and high temperature laminated resin (PTM&W Industries 2013), a high
viscous adhesive epoxy resin with a hardener which will be labeled as high-temperature
epoxy. Based on the adhesive manufacturer data sheet recommendations, the NSM CFRP
bonded strips were cured, and the concrete blocks were subjected to thermal distress from
25C [77F] to 200C [392F] for three hours.
48


In summary the test setup had undergone three phases:
Phase 1: Concrete casting and curing
Phase 2: Bonding CFRP strips to concrete by two different types of adhesives,
then curing at room temperature to allow the adhesives to utilize the full strength.
Phase 3: Subjecting the concrete blocks to thermal distresses up to 200C [392F]
for three hours, and then testing the specimens in tension.
Throughout the literature review, NSM CFRP strips have proven to utilize the FRP
strength more than the CFRP bars. From the economical standing point, CFRP strips
consume less adhesive than CFRP bars by using smaller groove dimensions. CFRP strips
are also less labor intensive than the CFRP bars.
The failure expectancy of this test setup is either between the interface of the concrete-
NSM-CFRP or between the adhesive-NSM-CFRP. The temperature exposure and the
transition temperature, Tg, of the adhesive will determine the failure mode.
3.1.2 Preparation
The concrete blocks were casted in a wooden frame box that was prepared by the author
as shown in Fig. 3.1. The NSM CFRP-concrete interface blocks were designed with the
49


dimensions stated above and shown in Fig. 3.2. Two CFRP strip were inserted in the
groove starting from the middle of the concrete interface block to the other edge of the
specimens. The CFRP strip extended 100 mm [4 in] outside the edges of the concrete
block interface to allow for the MTS machine grips to clamp the specimen and apply
tension force as shown in Fig. 3.3.
3.1.3 Materials
Concrete was mixed in the laboratory for a specified compressive strength of 40 MPa
[5,800 psi]. The 28-day strength was measured to be 41.8 MPa [6,060 psi] based on the
average capacity of test cylinders (0100 x 200 mm [04x8 in.]). The CFRP strip used is
comprised of carbon fibers (4,626 MPa [700 ksi]) impregnated with a bisphenol epoxy
vinyl ester resin (Hughes Brothers 2008). The strip surface is treated to improve bond
when used with an adhesive. Table 3.1 summarizes typical engineering properties of the
CFRP. Two kinds of bonding agents were utilized to install the CFRP into a concrete
element: ordinary and high-temperature epoxy adhesives. The ordinary epoxy is a two-
part compound consisting of a hardener and a resin and is injected into a narrow groove
using a specially-designed nozzle. This mercaptan-free epoxy is viscous and has a
capacity of at least 57 MPa [8,267 psi]. The high-temperature epoxy includes low
viscosity and demonstrates good compatibility with structural fibers. It is particularly
suitable for structural bonding when significant thermal stress is expected. The bonding
agent will maintain its geometric stability until a temperature of 180C [356F] is reached
without showing premature distortion and shrinkage (PTM&W Industries 2013). A
50


hardener for this bonding material was selected according to the manufacturers
recommendation, which had favorable characteristics for wetting fabrics up to an
intermediate temperature range between 149C [300F] and 177C [350F], The glass
transition temperature specified by the manufacturer is at 170C [338F], The high-
temperature epoxy requires three-stage curing (PTM&W Industries 2013): i) a prepared
mixture is applied at room temperature and allows curing for a minimum of 18 hours; ii)
it is exposed to gradually increasing temperatures up to 177C [350F] for 12 hours; and
iii) the heated adhesive is cooled down to room temperature.
3.1.4 Test Specimen
To examine the behavior of an NSM CFRP strengthening system, an isolated interface
configuration was used as shown in Fig. 3.2. Such a test protocol can readily represent
NSM CFRP strips installed in a reinforced concrete beam with emphasis on the effective
tensile zone of the strengthened beam. A total of forty-eight (48) specimens were
prepared (Tables 3.2). Each specimen included one concrete block (200 mm [8 in]
longx 100 mm [4 in] deepx 100 mm [4 in] wide) having a groove along the strengthening
direction. The groove size (25 mm [1 in] deepx 13 mm [0.5 in] wide) was designed as per
the recommendation of ACI.440.2R-08 (ACI 2008): the width and depth of the groove
were greater than 3 and 1.5 times the CFRP thickness and width, respectively. Upon
complete curing of the concrete specimens, CFRP strips were bonded with the adhesives
mentioned earlier. The groove was first cleansed with an air brush to eliminate surface
dirt that could degrade bond between the adhesive and concrete surface [Fig. 3.4 (a)].
51


Two precut CFRP strips (16 mm [0.63 in] widex2 mm [0.08 in] thickx200 mm [8 in]
long, each) were positioned with each bonding agent injected into the groove [Fig. 3.4
(b)]. Both ends of the groove were unbonded using styroform pieces to prevent the
premature failure of the interface caused by stress concentrations when mechanically
loaded. The assembled test specimens were cured according to the manufacturers
recommendation.
3.1.5 Thermal Exposure
After adequate curing of the adhesively-bonded NSM CFRP strips, all specimens were
exposed to elevated temperatures ranging from 25C [77F] to 200C [392F] at a typical
interval of 25C [77F] for three hours using a digital-control electric furnace [Fig. 3.5
(a) ]. It should be noted that the furnace was preheated to a designated temperature before
the thermal exposure of the NSM CFRP-concrete interface commenced. The temperature
of the specimens was monitored using a laser thermometer gun, as shown in Fig. 3.5 (a),
to ensure the temperature inside the furnace. The thermometer gun confirms the
thermocouple reading as shown in the Fig 3.6. When the planed thermal exposure was
completed, each specimen was cooled down to room temperature for one day [Fig. 3.5
(b) ]. The variation of concrete strength within the temperature range studied in this
experimental program was previously examined by Kim et al. (2012) and thus additional
material testing was not conducted.
52


3.1.6 Test Setup and Instrumentation
The specimens were positioned to an MTS machine for mechanical testing, as shown in
Fig. 3.5 (c). The CFRP strips were gripped and monotonic tension was applied at a
loading rate of 0.1 mm/sec [0.004 in/sec] until interfacial failure took place. A clamping
system was used [Fig. 3.5 (b) and (c)] to preclude tensile splitting failure of the concrete
in the vicinity of the gap between the two CFRP strips (i.e., mid-length of the specimen).
The load applied was measured with the built-in load cell and corresponding
displacement was recorded by the stroke of the machine.
3.2 Material-Level Testing
3.2.1 Dynamic Mechanical Analysis
Dynamic Mechanical Analysis (DMA) [fig.3.7] was used to measure the CFRP and
adhesive epoxy response. DMA uses an imposed oscillating stress or strain. Some level
of deformation will be caused by the dissipation of stress imposed on the samples. Since
most adhesives are viscoelastic in nature, some level of deformation will be recovered
when stress is released. Using Hooks law, the DMA machine measures the phase lag
stress strain. The computerized DMA instrument measures the storage and recovered
modulus and the loss energy as well. The ratio of the loss modulus over the storage and
recovered modulus will determine tan(8). The peak of tan(S) is the good indication of
verifying the transition temperature (Tg) of the different type of adhesives and the CFRP
53


strip. The test samples were conducted at the mechanical department according to ASTM
E1356-03. The results are shown in Fig. 3.8 for the CFRP strip [Fig. 3.8(a)], the high
temperature epoxy [Fig. 3.8(b)], and the ordinary epoxy [Fig. 3.8(c)], The high
temperature and the ordinary epoxy were specified as a thermoset resin. A sudden drop in
the storage modulus was shown by the ordinary epoxy which may indicated a melting
possibility and thus losing it properties. The glass transition temperature (Tg) of the high
temperature epoxy is at 163C [325F], and the ordinary epoxy appeared to be
approximately 65C [149F],
3.2.2 Concrete Testing under Thermal Distress
Kim et al. (2012) performed concrete compressive strength testing over a O100 x 200
mm [04x8 in.] diameter cylinders. The results were discussed earlier under Materials
(3.1.3). The results are summarized in Fig. 3.9(a). From the results the temperature
specified seemed to have minimal effect on the concrete strength.
3.2.3 CFRP Strip Testing under Thermal Distress
The performance of CFRP strips under tension load was consistent over the specified
temperatures. The failure mode was along the fiber line and might be caused by a CFRP
slip between the grips. The results are summarized in Fig. 3.9(b). From the results and
literature review the specified temperature has no effect on the CFRP tension capacity.
54


3.2.4 Ordinary Epoxy and High-Temperature Coupons
The tensile testing of the ordinary epoxy and the high-temperature epoxy were
determined by ASTM D638 Standard Test Method for Tensile Properties of Plastics
(2010) to verify the effect of thermal distress on the epoxy resin. The results are
summarized in Figs. 3.9(c) and 3.9(d). The results does not fully confirm with the
manufacturer specification data sheet, mainly due to the residual tensile testing where
some of the properties were recovered after thermal curing and cooling process. A further
investigation is needed in this field. The failure mode of the adhesive epoxy coupons and
the CFRP strip is shown in Figure 3.10.
3.3 Experimental Results
3.3.1 Analysis of Variance for Concrete at Elevated Temperatures
The strength variation of concrete in compression is summarized in Table 3.3, depending
upon the degree of temperature exposure. Although the strength tended to decrease with
an increasing temperature, an insignificant difference was observed in all test categories
within a range between 41.8 MPa [6,060 psi] to 39.2 MPa [5,690 psi]. To clarify the
temperature-dependent strength issue of the concrete, analysis of variance (ANOVA) was
performed at a level of significance a = 0.05. Hypotheses were proposed whether the
55


temperature range studied would have any effect on the variation of concrete compressive
strength. The following was tested using Eq. 1:
H0 : All the means of the tested concrete strength are equal
7/a: Not all the means of the tested concrete strength are equal
ms_
F =
( k \
2>2
V i=1
Ik
(1)
where F is the F distribution of the tested concrete; m is the sample size per test group; s2
x
is the sample variance; is the sample variance per group; and k is the number of the
means. The degree of freedom is defined as DOF = (k-l,n-k) to determine the critical F
distribution value (e.g., F,05 for a level of confidence of 0.05). The calculated F
distribution value was 0.87, which was not in the critical region F,05 of 2.85. The H0
hypothesis, therefore, was not rejected. It implies that insufficient statistical evidence has
existed to conclude the thermal exposure varying from 25C [77F] to 200C [392F]
influenced the strength of the concrete.
3.3.2 Interfacial Strength of NSM CFRP
Table 3.2 lists the failure load of the specimens bonded with the ordinary epoxy adhesive
and high-temperature epoxy. The average capacity of the NSM CFRP-concrete interface
56


without experiencing thermal distress for ordinary epoxy (the OE25 series) was 21.5 kN
[4.8 kip] with a standard deviation of 2.5 kN [0.6 kip]. The interfacial capacity was
preserved up to a temperature of 75C [167F], This illustrates that the glass transition
temperature (Tg) of the ordinary epoxy appeared to be approximately 75C [167F] and
hence the performance of the NSM CFRP strengthening system was not influenced by
the degree of thermal exposure within this temperature boundary. The specimens
subjected to a range between 100C [212F] and 150C [302F] exhibited some
degradation in interfacial capacity, including an average decrease of 20.8% in
comparison to the capacity of the 25C [77F] category. The rate of strength decrease
was significant beyond a temperature of 150C [302F], The capacity of the specimens
exposed to temperatures of 175C [347F] and 200C [392F] was 38.6% and 89.8%
lower than that of the specimens tested at 25C [77F], The strength of the specimens
subjected to 200C [392F] was almost negligible. The NSM CFRP strengthening system
bonded with the ordinary epoxy, therefore, will not function as designed when the
surrounding temperature exceeds 75C [167F] and will demonstrate a noticeable
decrease in its interfacial capacity over 150C [302F], even though an insulation
material covers the strengthening system. It means that the fire endurance time of NSM
CFRP-strengthened members bonded with such an adhesive material needs to be
estimated until critical temperatures are reached for design and practice.
The interfacial strength of the specimens bonded with the high-temperature epoxy is
given in Table 3.2. The ones without thermal exposure (the HE 25 series) demonstrated
an average interfacial strength of 17.2 kN [3.9 kip], which was 20% lower than the
57


strength of the OE series at 25C [77F], The interfacial capacity of the specimens tended
to decrease with an increasing temperature, while the degree of strength reduction was
not as obvious as the case of the ordinary-epoxy bond. For instance, the strength drop of
the HE series was 28.5% when a temperature changed from 25C [77F] to 200C
[392F], which was a substantially low rate when relative to the 89.8% of the OE series
under the same condition. According to the test results of the HE specimens shown in
Table 3.2, the glass transition temperature of the high-temperature epoxy seemed to be at
around 175C [347F], This value well conforms to the glass transition temperature
reported by the manufacture, 170C [338F], A graphical comparison on the thermal
performance of the NSM CFRP-concrete interface bonded with the ordinary or high-
temperature epoxy is made in Fig. 3.11. The strength gap between these two cases was,
by and large, maintained up to a temperature of 150C [302F] beyond which the
specimens bonded with the high-temperature epoxy displayed superior behavior to those
with the ordinary epoxy.
3.3.3 Load-Displacement Response
Figure 3.12 shows the load-displacement behavior of the specimens bonded with the
ordinary epoxy. The OE25 series exhibited a linear response until an abrupt load-drop
was associated due to bond failure [Fig. 3.12 (a)]. The stiffness of the respective
specimens was more or less similar to each other, while their strength was slightly
different. This can be explained by the fact that the amount of epoxy inside the groove
was not the same because of experimental randomness to a certain extent. The specimens
58


subjected to elevated temperatures demonstrated similar behavior [Fig. 3.12 (b) to (g)],
while a local load drop was noticed in some cases. Such an observation illustrates that the
interface was partially damaged due to the thermal exposure. The specimens at 200C
[392F] showed somewhat ductile failure with a substantially low capacity. It should be
noted that Specimen OE200-3 failed in a premature manner when tested and thus its load-
displacement response was not provided in Fig. 3.12(h). A summary of load displacement
over the range of different temperature for ordinary epoxy is shown in Fig. 3.14(a).
Figure 3.13 reveals the behavior of the specimens bonded with the high-temperature
epoxy. The response of these cases was virtually the same as that of the former cases
discussed in Fig. 3.12. It may be of interest to note that the cured high-temperature epoxy
looked like brittle glass that was different from the ordinary epoxy and thus shows
adhesive failure at low temperatures and bond failure at high temperatures.
3.3.4 Failure Mode
The interface failure of selected specimens is presented in Fig. 3.15. Although
infinitesimal deformation was recorded with an increasing mechanical load (Figs. 3.12
and 3.13), a morphological change along the CFRP-concrete interface was not apparent
until a peak load was achieved. When the failure of the interface was imminent, several
hairline cracks formed along the bond-line and further developed, as shown in Fig.
3.15(a). At this stage, stress redistribution was expected between the CFRP system and
surrounding concrete. The failure of the CFRP was localized along the groove [Fig.
3.15(b)], which indicated the existence of an effective failure zone in the vicinity of the
59


installed NSM CFRP. A relationship among this effective failure zone, the configuration
of the CFRP system, and the properties of the surrounding concrete appears to be an
interesting topic for future research. A secondary concrete crack was observed near the
mid-length of the specimen because of the spatial gap between the two CFRP strips. The
failure mode of the HE specimens bonded with the high-temperature epoxy [Fig. 3.15(c)]
was analogous to that of the OE categories. Additional failure modes of ordinary epoxy
and high temperature epoxy at 50C [122F] and 200C [392F] are shown in Fig.
3.16(a)-(d).
3.4 Summary and Conclusion
This chapter has discussed the residual behavior of NSM CFRP strips embedded in a
concrete substrate when subjected to elevated temperatures ranging from 25C [77F] to
200C [392F], Two types of adhesives were used to bond the strips: ordinary and high-
temperature epoxies. Test specimens were conditioned in the predefined thermal
environment for three hours, cooled down to room temperature, and mechanically loaded
to failure. Interfacial strength between the CFRP and concrete was measured and
corresponding failure mode was studied. These technical results are part of an ongoing
research program examining the thermal performance of an NSM CFRP strengthening
system for concrete structures with focus on material- and structure-level investigations.
Some preliminary conclusions are drawn as follows.
60


According to the statistical analysis, the variation of concrete strength in
compression was not influenced by a temperature range from 25C [77F] to
200C [392F], The capacity of the CFRP-concrete interface was, however,
affected by thermal exposure within this boundary. The specimens bonded with
the ordinary epoxy preserved interfacial strength until 75C [167F] beyond
which a noticeable strength decrease was observed, in particular over 175C
[347F] The ones with the high-temperature epoxy, by and large, maintained
their interfacial strength up to 200C [392F] even though a trend of strength-
reduction was associated with an increasing temperature.
The interfacial capacity of the ordinary epoxy was higher than the high
temperature epoxy, but the high temperature epoxy showed a consistent capacity
over the temperature increase.
61


Table 3.1, Properties of CFRP strip from the manufacturer (Hughes Brothers 2008)
Property Value
Width 16 mm [0.63 in]
Thickness 2 mm [0.079 in]
Cross-sectional area 31.2 mm2 [0.49 in2]
Tensile strength 2068 MPa [300 ksi]
Tensile modulus 124 GPa [18,000 ksi]
Ultimate strain 0.017
Coefficient of thermal expansion (transverse) 74 to 104 x 10'6 (/C) [41 to 58 x 10'6(/F)]
Coefficient of thermal expansion (longitudinal) -9 to Ox 10-6(/C) [4 to Ox 10 6 (/F)|
62


Table 3.2, Test specimens load and failure mode of ordinary epoxy and high temperature epoxy
Test Specimens bonded with ordinary epoxy
ID Failure load Failure Mode ID Failure load Failure Mode
kN kip] kN kip]
Each Average Each Average
OE25-1 21.2 [4.8] 21.4 [4.8] Resin Failure OE125-1 17.5 [3.9] 16.2 [3.6] Interface Failure
OE25-2 19.1 [4.3] OE125-2 13.3 [3.0]
OE25-3 24.1 [5.4] OE125-3 17.6 [4.0]
OE50-1 18.2 [4.1] 20.3 [4.6] Resin Failure OE150-1 13.4 [3.0] 17.8 [4.0] Interface Failure
OE50-2 23.3 [5.2] OE150-2 19.4 [4.4]
OE50-3 19.4 [4.4] OE150-3 20.6 [4.6]
OE75-1 19.4 [4.4] 20.7 [4.7] Interface Failure OE175-1 14.6 [3.3] 13.2 [3.0] Interface Failure
OE75-2 20.7 [4.7] OE175-2 11.3 [2.5]
OE75-3 21.9 [4.9] OE175-3 13.9 [3.1]
OElOO-1 15.5 [3.5] 17.1 [3.8] Interface Failure OE200-1 3.1 [0.7] 2.2 [0.5] Interface Failure
OElOO-2 16.6 [3.7] OE200-2 1.3 [0.3]
OElOO-3 19.1 [4.3] OE200-3 N/A
Test specimens bonded with high-temperature epoxy
ID Failure load Failure Mode ID Failure load Failure Mode
kN kip] kN kip]
Each Average Each Average
HE25-1 16.9 [3.8] 17.2 [3.9] Resin Failure HE125-1 11.4 [2.6] 12.8 [2.9] Interface Failure
HE25-2 16.5 [3.7] HE125-2 13.4 [3.0]
HE25-3 18.1 [4.1] HE125-3 13.6 [3.1]
HE50-1 16.4 [3.7] 16.1 [3.6] Resin Failure HE150-1 14.6 [3.3] 15.2 [3.4] Interface Failure
HE50-2 16.5 [3.7] HE150-2 15.9 [3.6]
HE50-3 15.3 [3.4] HE150-3 15.0 [3.4]
HE75-1 12.8 [2.9] 15.1 [3.4] Resin Failure HE175-1 11.4 [2.6] 13.1 [2.9] Interface Failure
HE75-2 14.5 [3.3] HE175-2 13.0 [2.9]
HE75-3 18.1 [4.1] HE175-3 15.0 [3.4]
HE100-1 16.6 [3.7] 16.3 [3.7] Resin Failure HE200-1 13.2 [3.0] 12.3 [2.8] Interface Failure
HE100-2 18.8 [4.2] HE200-2 12.2 [2.7]
HE100-3 13.5 [3.0] HE200-3 11.7 [2.6]
N/A: premature failure during test
63


Table 3.3, Compressive concrete strength depending upon temperature (Kim et al. 2012)
ID Test strength (MPa [psi]) ID Test strength (MPa [psi]) ID Test strength (MPa [psi])
Each Average Each Average Each Average
C25-1 41.7 [6,050] 41.7 [6,050] C100-1 40.8 [5,9201 40.8 [5,920] C200-1 36.6 [5,3101 39.2 [5,690]
C25-2 41.2 [5,980] C100-2 42.1 [6,110] C200-2 41.2 [5,980]
C25-3 42.1 [6,110] C100-3 39.1 [5,670] C200-3 38.8 [5,630]
C50-1 41.2 [5,980] 41.8 [6,060] C125-1 38.4 [5,570] 40.4 [5,860] ANOVA: Level of confidence a = 0.05 A =0.87 DOF = (6,14) Critical region F05 = 2.85 Conclusion: F < F05
C50-2 44.1 [6,400] C125-2 41.2 [5,980]
C50-3 40.1 [5,820] C125-3 41.7 [6,050]
C75-1 42.6 [6,180] 41.2 [5,980] C150-1 37.1 [5,380] 40.0 [5,800]
C75-2 39.1 [5,670] Cl 50-2 42.8 [6,210]
C75-3 41.7 [6,050] C150-3 39.3 [5,700]
64


Figure 3.1. Wooden mold in which concrete blocks were formed
Top View
E
E
E
E
Concrete
Epoxy
CFRP
13mm
100mm
Figure 3.2. NSM CFRP-concrete interface test specimen dimension (not to scale)
65
100mm


(a)
Figure 3.4. Test specimen: (a) air-blasting before CFRP-bonding; (b) CFRP installation
Figure 3.5. Test setup: (a) high temperature exposure; (b) conditioned specimen; (c) tension test
66


Temperature (C) Temperature (C)
(a) (b)
200 i
175
150 -
125 -
100
75
50
25
0
Temperature
--1----1---(
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(c)
O
2
5
a
E
l-
200 i
175
150 -
125 -.
100
75
50 -
25 -
Temperature
0 -I--------1------1-------1-------1------1-------1------1-------1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(d)
200 -I
175 -
150 -
125 -
100 -
75 -
50 -
25 -
Temperature
0 -l-------1------,------1------1------1-----1------1------1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(e)
o
o
0
3
0
0
Q.
E
£
200
175
150
125
100
75
50 H
25
Temperature
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(f)
67


200
_ 175
O
^ 150
= 125
15
g> 100
Q.
E
75
50
25
0
-Temperature
---1----1---------1------1-------1------1------1-------1
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
O
o
£
=
2
ai
Q.
200
175
150
125 -
100 -
75 -
50
25 ^
0
200C
Temperature
200C
-| 175C
" 150C
125C
100C
^ 75C
""" 50C
Time (hrs)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(g) (h)
Figure 3.6. Thermocouple temperature readings over a tliree hour range acquired from the data logger: (a)
at 50C [122F]; (b) at 75C [167F]; (c) at 100C [212F]; (d) at 125C [257F]; (e) at 150C [302F]; (f)
at 175C [347F]; (g) at 200C [392F]; (h) combined temperature
(a) (b) (c)
Figure 3.7. Dynamic mechanical analysis (DMA): (a) testing machine; (b) DMA clamps; (c) DMA
specimens
68


(a)
(b)
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
(c)
Figure 3.8, Dynamic Mechanical Analysis (DMA) results: (a) CFRP strip; (b) high temperature epoxy; (c)
ordinary epoxy
69


1.2
.c 1 -
c 0) 05 0.8 -
~a a> N 0.6 -
cz E 0.4 -
o z 0.2 -
0 -
a Individual
----Average
25 50 75 100 125 150 175 200
Temperature (C)
(a)
Temperature (C)
(b)
CO
o
1)
18 1
1.6 -
1.4 -
1.2 -
1 -
0.8 -
0.6 -
0.4 -
0.2 -
0 -
A $/ A
...A A
A A A
A
A
A
A
-Average OE
CD
D.
1.2
1 -I
0.8
0.6
0.4 -I
0.2
0
80
70
60
50 H
40
30
20
10 H
0
A A
A A
A
A
a Individual OE
1.8
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
0
-kt'a'"'"a"''a
a Individual
----Average
25 50 75 100 125 150 175 200
Temperature (C)
(b)
A Individual HE
-----Average HE
III
.X A
A
A
A A
A A
A "V' A
A ft
50 75 100 125 150 175 200
Temperature (C)
(d)
A A
A\ A A 4
' A ' / \ ,, \
A A A 1 A A X A A
a Individual HE
.....Average HE
25 50 75 100 125 150 175 200
Temperature (C)
(e)
0 25 50 75 100 125 150 175 200
Temperature (C)
(f)
Figure 3.9. Temperature-dependent residual strength measured: (a) concrete in compression; (b) CFRP
strip in tension; (c) ordinary epoxy in tension; (d) high-temperature epoxy in tension; (e) normalized
strength of OE; (f) normalized strength of HE
70


(a) (b) (c)
Figure 3.10. Failure mode: (a) HE epoxy coupons at 200C [392F]; (b) OE epoxy coupons at 200C
[392F]; (c) CFRP strip at 200C [392F]
d
to
o
o
to
25
20 -I
15
10
5 1
A-.
A
A Individual
----Average
,A 2
'A
A
0 25 50 75 100 125 150 175 200
Temperature (C)
(a)
E
5
25
20
15
10
A Individual
----Average
A A ''AA
A A
0 25 50 75 100 125 150 175 200
Temperature (C)
(b)
(c)
Figure 3.11. Temperature-dependent interfacial strength: (a) specimens bonded with ordinary epoxy; (b)
specimens bonded with high-temperature epoxy; (c) temperature-dependent average interfacial strength
71


Displacement (mm)
Displacement (mm)
(a)
(b)
Displacement (mm)
Displacement (mm)
(c)
(d)
Displacement (mm)
Displacement (mm)
(e)
(f)
72


---OE200-1
---OE200-2
Displacement (mm)
15
(g)
(h)
Figure 3.12. Load-displacement of specimens bonded with ordinary epoxy: (a) at 25C [77F]; (b) at 50C
[122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C [257F]; (f) at 150C [302F]; (g) at
175C [347F]; (h) at 200C [392F]
73


Load (kN) Load (kN) Load (kN)
Displacement (mm)
Displacement (mm)
(a)
(b)
Displacement (mm)
Displacement (mm)
(c)
(d)
Displacement (mm) Displacement (mm)
(e)
(f)
74


(g) (h)
Figure 3.13. Load-displacement of specimens bonded with high-temperature epoxy: (a) at 25C [77F|; (b)
at 50C [122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C [257F]; (f) at 150C [302F]; (g)
at 175C [347F]; (h) at 200C [392F]
(a) (b)
Figure 3.14. Load-displacement comparison of exposure temperatures: (a) bonded with ordinary epoxy; (b)
bonded with high-temperature epoxy
(a) (b) (c)
Figure 3.15. Failure mode: (a) interface failure of ordinary epoxy; (b) interfacial failure of ordinary epoxy
after test; (c) interfacial failure of high-temperature epoxy after test
75


(a) (b) (c) (d)
Figure 3.16. Failure mode: (a) ordinary epoxy at 50C [122F]; (b) ordinary epoxy at 200C [392F]; (c)
high-temperature epoxy at 50C [122F]; (d) liigh-temperalnre epoxy at 200C[392F]
76
too


4. Bond Performance of Concrete-Adhesive Interface at Elevated Temperatures
4.1 Fabrication and Specimen Design
4.1.1 General Overview
Seventy-two (72) non-reinforced concrete prisms were made with compressive strength,
fc of 40 MPa [5,800 psi] with each dimensioned as following (400 mm [16 in]
longxlOO mm [4 in] widex 80 mm [3.15 in] deep). The prisms were split in half by
using a wood divider sheet as shown in Fig. 4.1. After concrete curing, every twenty-four
(24) concrete prism with the different type of adhesive Power Fasteners T308+ (Power
Fastener, 2011), a two components epoxy adhesive anchoring system which will be
labeled as high viscosity adhesive epoxy; BASF MBrace Saturant (MBrace, 2007), also a
two components epoxy adhesive anchoring system will be labeled as Low viscosity
adhesive epoxy; and high temperature laminated resin (PTM&W Industries 2013), a high
viscous adhesive epoxy resin with a hardener which will be labeled as high-temperature
epoxy. The two pieces-prism were bonded together to form one, and then cured based on
the adhesive manufacturer data sheet recommendations. After adhesive curing, a small
groove was created by a stationary electric saw, dimensioning 10 mm [0.4 in] deep. The
concrete prism were subjected to thermal distress from 25C [77F] to 200C [392F] for
three hours.
77


The failure expectancy of this test setup is at the bonding location of the two prism pieces,
the interface of the concrete-adhesive or between the adhesive-adhesive. The temperature
exposure and the transition temperature, Tg, of the adhesive, and the adhesive thickness
will determine the failure mode.
4.1.2 Concrete and Adhesive Material Properties
Concrete was mixed in the laboratory under the same conditions of the previous
experiment. The specified compressive strength was 40 MPa [5,800 psi] after 28 days.
Dog-bone shaped coupons were made from the three types of adhesive. Twenty-four (24)
coupons were made of each adhesive type. The coupons are 100 mm [4 in] long x 5 mm
[0.2 in] wide x 10 mm [0.4 in] deep as show in Fig. 4.2. The final shape of different
adhesive types testing coupon are shown in Fig. 4.3. The coupons were then exposed to
temperature distress from 25C [77F] to 200C [392F] for three hours. The high-
viscosity adhesive is a two-part compound consisting of a harder and a resin and is
injected onto one side of the prism and then bonded together. This mercaptan-free
adhesive is viscous and has a capacity of at least 57 MPa [8,267 psi]. The low-viscosity
adhesive is a two-part (Part A to Part B) mixed with a 100:30 ratio by weight applied
after manufacturer recommended mixing on one side of the prism and bonded the other
prism piece. The adhesive compression capacity is 86.2 MPa [12,500 psi]. The low-
viscosity adhesive has a transition temperature, Tg, of 71C [163F] (Mbrace, 2007).
Material properties of this bonding agent are provided in Table 4.1. The high-
temperature epoxy is particularly suitable for structural bonding when significant thermal
78


stress is expected. The transition temperature, Tg, of this epoxy is 180C [356F]
(PTM&W Industries 2013). The manufacturers recommends a hardener for this bonding
material mixed over a ratio of 100:22 by weight (PTM&W Industries 2013).
4.1.3 Testing Specimen and Thermal Exposure
The test was conducted per ASTM C78, a standard testing method with third-point
loading (ASTM, 2009). The test set-up investigates the prism bond strength of 72
concrete prism specimens. After concrete curing, each 24 prism were bonded with
different type of adhesive and cured according to the manufacturer recommendations.
The concrete prisms are all identical and mixed with the same concrete mix design. The
test set up is shown in Fig. 4.5(b). To the author knowledge, there was no particular test
preparation found that provide the bonding performance of concrete-adhesive interface at
elevated temperatures.
After proper curing of the adhesive bonded prisms, specimens were exposed to elevated
temperature over the range from 25C [77F] to 200C [392F] over a typical interval of
25C [77F] for three hours using a digital-control electric furnace and verified by using
a temperature data logger [Fig. 4.6], The data logger shows consistent oven temperature
over the testing temperature range [Fig. 4.7(a)-(e)]. It should be noted that the furnace
was preheated to a designated temperature before the thermal exposure of the concrete
prisms and coupons.
79


4.1.4 Instrumentation
The concrete prisms were positioned over a steel fabricated three-point supports that are
350mm [13.8 in] apart, then placed in an MTS machine as shown in Fig. 4.4(a)-(b). The
compression loading rate was of 0.1 mm/sec [0.004 in/sec] until interfacial failure
happened. Before flexural concrete prism test took place, the specimens were cut with
notches over the bonding location of the specimen by a diamond saw with high stability
and accuracy [Fig 4.8], The depth of the notches was 10mm [0.4 in]. This notch was
placed to keep the failure mode at the bonding location and prevent from having a
concrete failure between the supports. The load applied was measured with the built-in
load cell and corresponding displacement was recorded by the stroke of the machine.
The dynamic modulus test was performed as per ASTM C215 (2008) using the DK-4000
machine [Fig. 4.10(a)], The transverse frequency was measured by using the impact test
method. An accelerometer with an output signal was attached to one end of the prism.
The concrete prism was supported by a standard two-point aluminum pad specific for this
test and impacted by a standard hammer at consistent specified locations over all the
concrete prisms. The response frequency domain wave was recorded by DK-4000 [Fig.
4.10(b)], Please note that the dynamic modulus test was performed after the notch in the
concrete was made, the sample was fully concrete and adhesive cured, and the concrete
prisms were subjected to thermal distresses.
80


4.2 Material-Level Testing
4.2.1 Stress-strain Response of Adhesive Resin
Table 4.2 lists the failure tensile load of the coupons made with different adhesives that
was specified earlier at elevated temperature. Even though the residual tensile capacity of
the adhesive coupons did not reflect failure around the glass transient temperature, the
results remains fairly consistent throughout. The average residual tensile capacity of high
viscosity adhesive was 25.4 MPa [3.68 ksi]. The residual tensile capacity of the low
viscosity adhesive epoxy coupons did show a trend drop in the tensile capacity at Tg
around 75C [167F] but the ultimate capacity was 58.8 MPa [8.53 ksi]. The tensile
ultimate strength capacity of the low viscosity adhesive epoxy coupon was 55.2 MPa
[8,000 psi] as per the manufacturer data sheet [table 4.1], The High temperature adhesive
epoxy coupons did not perform as it was specified in the data sheet. This behavior is due
to the complex curing requirement that this adhesive requires. As per the data sheet, the
curing requires a 18 hours set after mixing the adhesives then gradually heat to 65.5C
[150F] for 3-4 hours, then the manufacturer also specifies to slowly raise the
temperature to 121C [250F] and hold for 3-4 hours. Lastly the manufacturer
recommends turning off the curing oven and letting it cool down to room temperature
before service. The glass transient temperature was not clear in this test, which was
examined earlier and found to be at 175C [347F] from the NSM CFRP concrete block
tests.
81


4.2.2 Storage Modulus of Adhesive Resin
The storage modulus of the adhesive resin was determined by using the dynamic
mechanical analysis (DMA). The DMA is a is a computerized instrument that measures
the storage modulus under a heat flow difference and was used to verify the transition
temperature (Tg) of the different type of adhesives. The tests were performed according to
ASTM E1356-03. The results shown in Fig. 4.10(a) identify a Tgof76C [169F] for low
viscosity adhesive, while Fig. 4.9(b) shows a Tg of 65C [149F] for high viscosity, and
Fig. 4.9(c) shows a Tg 163C [325F] for high temperature adhesive.
4.2.3 Frequency Test
The frequency test was performed on the concrete prisms bonded with the three different
types of adhesives discussed earlier and subjected to different thermal distresses. The
frequency test was performed according to ASTM C215 (2008). The prisms were marked
and had the same conditions throughout the testing period. The results will be are shown
in Fig. 4.11(a).
82


4.3 Experimental Results
4.3.1 Dynamic Elastic Modulus of Concrete Prisms
The dynamic elastic moduli of concrete prisms were determined by the standard test
method for fundamental transverse, longitudinal, and torsional frequencies of concrete
specimens (ASTM C215, 2008). Table 4.3 list the dynamic elastic modulus over the
specified temperature range. The high viscosity specimens that was not subjected to
thermal distresses HV-25 shows a dynamic elastic modulus of 20.44 GPa [2964.6 ksi].
The high viscosity specimens subjected to 200C [392F] temperature shows 51.8% drop
with a dynamic elastic modulus of 9.85 GPa [1428.6],
The low viscosity (LV) dynamic elastic modulus starts with 21.09 GPa [3058.85 ksi] at
room temperature and dropped to 7.36 GPa [1067.48 ksi] at 200C [392F], The LV
dynamic elastic modulus was preserved up to 75C [167F] that confirms the glass
transition temperature (Tg) specified by the manufacturer at 71C [163F], The dynamic
elastic modulus of the LV loses 65.1% of its original capacity.
The high temperature (HT) dynamic elastic modulus starts with 21.09 GPa [3058.85 ksi]
without being subjected to any thermal distresses. The dynamic elastic modulus was
fairly consistent up to 125C [257F] with a drop of 3%. The dynamic elastic modulus
dropped 24.9% at 175C [347F], The overall drop in the dynamic elastic modulus at
200C [392F] was 51.7%. Figure 4.11(a) shows the frequency readings with the impact
83


test method. By using the frequency from the impact testing, the dynamic modulus of
elasticity was found through the ASTM C215 (2008) equations [Eq. 2 and 3], The overall
dynamic elastic modulus is presented in figure 4.11(b).
E: Dynamic modulus of elasticity (Pa)
n: Transverse frequency
M: Mass of the specimen
C: Dimensions of the specimen
L: Length of specimen
t: Thickness of specimen
b: Width of specimen
T: Correction factor dependent on the radius of gyration to length of specimen and the
Poissns ratio
E = CMn2 (2)
773
C = 0.9464 (3)
bt3
4.3.2 Load Displacement Response of the Bond Test Concrete Prisms
Figure 4.12(a)-(b) shows the ultimate load and capacity of the concrete prism specimens
bonded with high viscosity adhesive after subjected to thermal distresses. As it is clear,
there is no specific trend in the specimens results even when utilizing the bonded area in
the concrete prisms. Since, it was observed that the thickness of the adhesive varies
84


between one type to another depending on many factors e.g. the quantity, application
technique, and human error. It was proposed by the author to use the thickness as a way
to normalize the capacity of the concrete prisms. With this being in mind the normalized
capacity utilized the bonded volume rather than the bonded area. Figure 4.12(c) shows a
decreasing trend, although the glass transition temperature (Tg) is not clear.
Similarly, the low viscosity adhesive ultimate load and capacity [Fig. 4.13(a)-(b)] shows
no clear trend. After normalizing the results over the thickness of the adhesive, it was
found a decreasing trend [Fig. 4.13(c)],
As for high temperature adhesive concrete prism, the ultimate load and capacity were not
clear as well [Fig. 4.14(a)-(b)]. The normalized ultimate capacity shows a consistent
decreasing trend [Fig. 4.14(c)],
Table 4.4a lists the temperature dependent adhesive-concrete interfacial capacity of the
concrete prism specimens bonded with high viscosity adhesive, Table 4.4b list the
adhesive-concrete interfacial capacity of the concrete prism specimens bonded with low
viscosity adhesive at the specified temperature range, and Table 4.4c lists the high
temperature adhesive-concrete interfacial capacity of the temperature dependent concrete
prisms. The average capacity of the bond performance of concrete prisms bonded with
high viscosity adhesive without being subjected to thermal distresses was 0.31 MPa/mm
[1142 psi/in]. At 200C [392F], the normalized bond capacity dropped by 68.5%.
Similarly the low viscosity normalized bond capacity dropped by 68.2% between the
85


specimens not subjected to thermal distresses and specimens subjected to 200C [392F],
The concrete prisms bonded with high temperature adhesive was 0.47 MPa/mm [1731.5
psi/in] at room temperature and dropped to 0.24 MPa/mm [884.2 psi/in] at 200C [392F],
The normalized capacity drop was the least compared to the other types of adhesive with
48.8%. The concrete prisms bonded with high temperature adhesive have the highest
capacity [Fig. 4.15], Even the glass transition temperature (Tg) was not clear in this test,
the results follows the same trend of the frequency [Fig. 4.11(a)] and the dynamic elastic
modulus [Fig. 4.11(b)],
4.3.3 Failure Mode of Bonded Concrete Prisms
The interface failure of the concrete prisms bonded with different types of adhesive is
shown in Fig. 4.16. The failure mode of the high viscosity specimen was typically bond
failure where the adhesive fails at both temperatures 50C [122F] and 200C [392F] as
shown in Fig. 4.16(a)-(b). The other two types of adhesives, low viscosity and high
temperature failure was within the concrete bonding. The concrete prisms bonded with
high viscosity adhesive have a thickness almost double the concrete prisms bonded with
high temperature. A relationship can be drawn from the thickness of the adhesive and the
normalized bond failure that the thickness of the adhesive subjected to thermal distresses
can be an interesting topic for future research.
86


4.4 Summary and Conclusion
This Chapter has discussed the bond performance of the concrete-adhesive interface at
elevated temperatures. The concrete prisms were subjected to three hours of a
temperature ranging from 25C [77F] to 200C [392F], Three different types of
adhesive were used to bond the non-reinforced concrete prisms: (1) high viscosity
adhesive; (2) low viscosity adhesive; (3) high temperature adhesive. The author used the
normalized capacity that utilized the ultimate capacity over the thickness of the adhesive,
in other words the adhesive bonded volume was used to normalize the capacity not the
bonded area due to thickness variation among the specimens between different types of
adhesive and between the same adhesive type as well. Since the glass transient
temperature was not reflected by the concrete prisms, additional material testing was
performed to check the Tg using the coupon testing and the differential scanning
calorimetry.
According to the temperature dependent concrete prisms load displacement
results, the specimen bonded with high temperature maintained the highest bond
performance capacity. The load displacement graph was linearly decreasing with
temperature.
The thickness of the adhesive was used to normalize the ultimate capacity instead
of using the bonded area alone due to a non-uniform quantity of adhesive used to
bond the concrete prisms and other mentioned factors.
87


Table 4.1. Properties of low-viscosity adhesive from the manufacturer (MBrace, 2007)
Property Tensile Value Compressive Value Flexural Value
Yield Strength 54 MPa [7,900 psi] 86.2 MPa [12,500 psi] 138 MPa [20,000 psi]
Strain at Yeild 2.50% 5.00% 3.80%
Elastic Modulus 3034 MPa [440 ksi] 2620 MPa [380 ksi] 3724 MPa [540 ksi]
Ultimate Strength 55.2 MPa [8,000 psi] 86.2 MPa [12,500 psi] 138 MPa [20,000 psi]
Rupture Strain 3.50% 5.00% 5.00%
Table 4.2. Temperature dependent residual strength of adhesive coupons: (HV) high viscosity adhesive
coupons; (LV) low viscosity adhesive coupons; (HT) high temperature adhesive coupons
HV LV HT
Temp Stress Stress Stress
C MPa MPa MPa
25 22.2 57.8 21.7
25 15.8 55.4 20.0
25 N/A 55.4 22.6
50 14.5 50.4 28.7
50 22.2 65.9 32.7
50 19.5 69.2 20.7
75 34.8 60.1 12.0
75 26.6 70.8 12.2
75 31.2 73.0 29.4
100 28.5 67.2 35.0
100 16.1 71.9 24.9
100 29.6 64.7 24.4
125 28.8 75.5 28.6
125 20.0 79.6 15.7
125 36.6 50.8 14.0
150 N/A 60.7 26.8
150 23.0 47.8 20.8
150 34.8 76.9 38.2
175 38.2 45.5 28.0
175 35.0 19.6 14.5
175 15.6 49.3 18.1
200 14.8 47.8 36.0
200 32.9 23.2 23.2
200 17.8 72.7 22.3
88


Table 4.3. Average dynamic elastic modulus response at elevated temperatures
Temp Mbrace T-308 High-temp
LV HV HT
DEM (GPa) Average DEM (GPa) Average DEM (GPa) Average
25 22.55 21.09 20.06 20.44 21.89 20.44
25 21.32 20.06 19.96
25 19.36 18.21 18.12
25 20.12 22.00 21.89
25 22.74 20.65 18.93
25 22.74 22.22 22.42
25 19.20 20.65 18.93
25 21.93 21.43 21.62
25 21.63 20.48 20.26
25 20.84 20.48 20.26
25 22.04 20.87 18.76
25 18.55 18.23 22.22
50 21.42 19.47 19.47 18.83 20.23 19.64
50 21.81 19.10 19.85
50 20.24 19.47 19.47
50 18.36 19.10 19.10
50 20.73 19.55 19.60
50 20.34 19.18 19.60
50 19.96 19.18 18.51
50 18.83 18.07 19.60
50 17.81 18.58 20.22
50 18.18 18.21 20.22
50 18.18 17.84 19.84
50 17.81 18.21 19.46
75 19.39 19.55 18.06 17.98 21.25 20.94
75 19.39 17.70 20.47
75 19.01 18.06 20.86
75 18.63 18.06 20.47
75 18.80 17.78 20.82
75 18.07 17.78 20.42
75 18.07 18.14 20.42
75 17.71 17.78 20.82
75 21.46 18.19 21.75
75 21.07 18.19 21.35
75 21.46 18.19 21.35
75 21.48 17.83 21.34
89


Table 4.3. Average dynamic elastic modulus response at elevated temperatures (cont.)
Temp Mbrace T-308 High-temp
LV HV HT
DEM (GPa) Average DEM (GPa) Average DEM (GPa) Average
100 17.87 18.12 15.23 15.30 20.43 20.22
100 17.52 17.31 21.21
100 17.17 13.61 20.82
100 17.52 14.25 18.19
100 16.43 14.70 20.65
100 16.43 12.52 20.26
100 16.10 16.70 21.04
100 16.10 17.39 18.75
100 20.38 14.45 20.33
100 20.77 17.55 20.71
100 20.38 15.45 19.95
100 20.77 14.45 20.33
125 15.80 16.23 17.67 15.72 17.77 19.73
125 19.38 16.59 19.96
125 14.15 15.55 20.34
125 18.64 14.54 19.58
125 17.13 15.07 20.78
125 15.07 15.07 20.39
125 14.10 14.10 18.86
125 15.74 15.74 18.86
125 18.85 16.38 20.91
125 16.71 17.11 20.53
125 13.76 15.06 18.29
125 15.36 15.73 20.53
150 16.68 11.78 12.12 12.71 15.59 16.37
150 17.03 12.12 15.59
150 16.68 11.84 14.63
150 16.68 12.12 15.91
150 8.84 13.27 16.17
150 9.09 13.57 15.48
150 8.59 13.57 15.48
150 8.10 13.57 16.17
150 10.27 12.28 17.95
150 10.27 13.48 17.95
150 10.53 12.28 17.95
150 8.53 12.28 17.60
90


Table 4.3. Average dynamic elastic modulus response at elevated temperatures (cont.)
Mbrace T-308 High-temp
Temp LV HV HT
DEM (GPa) Average DEM (GPa) Average DEM (GPa) Average
175 11.42 13.04 14.06
175 11.14 13.35 13.15
175 10.87 13.04 13.15
175 11.14 13.66 13.15
175 14.36 14.71 16.24
175 14.03 12.36 15.04 12.77 16.24 15.35
175 13.71 14.39 16.58
175 14.03 14.71 16.24
175 12.11 10.53 16.59
175 11.82 10.26 16.25
175 11.53 9.99 16.25
175 12.11 10.53 16.25
200 7.52 8.94 9.29
200 7.07 9.69 9.54
200 7.29 9.19 9.54
200 7.52 12.41 9.54
200 7.32 9.00 9.25
200 6.89 7.36 8.76 9.85 10.28 9.88
200 7.32 9.00 10.54
200 6.89 8.76 9.00
200 7.81 10.68 10.52
200 7.35 10.40 10.26
200 7.58 10.68 10.52
200 7.81 10.68 10.26
91


Table 4.4a. Temperature dependent adhesive-concrete interfacial capacity of high viscosity adhesive
High Viscosity Adhesive
Temp Pu Average MPa Average Thickness Average Normalized Normalized
C (kN) (KN) (MPa) mm Thickness MPa/mm Average
25 2.92 0.50 1.60 0.31
25 3.34 3.08 0.57 0.52 1.90 1.67 0.30 0.31
25 2.98 0.50 1.50 0.33
50 5.81 0.95 2.20 0.43
50 4.25 4.32 0.73 0.73 1.80 1.90 0.41 0.38
50 2.90 0.51 1.70 0.30
75 4.04 0.66 2.10 0.32
75 3.19 2.97 0.54 0.50 1.80 1.73 0.30 0.28
75 1.68 0.30 1.30 0.23
100 0.14 0.02 1.40 0.02
100 2.75 1.96 0.45 0.32 2.00 1.80 0.23 0.16
100 2.99 0.48 2.00 0.24
125 2.34 0.39 2.10 0.19
125 2.19 2.41 0.38 0.41 2.20 2.10 0.17 0.19
125 2.72 0.45 2.00 0.22
150 2.61 0.43 2.00 0.21
150 0.98 1.93 0.16 0.32 1.90 1.97 0.08 0.16
150 2.18 0.36 2.00 0.18
175 3.16 0.51 2.10 0.24
175 0.92 1.77 0.16 0.29 1.90 2.00 0.08 0.14
175 1.23 0.20 2.00 0.10
200 1.41 0.24 2.10 0.11
200 1.55 1.26 0.26 0.21 2.30 2.13 0.11 0.10
200 0.84 0.14 2.00 0.07
92


Table 4.4b. Temperature dependent adhesive-concrete interfacial capacity of low viscosity adhesive
Low Viscosity Adhesive
Temp Pu Average MPa Average Thickness Average Normalized Normalized
C (kN) (KN) (MPa) mm Thickness MPa/mm Average
25 4.83 0.77 1.40 0.55
25 2.98 3.81 0.50 0.62 1.20 1.33 0.41 0.46
25 3.63 0.59 1.40 0.42
50 5.60 0.95 1.70 0.56
50 5.54 4.92 0.94 0.83 1.80 1.67 0.52 0.49
50 3.62 0.60 1.50 0.40
75 4.04 0.67 1.60 0.42
75 3.19 2.97 0.54 0.50 1.50 1.47 0.36 0.33
75 1.68 0.27 1.30 0.21
100 1.69 0.28 1.50 0.19
100 3.99 2.54 0.64 0.42 1.70 1.60 0.38 0.26
100 1.94 0.32 1.60 0.20
125 4.42 0.74 1.70 0.43
125 1.50 2.62 0.25 0.44 1.50 1.60 0.17 0.27
125 1.95 0.33 1.60 0.21
150 2.31 0.38 1.60 0.24
150 1.10 1.81 0.19 0.30 1.60 1.60 0.12 0.19
150 2.04 0.34 1.60 0.21
175 1.85 0.31 1.80 0.17
175 2.75 2.15 0.44 0.35 1.90 1.83 0.23 0.19
175 1.87 0.30 1.80 0.17
200 3.30 0.55 2.00 0.27
200 0.46 1.72 0.08 0.29 1.90 1.93 0.04 0.15
200 1.41 0.24 1.90 0.12
93


Table 4.4c. Temperature dependent adhesive-concrete interfacial capacity of high temperature adhesive
High Temperature Adhesive
Temp Pu Average MPa Average Thickness Average Normalized Normalized
C (kN) (KN) (MPa) mm Thickness MPa/mm Average
25 2.10 0.42 0.90 0.47
25 2.99 2.01 0.50 0.37 0.80 0.77 0.62 0.47
25 0.93 0.19 0.60 0.32
50 2.77 0.47 1.00 0.47
50 2.21 3.09 0.38 0.51 0.90 1.03 0.42 0.49
50 4.31 0.69 1.20 0.58
75 4.00 0.67 1.60 0.42
75 3.40 3.96 0.58 0.66 1.30 1.47 0.44 0.45
75 4.50 0.73 1.50 0.48
100 2.36 0.39 1.20 0.33
100 2.31 2.22 0.39 0.37 1.00 1.10 0.39 0.34
100 1.99 0.33 1.10 0.30
125 2.36 0.39 1.10 0.36
125 2.35 2.36 0.39 0.39 0.90 1.10 0.44 0.37
125 2.36 0.39 1.30 0.30
150 1.91 0.31 0.90 0.34
150 0.96 1.26 0.16 0.21 0.60 0.70 0.27 0.29
150 0.93 0.15 0.60 0.26
175 1.06 0.18 0.90 0.20
175 0.92 1.52 0.15 0.25 0.90 0.97 0.17 0.25
175 2.57 0.41 1.10 0.38
200 0.11 0.02 0.50 0.04
200 1.06 1.01 0.17 0.17 0.90 0.83 0.19 0.17
200 1.86 0.32 1.10 0.29
94


Figure 4.1. Concrete beam mold being prepared for concrete casting
r
HV 25C cd HV 200C

LV 25C * 1 LV 200C

HT 25C HT 200C
Figure 4.3. Adhesive coupons: (a) during testing; (b) before and after thermal exposure
95


Stress (MPa)
80 -I
70 -
60 -
50 -
40
30 -
20
10 -
0 --
0
A Individua HV
----Average HV
A
A
A
A
V A
A
A
&
A
1
25 50 75 100 125 150 175 200
Temperature (C)
(a)
co
Q.
2
w
80
70
60
50
40
30
20
10
0
X
A
80
70
60
50
40 -
30 -
20
10
0
A A
A A A N ,-A
Individual LV A A
Average LV
50 75 100 125 Temperature (C) 150 175 200
25
A Individual HT
.....Average HT
(b)
A
A

A A
/frs. A
A V A
0 25 50 75 100 125 150 175 200
Temperature (C)
(c)
Figure 4.4. Temperature dependent residual capacity of adhesive coupons: (a) high viscosity adhesive
coupons; (b) low viscosity adhesive coupons; (c) high temperature adhesive coupons
(a) (b)
Figure 4.5. Test setup: (a) specimen details (not to scale); (b) test set-up configuration using the MTS
machine
96


Figure 4.6. Temperature data-logger
97


Temperature (C) Temperature (C)
Temperature
Time (hrs)
(a)
200 -I
175 -
150
125
100 -
75 -
50 -
25
0 -
Temperature
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Time (hrs)
(c)
4.0
200
175
150
125
100
75
50
25
0
Temperature
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(e)
200
~ 175
150
| 125
l 100
a
E 75
|2
50
25
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(b)
200
175 -
150
125 --------------------------------,
100 -
75 -
50 -
25 Temperature
0 -I----1----,----,----1----,----,----,----,
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(d)
o
o
0)
200
175
150
125
100
75
50
25
0
-Temperature
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
Time (hrs)
(f)
98


200
175
0
^ 150

= 125
15
a 100
a
1 75
50
25
0
Temperature
0.0 0.5 1.0
1.5 2.0 2.5
Time (hrs)
(g)
3.0 3.5 4.0
O
o
£
3
5
<5
Q.
E
200
175
150
125
100 -
75 -
50
25
0
Temperature
200C
200C
175C
150C
100C
75C
50C
0.0 0.5 1.0
1.5 2.0 2.5
Time (hrs)
(h)
3.0 3.5 4.0
Figure 4.7. Temperature recording acquired from the data logger: (a) at 50C [122F]; (b) at 75C [167F];
(c) at 100C [212F]; (d) at 125C [257F]; (e) at 150C [302F]; (f) at 175C [347F]; (g) at 200C
[392F]; (h) combined temperatures
99


Full Text

PAGE 1

AN EXPERIMENTAL INVESTIGATION INTO THE BEHAVIO R OF CONCRETE ELEMENTS REROFITT ED WITH NSM COMPOSITE STRIPS AT ELEVATED TEMPERATURES By ABDUL RAHMAN NAMROU B.S., North Dakota State University, 2011 A thesis submitted to the Faculty of the Graduat e School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Science Civil Engineering 2013

PAGE 2

2013 ABDUL RAHMAN NAMROU ALL RIGHTS RESERVED

PAGE 3

ii This thesis for the Master of Science de gree by Abdul Rahman Namrou has been approved for the Civil Engineering Program By Yail Jimmy Kim, Chair Kevin Rens Frederick Rutz November 12, 2013

PAGE 4

iii Namrou, Abdul Rahman. (M.S. Civil Engineering) An Experimental Investigation into the Behavi or of Concrete Elements Retrofitted with NSM Composite Strips at Elevated Temperature Thesis directed by Associate Professor Dr. Yail Jimmy Kim A BSTRACT Near surface mounted (NSM) fiber reinforced polymer (FRP) is another strengthening alter native of ex ternally bonded fiber reinforced polymers NSM FRP is a promising alternative technology that has emerged for enhancing the strength capacity of concrete structures. Most laboratory researches have focused mainly on the overall member performance and/or th e bonding performance of the NSM bars or strips. Limited research has focused on the effect of temperature exposure on NSM FRP performance. The results of an experimental program performed on forty eight (48) concrete block specimen with NSM carbon fiber r einforced polymer (CFRP) strengthening systems at elevated temperatures that reaches to 200C [392F] to investigate flexural performance. The effect of using two different adhesive systems (epoxy anchoring system) with manufacturer recommendation at ordin ary and high temperature exposures is also studied. The adhesive was injected in a NSM groove size (25 mm [1 in] deep 13 mm [0.5 in] wide) the width and depth of the groove were greater than 3 and 1.5 times the CFRP thickness and wid th, respectively. Test results show that the interfacial strength of the specimens bonded with the ordinary epoxy is maintained until 75C [167F] is reached, while the strength noticeably decreases with an increasing temperature above this limit. The spec imens with the high temperature epoxy preserve interfacial capacity up to 200C

PAGE 5

iv [392F] despite a trend of strength decrease being observed. The failure of the test specimens is brittle irrespective of adhesive type. Interfacial damage is localized along t he bond line with the presence of hairline cracks that further develop when interfacial failure is imminent. This thesis also presents an experimental result concerning the bond performance of concrete adhesive at elevated temperatures that reaches to 200 C [392F] applied for three hours Then, the concrete prisms were tested under three point flexural loading The experimental pro gram is comprised of seventy two (72 ) specimens bonded with low viscosity, high viscosity adhesives and high temperatur e adhes ive and their comparative performance is of interest in the present investigation. Emphasis is placed on the residual capacity of the conditioned bond concrete interface and corresponding failure mode. For high temperature exposure, it is shown that the hi gh temperature lami nated adhesive outperforms the high and low viscosity adhesive s by remaining fairly consistent and allowing the strengthening system to remain effective for up to three hours of 200C [392F]. The form and content of this abstract are a pproved. I recommend its publication. Approved: Jimmy Kim

PAGE 6

v ACKNOWLEDGEMENTS This thesis would not hav e been possible without the support of so many people in many forms It was the product of a large measure of serendipity with people who have changes the course of my life. I truly believe t hat every person I come across with have some effect on my life be it the most apparently insignificant way. On behalf of this, I must also acknowledge all those of whom I have come across with and have led me to the path that I chose. Some have had much larger roles than others. First, I must take to opportunity to acknowledge and thank my parents for their endless love and support. They have implanted in me high work ethics and confidence to prosper. I would like to thank my sisters for standing by my parents when I am gone. With this believe in mind, It is with immense gratitude that I acknowledge the deepest appreciation to my advisor Associate Professor Yail Jimmy Kim, for his patience, vision and guidance to p ursue my goals. Without his encouragement, I would never have Also, I would like to acknowledge the department chair and committee chair of civil engineering Dr. Kevin Rens f or generous support and making th e transition from North Dakota State University easy. With much r espect, I would like to thank Dr Khalil Jarrar J.D. for his moral support; I learned from him that there is no limit for educational content.

PAGE 7

vi This thesis work was supported by the United S tates Department of Transportations through the Mountain Plains Consortium Program and t he University of Colorado

PAGE 8

vii TABLE OF CONTENTS Chapter 1. Introduction ................................ ................................ ................................ .................... 15 1.1 General ................................ ................................ ................................ ......................... 15 1.2 Research Significance ................................ ................................ ................................ .. 19 1.3 Scope ................................ ................................ ................................ ............................ 20 1.4 Thesis Organization ................................ ................................ ................................ ..... 22 2. Literature Review ................................ ................................ ................................ ........... 24 2.1 Near Surface Mounted FRP Strengthening ................................ ................................ 24 2.1.1 FRP Techniques ................................ ................................ ................................ ........ 24 2.1.2 FRP Applications ................................ ................................ ................................ ...... 25 2.2 Background of Nea r Surface Mounted Techniques ................................ .................... 25 2.2.1 FRP Failure Modes ................................ ................................ ................................ ... 27 2.2.2 Bonding Failure ................................ ................................ ................................ ........ 27 2.2.3 Strengthened Capacity of the Element ................................ ................................ ...... 28 2.2.4 NSM Groove Spacing and Depth Recommendation ................................ ................ 29 2.2.5 Int erfacial Behavior of NSM CFRP Bonded to a Concrete Substrate ...................... 30 2.2.6 Failure Modes of NSM Bonded Specimens ................................ .............................. 31 2.3 High Temperat ure Effect ................................ ................................ ............................. 32 2.3.1 Effect of Elevated Temperatures on CFRP Materials ................................ ............... 32 2.3.2 Behavior of Concrete Members Strengthened with NS M CFRP at High Temperature ................................ ................................ ................................ ....................... 33 2.4 Performance Based Fire Safety Design ................................ ................................ ....... 34 2.4.1 History of Performance Based Codes ................................ ................................ ....... 34

PAGE 9

viii 2.4.2 Advantages an d Disadvantages of Prescriptive Based Design ................................ 36 2.4.3 Advantages a nd Disadvantages of Performance Based Design ............................... 37 2.4.4 Development and Evaluation of Design Fire ................................ ............................ 39 2.4.5 Importance Factor Approach ................................ ................................ .................... 39 2.5 Conclusion ................................ ................................ ................................ ................... 40 3. Behavior of Near Surface Mounted CFRP Concrete Interface in Thermal Distress .... 48 3.1 Fa brication and Specimen Design ................................ ................................ ............... 48 3.1.1 General Overview ................................ ................................ ................................ ..... 48 3.1.2 Preparation ................................ ................................ ................................ ................ 49 3.1.3 Materials ................................ ................................ ................................ ................... 50 3.1.4 Test Specimen ................................ ................................ ................................ ........... 51 3.1.5 Thermal Exposure ................................ ................................ ................................ ..... 52 3.1.6 Test Setup and Instrumentation ................................ ................................ ................ 53 3.2 Material Level Testing ................................ ................................ ................................ 53 3.2.1 Dynamic Mechanical Analysis ................................ ................................ ................. 53 3.2.2 Concrete Testing under Thermal Distress ................................ ................................ 54 3.2.3 CFRP Strip Testing under Thermal Distress ................................ ............................ 54 3.2.4 Ordinary Epoxy and High Temperature Coupons ................................ .................... 55 3.3 Experimental Results ................................ ................................ ................................ ... 55 3.3.1 Analys is of Variance for Concrete at Elevated Temperatures ................................ .. 55 3.3.2 Interfacial Strength of NSM CFRP ................................ ................................ ........... 56 3.3.3 Load D isplacement Respon se ................................ ................................ ................... 58 3.3.4 Failure Mode ................................ ................................ ................................ ............. 59

PAGE 10

ix 3.4 Summary and Conclusion ................................ ................................ ............................ 60 4. Bon d Performance of Concrete Adhesive Interface at Elevated Temperatures ............ 77 4.1 Fabrication and Specimen Design ................................ ................................ ............... 77 4.1.1 General O verview ................................ ................................ ................................ ..... 77 4.1.2 Concrete and Adhesive Material Properties ................................ ............................. 78 4.1.3 Testing Specimen and Thermal Exposure ................................ ................................ 79 4.1.4 Instrumentation ................................ ................................ ................................ ......... 80 4.2 Material Level Testing ................................ ................................ ................................ 81 4.2.1 Stress strain Response of Adhesive Resin ................................ ................................ 81 4.2.2 Storage Modulus of Adhesive Resin ................................ ................................ ......... 82 4.2.3 Frequency Test ................................ ................................ ................................ .......... 82 4.3 Experimental Results ................................ ................................ ................................ ... 83 4.3.1 Dynamic Elastic Modulus of Concrete Prisms ................................ ......................... 83 4.3.2 Load Displacement Response of the Bond Test Concrete Prisms ............................ 84 4.3.3 Failure Mode of Bonded Concrete Prisms ................................ ................................ 86 4.4 Summary and Conclusion ................................ ................................ ............................ 87 5. Structural Concrete Performance under Fire ................................ ............................... 105 5.1 Introduction ................................ ................................ ................................ ................ 105 5.2 Importance Factor Evaluation Approach ................................ ................................ ... 106 5.2.1 Importance Factor Calculation ................................ ................................ ................ 107 5.2.2 Different Type Contribution to t he Overall Influence Factor ................................ 109 5.3 Importance Factor Analysis for NSM CFRP Concrete Interface Interaction ............ 113 5.4 Importance Fa ctor Analysis for the Bond Performance Concrete adhesive Interface 114

PAGE 11

x 5.5 Utilization of the Importance Factor into Performance based Approach .................. 116 5.6 Conclusion ................................ ................................ ................................ ................. 118 6. Conclusion and Recommendation ................................ ................................ ............... 123 6.1 Recommendation for Future Work ................................ ................................ ............ 127 References ................................ ................................ ................................ ........................ 128 Appendix A ................................ ................................ ................................ ................................ ...... 128 B ................................ ................................ ................................ ................................ ....... 1 5 2 C ................................ ................................ ................................ ................................ ...... 1 7 0 D ................................ ................................ ................................ ................................ ...... 1 7 2 E ................................ ................................ ................................ ................................ ...... 1 7 3 F ................................ ................................ ................................ ................................ ....... 1 8 1

PAGE 12

xi LIST OF TABLES Table 2.1. FRP Te (Adapted from ACI 440.2R 08) 4 2 2.2. Bridge features weightage factors (Kodur and Naser 2013) ................................ ....... 43 2.3. Risk grades w.r.t the overall class coefficient and the importance factor (Kodur and Naser 2013) ................................ ................................ ................................ ........................ 44 3.1. Properti es of CFRP strip from the manufacturer (Hughes Brothers 2008) ................ 62 3.2. Test Specimens load & failure mode of ordinary epoxy & high temperature epoxy 63 3.3 Compressive concrete strength depending up on temperature (Kim et al. 2012) ........ 64 4.1. Properties of low viscosity adhesive from the manufacturer (MBrace, 2007) ........... 88 4.2. Temperature dependent stress response of adhesive coupons: (HV) high viscosity adhesive coupons; (LV) low viscosity adhesive coupons; (HT) high temperature adhesive coupons ................................ ................................ ................................ .............................. 88 4.3. Dynam ic elastic modulus response over the specified temperature ........................... 89 4.4 a Temperature dependent adhesive concrete interfacial capacity of high viscosity adhesive ................................ ................................ ................................ .............................. 92 4.4b. Temperature dependent adhesive concrete interfacial capacity of low viscosity adhesive ................................ ................................ ................................ .............................. 9 3 4.4c. Temperature dependent adhesive concrete interfacial capacity of high temperature adhesive ................................ ................................ ................................ .............................. 9 4 5.1. Parameters used to predict the overall class coefficient for NSM CFRP concrete blocks ................................ ................................ ................................ ............................... 120 5.2. Parameters used to predict the overall class coeffici ent for bond performance concrete prisms ................................ ................................ ................................ ................ 1 21 5.3. Importance factor table criteria based on Kodur and Naser (2013) propose ............ 1 21

PAGE 13

xii LIST OF FIGURES Figure 2.1. Pre cured system aramid type FRP installation (Carmichael and Barnes 2005). ....... 44 2.2. Groove dimen sion limit ations (ACI 440.2R 08, 2008) ................................ .............. 44 2.3. Development of Design Fires (Custer 1995) ................................ .............................. 45 2.4. Evaluation of Alt ernative Designs (Custer 1995) ................................ ....................... 46 2.5. The influencing characteristics of fire hazard in bridges (Kodur and Naser 2013) .... 46 2.6. The contri bution to overall influence factors from different classes (Kodur and Naser 2013) ................................ ................................ ................................ ................................ .. 47 3.1. Wooden mold in whi ch concrete blocks were for med ................................ ................ 65 3.2. NSM CFRP concrete interface test spe cimen dimension (not to scale) ..................... 65 3.3. Interface test specimens af ter applying an epoxy adhesive ................................ ........ 65 3.4. Test specimen: (a) air blasting before CFRP bonding; (b) CFRP installation ........... 66 3.5. Test setup: (a) high temperature exposure; (b) conditio ned specimen; (c) tension test 66 3.6. Ther m ocouple temperature readings over a thr ee hour range acquired from the data logger: (a) at 50C [122F]; (b) at 75C [167F]; (c) at 100C [212F]; (d) at 125C [257F]; (e) at 150C [302F]; (f) at 175C [347F]; (g) at 200C [392F] ...................... 68 3.7. Dynamic M echa nical A nalysis (DMA); (a) testing mach ine; (b) DMA clamps; (c) DMA specimens ................................ ................................ ................................ ................. 68 3.8. Dynamic Mechanical Analysis (DMA) results: (a) CFRP strip; (b) high temperature; (c) ordinary epoxy ................................ ................................ ................................ .............. 69 3.9. Temperature dependent strength measured: (a) concrete in compression; (b) CFRP strip in tension; (c ) ordinary epoxy in tension; (d) high temperature epoxy in tension ; (e) normalized strength of OE; (f) normalized strength of HE ................................ .............. 70 3.10 Failure mode: (a) HE epoxy coupons at 200C [392F]; (b) OE epoxy coupons at 200C [392F]; (c) CFRP strips at 200C

PAGE 14

xiii 3.1 1 Temperature dependent interfacial strength: (a) specimens bonded with ordinary epoxy; (b) specimens bonded with high temperature epoxy; (c) temperature dependen t average interfacial strength ................................ ................................ ................................ 71 3.12 Load displacement of specimens bonded with ordinary epoxy: (a) at 25C [77F]; (b) at 50C [122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C [257F]; (f) at 150C [302F]; (g) at 175C [347F]; (h) at 200C [392F] ................................ ......... 73 3.1 3 Load displacement of specimens bon ded with high temperature epoxy: (a) at 25C [77F]; (b) at 50C [122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C [257F]; (f) at 150C [302F]; (g) at 175C [347F]; (h) at 200C [392F] ..................... 75 3.14 Load displacement comparison of exposure t emperatures: (a) bonded with ordinary epoxy; (b) bonded with high temperature epoxy ................................ ............................... 75 3.15. Failure mode: (a) interface failure of ordinary epoxy ; (b) interfacial failure of ordinary epoxy after test; (c) interfacial failure of hig h temperature epoxy aft er test ....... 75 3.1 6 Failure mode: (a) ordinary epoxy at 50C; (b) ordinary epoxy at 200C; (c) high temperature epoxy at 50C; (d) h igh temperature epoxy at 200C ................................ ... 7 6 4.1. Concrete beam mold being prepared for concrete casting ................................ .......... 95 4.2. Adhesive coupon prepar ation and casting: (a) mold shape; (b) high viscosity epoxy; (c) low viscosity epoxy ................................ ................................ ................................ ...... 95 4.3. Adhesive coupon: (a) during testing; (b) before and after thermal exposure ............. 95 4.4. Temperature dependent residual capacity of adhesive coupons : (a) high viscosit y adhesive coupons; (b) low viscosity adhesive coupons; (c) high temperature adhesive coupons ................................ ................................ ................................ .............................. 96 4.5 Test set up: (a) specimen details (not to scale); (b) test set up configuration using the MTS machine ................................ ................................ ................................ ..................... 96 4.6 Temperature data logger ................................ ................................ ............................. 97 4.7 Temperatu re recording acquired from the data logger: (a) at 50C [122F]; (b) at 75C [167F]; (c) at 100C [212F]; (d) at 125C [257F]; (e) at 150C [302F]; (f) at 175C [347F]; (g) at 200C [392F] ; (h) combined temperatures ................................ ... 99 4.8 Electric stationary diamo nd saw ................................ ................................ ................. 99 4.9 DMA results of adhesive: (a) low viscosity; (b) high viscosity; (c) high temperature ................................ ................................ ................................ ................................ .......... 100

PAGE 15

xiv 4. 10 Frequency test: (a) frequency test set up; (b) frequency test reading ..................... 10 0 4.11 Residual impact resonant frequency of interface test speci men (dot = individual; line = average): (a) frequency response; (b) temperature dependent dynamic elastic modulus ................................ ................................ ................................ ............................ 10 1 4.12 High viscosity adhesive concrete prism specimens: (line = individual; dot = average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity .................... 10 1 4.1 3 Low viscosity adhesive concrete prism specimens: (line = individual; dot = average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity .................... 10 2 4.14 High temperature adhesive concrete prism specimens: (line = in dividual; dot = average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity .................... 10 3 4.15 Normalized bond performance concrete prism specimens capacity ........................ 10 3 4.16 Failure mode of concrete prisms bonded with different types of adhesive: left spec imen (LV); central specimen (HV); right specimen (HT): (a) at 50C [122F]; (b) at 200C [392F] ................................ ................................ ................................ .................. 104 5.1. Overall class coefficient to assign fire risk grade of NSM CFRP concrete blocks .. 1 22 5. 2 Importance factor based on the overall class coefficient of NSM CFRP concrete interface: (a) for HE epoxy; (b) for OE epoxy ................................ ................................ 122 5.3 Overall class coefficient to assign fire risk grade of bond performance concrete adhesive interface ................................ ................................ ................................ ............. 1 23 5. 4 Importance factor based on the overall class coefficient of bond perfor mance adhesive concrete interface: (a) for HT epoxy; (b) for LV epoxy; (c) for HV epoxy ..... 123

PAGE 16

15 1. Introduction 1.1 General Structural strengthening is frequently required for upgrading the performance of deteriorated concrete structures. Externally b onded carbon fiber reinforced polymer (CFRP) sheets have been broadly used, while an alternative technique called the near surface mounted (NSM) method is emerging (De Lorenzis and Teng 2007). The NSM method employs CFRP strips or rods inserted into the na rrow precut groove of a concrete member and permanently bonded with an adhesive. As is for the case of the externally bonded CFRP sheet application, epoxy adhesives are dominantly used for bonding NSM CFRP reinforcement. NSM strengthening techniques may re quire more initial costs than conventional externally bonded CFRP sheet application (Aidoo et al. 2006); however, their superior bond and durability can provide better long term performance. A number of laboratory scale projects have been carried out to ex amine the effect of NSM CFRP on enhancing structural behavior in flexure ( El Hacha et al. 2004; Hassan and Rizkalla 2004 ) and in shear (Chaallal et al. 2011). Field demonstration has been undertaken as well. Alkhrdaji et al. (1999) performed a strengthenin g project using a decommissioned reinforced concrete slab bridge with multiple spans. CFRP strengthening was intended to increase the flexural capacity of the strengthened span by 30% in comparison to that of an unstrengthened span. Test results revealed t hat both strength and stiffness of the bridge were improved due to the presence of NSM CFRP. Stone et al. (2002) strengthened a simply supported solid slab bridge. Prior to installing NSM CFRP strips, a non

PAGE 17

16 destructive test was done to examine the distribu tion of internal reinforcing steel. Load tests using a truck were conducted before and after the strengthening work to assess the efficacy of the NSM CFRP. Fire is a latent problem for all constructed structural members because thermal stress can signifi cantly degrade their functionality. Such an extreme event has been an issue for residential or office buildings over several decades. Given infrastructure systems are concerned with transporting hazard materials (e.g., flammable or combustible commodities) bridge structures are expos ed to potential fire hazards (Ga rlock et al. 2012). Fire resistance of existing members needs to be ensured so that catastrophic events can be precluded. This crucial requirement is valid for CFRP strengthened structures. Accor ding to a literature search (to be discussed in the Background section), the rehabilitation community is aware of the significance of thermally induced detrimental effects associated with CFRP application. Extensive effort has been done to investigate the behavior of CFRP strengthened concrete structures subjected to high temperature. The majority of research is dedicated to concrete members with externally bonded CFRP sheets (Foster and Bisby 2008). The application of NSM CFRP in this circumstance is limit edly studied ( Palmieri et al. 2012 ). All existing research programs are based on CFRP composites in conjunction with epoxy adhesives that are fundamentally susceptible to the degree of thermal exposure. The interaction between the NSM CFRP concrete and adh esive concrete in a thermal environment is another area to be explored, provided no research has been reported earlier. This thesis addresses these identified technical challenges and presents an experimental program concerning the interfacial behavior of

PAGE 18

17 NSM CFRP strips embedded in a concrete substrate and bonding adhesive to concrete subjected to elevated temperatures Foci of the study are bond deterioration, residual capacity, and failure characteristics. A pioneering endeavor is made to examine the the rmal performance of NSM CFRP concrete interface bonded with a high temperature epoxy that can overcome the limitation of ordinary epoxy adhesives. The results discussed here are taken from a test program that is devoted to understanding the fundamental the rmomechanical behavior of NSM CFRP for strengthening concrete structures in thermally induced stress states. Based on the 2013 ASCE Report Card on (ASCE, 2013) GPA is D + Being more co ncise, one ninth of North America bridges are classified as structurally deficient with a C + grade with an average age of currently 42 years and that an estimate investment of $20.5 billion per year for the next 15 years Over the past 20 years, the civil engineering applications has developed the use of fiber reinforced polymers (FRPs), and FRPs are now providing a number of innovative approaches for both new constructi on, and predominantly for rehabilitation and strengthening of existing structures (Bakis C.E., 2002) The carbon FRP (CFRP) composite is widely preferred for retrofit applications over the glass or aramid FRP (AFRP GFRP). The advantages of CFRP composites consists of resistance to corrosion, high strength and s tiffness, easy and prompt installation, reduced service costs, low density, and sustainable performance (Teng J.G., ICE 2003) (Kim YJ a. H., 2008) The achievement of the stress transfer is provided by the CFRP load carrying capacity and the

PAGE 19

18 matrix transfe r the load to the fibers. CFRP composites may be bonded externally to improve the load carrying capacity of the structure. According to Malek AM et al. and Kim YJ. et al., premature debonding failure due to mechanical or environmental loading may occur on the externally bonded CFRP due to stress concentration at the cut of point (Malek AM, 1998) (Kim YJ F. A. H., 2005) Debonding of CFRP is influenced by the normal and shear stresses along the bondline (Smith ST, 2001) Environmental loading such as wet dry ing cycles, freezing and thawing, and low temperature can deteriorate the concrete bonding interface (Green M.F., 2000) With the above mentioned drawbacks of externally bonded FRP as a strengthening and retrofit of deteriorated structural concrete, near surface mounted (NSM) FRP reinforcement has emerged as a promising alternative technology in both shear and flexure. Along the tensile soffit of the concrete structure, a small groove (slot) is cut and a CFRP composite may be strips, laminates, and bars a re inserted and bonded with an adhesive resin that may be epoxy or other cementitious grouts The NSM CFRP has improved bond characteristics and enhanced durability compared to externally bonded CFRP composites. A greater portion of full strength bonded CF RP is often able to utilize through this method because of the greater bonding characteristics that prevent premature debonding failures. Moreover, because the CFRP strengthening system is located within the member itself, protects it from the environment and fire damage.

PAGE 20

19 1.2 Research Significance A large research effort with the current emergence strengthening technique of NSM CFRP for reinforced concrete members has focused on strength gain with use of composite strips, failure mode, interfacial stres ses between concrete and CFRP at room temperature in both shear and flexural strengthening applications. Minimal research has been reported over the durability issues of the NSM FRP strengthening techniques for concrete. Due to the potential hazard of fire a challenge in CFRP strengthening applications may take place. The performance of the adhesives resins used to bond the CFRP to the structure may be vulnerable to high temperature due to a characteristics change. (Calberger T., 2009) The critical temper ature of an adhesive is called glass transient temperature, T g The effectiveness of CFRP strengthening is dependent upon the temperature exposure level. Although limited research regarding the fire safety of CFRP composite strip concrete strengthening hav e been recently published (Bisby L.A., 2005) there still exists inadequate information in this area. The promising strengthening method requires more experimental and analytical investigations. High temperature provides detrimental stress to CFRP strength ened concrete members because the strengthening system includes polymeric materials that are radically susceptible to thermal exposure. Although NSM CFRP strips are a strong alternative to traditional externally bonded CFRP sheets when upgrading constructe d concrete structures, their thermal response is not sufficiently elucidated in the research community. This research program emphasizes the interfacial behavior of the NSM CFRP embedded in a concrete substrate subjected to elevated temperatures, including a comparative study as to the performance of ordinary and high temperature epoxy adhesives.

PAGE 21

20 This thesis presents an on going research program concerning the performance of NSM CFRP concrete members performance subjected to high temperature. A brief revie w on literature relevant to the research is also provided. The experimental st udy reported here examines the m ateria l characteristics and bond performance of NSM CFRP concrete interface and the adhesive bond concrete interface at elevated temperatures. Th e objectives of this research were: To investigate experimentally the performance of NSM CFRP strengthening systems bonded to concrete wit h different type of adhesive subjected to high temperature exposure for three hours up to 200 C [392F]. As might real istically bridges and existing structures allow for evacuation to take place and to evaluate the potential behavior of these strengthening techniques under the effect of fire. To investigate experimentally the adhesive bond concrete interface subjected to elevated temperatures. To investigate experimentally the mechanical properties and characterization of different type of adhesives for the use of bonding NSM FRP strengthening applications under the same condition of high temperature exposure To propose a p erformance base d design equations for structures under the effect of fire. 1.3 Scope The performed work within this thesis involved experimental testing of forty eight (48) plain concrete blocks made with the same concrete mix design to eliminate stren gth

PAGE 22

21 discrepancies each dimensioned as following (200 mm [8 in] long 100 mm [4 in] deep 100 mm [4 in] wide) having a groove of size (25 mm [1 in] deep 13 mm [0.5 in] wide) along the strengthe ning direction molded where the CFRP strip is inserted and bonded with the different type of adhesives designated as ordinary epoxy known for its industrial name as Power Fasteners T308 + (Power Fastener, 2011), a two components epoxy adhesive anchoring sys tem; and the other as high temperature laminated resin ( PTM&W Industries 2013 ) a two part system epoxy The concrete blocks then oven cured over a temperature range from 25C [77F] to 200C [392F] for three hours. The other study presented in the thesi s is the bond performance of adhesive concrete interface. The performed work was done by molding seventy two (72) concrete prisms each dimensioned as following (400 mm [16 in] long 100 mm [4 in] wide 80 mm [3.15 in] deep). The beams were split in half during the mo lding process by a wooden sheet plate. After proper concrete curing, the specimens were bonded together with the same types of high viscosity adhesive known for its industrial name as Power Fasteners T3 08 + (Power Fastener, 2011), a two components epoxy adhesive anchoring system; low viscosity adhesive known as BASF Mbrace (Mbrace 2007); and high temperature adhesive laminated resin ( PTM&W Industries 2013 ), a two part system epoxy. The concrete prisms wer e cured as the s ame conditions up to 200C [392F] for three hours. A material property of the CFRP strips, adhesive epoxies, and concrete were determined over the temperature range mentioned earlier.

PAGE 23

22 1.4 The sis Organization Literature review on NSM FRP behavior in strengthening applications is presented in chapter 2 The properties, description of some existing structures, and proposed application models in a ddition to some design guidelines are discussed in this chapter. An overview of the experimenta l program of concrete blocks with NSM CFRP is discussed in chapter 3. A lso a detailed b ehavior of NSM CFRP concrete interface in thermal d istress along with the test scheme and material level Chapter 4 presents the bond performance of c oncrete adhesive I nterface at elevated t emperatures Additional mechanical properties testing of different types of adhesives are also tested at high temperatures. Chapter 5 discusses the design recomme ndation of the fire performance based methodology structures along wit h the typical objec tives o f the performance based specifications. Chapter 6 provides a summary from this experimental study and some recommendations for further experiments in this research area based on the evaluation of the results from the testing progr am. Appendix A presents additional detailed pictures of the process for the NSM CFRP concrete block testing that have been done in which you can see the fai lure mode Appendix B presents the bond performance of the concrete adhesive prisms in which you ca n see the failure mode from the pictures.

PAGE 24

23 Appendix C shows the dynamic mechanical analysis test specimens and set up. Appendix D shows the frequency tests performed on concrete prisms and the dynamic elastic modulus in pictures. Appendix E shows the thr ee types of adhesive epoxy coupon testing in pictures along with the fa ilure mode and stress strain figures Appendix F shows the CFRP testing in pictures.

PAGE 25

24 2. Literature Review 2.1 Near Surface Mounted FRP Strengthening 2.1. 1 FRP Techniques The nec essity of structural upgrade has been required to increase the capacity of many existing structures. Because FRP is a strong and light material, it has been extensively used in structural rehabilitation industry. FRP is a s trengthening techniques of conc re te structures that started with externally bonded systems, where sheets that could be fibers in one direction or multi directions are bonded to the externally to the surface of the concrete. Typical there are two methods of strengthening techniques where 1) FRP sheets are saturated on site along with the resin bonded to the concrete surface and called wet overlaid 2) FRP sheets are saturated and cured before site application and then applied to concrete surface with adhesive, this method is called pre cur ed system. Figure 2.1 shows the procured FRP installation of aramid fibers type along the soffit of the structure. With the externally bonded application of FRP, premature debon d ing failure has often occurred due to concentration of stress at the cut of po int (Kim YJ F. A. H., 2005) Researchers have been focused on NSM strengthening technique as an improved method of strengthening existing structures.

PAGE 26

25 2.1.2 FRP Applications The application of FRP to existing structures can be divided into two main catego ries: externally bonded FRP systems and near surface mounted NSM systems. The composite of FRP does not vary and typically involves one fiber type regardless of the application type. The types of FRPs available in the market are Carbon (C) glass (G) or a ramid (A) FRPs (ACI 440 2008) with carbon is the most commonly used. Typically, FRP materials ha ve higher ulti mate tensile strength relatively to steel yield the strength of 420 MPa [60 ksi] and ultimate strength of 520 MPa [75 ksi], however, steel is mo re elastic that FRPs with a typical value of 200 GPa [29,000 ksi] (Bertolotti, Eric A. 2012) As a result, the FRP failure rapture is considered to be brittle failure in tension with an approximate strain of 1 3%; by comparison, whereas steel ruptures at a pproximately 30% (Vasquez 2008). modulus of different types of FRP are shown in table 2.1. 2. 2 Background of Near Surface Mounted Techniques Structural strengthening with NSM reinforcement is one of the most promising techniques, where the original method was found in the literature dating to the mid of (De Lorenzis et al. 2000). S teel rods or bars were used as reinforcement and cement mortar was used as an adhesive (Burke 2008) Nowadays, FRP bars or strips take place of the steel rods or bars strengthening techniques and adhesive resin replaces the

PAGE 27

26 cementitious bonding. The concept of bonding FRPs to a concrete element is simply by inserting FRP rod or strip into a pre cut groove concrete opening and bonding it with adhesive resin. The revised technique of NSM FRP has been used in several applications reinforced concrete silos strengthening in Boston, Myriad Convention Center in Oklahoma City, and deck strengthening of the Naval Pier in San Diego (De Lorenzis et al. 2000). The results have shown higher level of strengthening efficiency, less mechanical damages and aging effects, and less susceptibility for fire damage (Tljsten et al. 2003). The full t esting experim ent was done on a full scale bridge in Missouri, that took over the FRP external bonded sheets system by 10% with respect to strength gain (Alkhrdaji et al. 1999) (Bruke 2008). A lot of researchers have focused ever since on the strength gain, groove chara cteristics and depth, FRP shape, adhesive types and bonding characteristics. In addition to that, researchers consider the overall performance of the structu ral element and the performance based bond. To increase the load carrying capacity NSM strips or b ars can be installed in a parallel grooves that are cut at a specified distance. The FRP bars are available as deformed steel with US corss sectional customary size area as number 3 or 4, while the FRP strips are available in two typical dimensions of 16 m m [0.63 in] wide and 2 to 4.5 mm [0.079 to 0.177 in] thick (Hughes Brothers 20 10)

PAGE 28

27 2.2.1 FRP Failure Modes At the meantime, there is no building code req uirement that addresses the FRP strengthening systems specifically, however, there is design recom mendation for the applications of NSM FRP concrete reinforcing systems. The available literature of the NSM FRP fail ure modes has focused on the three failure categories, debonding resistance failure the strengthened capacity of the element and the depth of the groove within the concrete. If adequate spacing of NSM FRP strips or bars was used, it will result in individual failure planes. If not adequate spacing distance of NSM FRP strip or bars was used, it will result in a singular failure plane and of c ourse decreases the overall load carrying capacity of the structure. 2.2.2 Bonding Failure To be able to create composite action, the adhesive resin comes into play to transfer the stress between the FRP bar or strip reinforcement and the concrete susbs trate (De Lorenzis et al. 2001) Most researchers represent the failure mode of the tensile testing by the beam pull out. El Hacha and Rizkalla (2004) found from eight T beam testing that the FRP has little impact on the deflection due to loading before c racking (elastic range) but improved significantly after cracking. The authors also found that the NSM CFRP strips resulted in higher capacity than NSM CFRP bars where he explains that this behavior is due to premature debonding of the NSM CFRP bars.

PAGE 29

28 Tal jsten and Nordin (2007) also found from their eight six meter long beam testing experiments that the NSM reinforcement did not contribute to the stiffness of the overall beam in the elastic range but after cracking the beam shows stiffer behavior and signi ficant increase in the load carrying capacity. Nine T beams were tested by Hassan et al. (2004) and found that premature debonding between the concrete adhesive interface failure happened in the specimens with bond lengths of 150 and 250 mm [5.9 and 9.8 in] with no improvement in strength in comparison to the control specimen. On the other hand, strength and stiffness gains were observed with bonding length of 500 to 700 mm [19.7 to 27.6 in] (Bruke 2008). Seracino et al (2007) conducted thirty six (36) p ush pull tests of concrete blocks bonded to NSM FRP strips restrained by adhesive inside a middle groove From t his analysis, t he researchers conducted a non linear regression analysis to derive some equations needed to determine the intermediate cracking debonding forces that according to the authors, it is controlled by the dimensions of the NSM FRP strip and the concrete compressive strength. Seracino et al (2007) recommended 200 mm [7.9 in] as a minimum bonded length for the FRP. 2.2.3 Strengthened Ca pacity of the Element Rectangular beams were tested by Barros et al. (2005) in four point bending with different NSM FRP and steel reinforcement. In most samples, failure occurred in the

PAGE 30

29 concrete cover spalling that includes the tension steel reinforcemen t and the NSM FRP strip. In some cases, concrete was pulled away from the internal steel. From th is research, the authors notice s strain increase from 62% to 91% of the ultimate tensile strain and a strength increase of 78% to 96% in comparison to the cont rol beams. They also n oted that the stiffness and cracking moment in creased in the strengthened beams. Teng et al. (2006) conducted similar testing scheme of concrete beams in four point bending like Barros et al. (2005) but with different NSM CFRP bondin g lengths. Stiffer behavior and enhanced strength gains of about 30% and 90% to beams with bonded lengths of 1200 mm to 1800 mm [47.2 to 70.9 in] respectively. 2.2.4 NSM Groove Spacing and Depth Recommendation According to Blaschko (2003) spalling of con crete corner may occur if the FRP reinforcement was placed close to the edge of the concrete element. The ACI Committee 440 (2008) stated the clear spacing of NSM groove strips should be two times the groove depth, and four times the grove depth from the s ide cover Hassan and Rizkalla (2004) stated that the clear spacing of the NSM groove strips is two times the FRP bar diameter, and f our times the FRP bar diameter from t he side cover Kang et al. (2005) specified that 40 mm [1.6 in] is the clear spacing o f the NSM groove strips and the side cover distance. Rashid, Oehlers and Seracino (2008) detailed that 53 mm [2.1 in] is the clear spacing of NSM groove strips and 3.5 times the FRP depth with respect to the side cover distance.

PAGE 31

3 0 ACI 440.2R 08 (2008) secti on 13.3 sets a limit for NSM minimum groove dimension The depth of the groove should be greater or equal to 1.5 times the diameter for FRP bars or maximum length of the FRP strip. The width of the groove is also greater or equal to 1.5 times diameter of t he FRP bars. For FRP strips the groove width is greater or equal of 3 times the thickness of the strip. More bonded surface will develop between the adhesive resin and the concrete when the depth of the FRP embedment increases in the concrete element as sh own in Fig. 2.2 In other words higher debonding strains and greater ductility will be developed (Oehlers et al. 2008). 2.2.5 Interfacial Behavior of NSM CFRP Bonded to a Concrete Substrate The interfacial behavior of NSM CFRP embedded in a concrete subs trate has been of interest for over a decade. Laboratory investigations were conducted using beam specimens (El Hacha et al. 2004; Hassan and Rizkalla 2004; Sena Cruz et al. 2004; Barros et al. 2005) and isolated block elements for a pull out test (Teng et al. 2006; Seracino et al. 2007). CFRP debonding frequently occurs at geometric discontinuities of a strengthened beam due to stress concentrations (e.g., flexural cracks and CFRP cut off points). Several failure modes are available for NSM CFRP systems: F RP rupture, adhesive concrete interfacial fracture, and adhesive splitting (El Hacha et al. 2004; Teng et al. 2006; Seracino et al. 2007). Bond strength of CFRP strips is generally greater than that of CFRP bars because of their geometric configuration and corresponding stress distribution (Blaschko 2003). Bond length of NSM CFRP is an important factor controlling the efficacy of a strengthening system. An insufficient bond length can cause

PAGE 32

31 premature debonding of the CFRP (Hassan and Rizkalla 2004). Other p arameters can also affect the behavior of strengthened beams such as concrete strength, CFRP aspect ratio, and groove size (Teng et al. 2006; Seracino et al. 2007). The embedment depth of CFRP strips appears to be related to a pseudo confining effect that may delay splitting failure of the strips (Seracino et al. 2007). Long term performance of NSM CFRP concrete interface has been stud ied in some experimental program s. Sena Cruz et al. (2006) tested a midspan hinged reinforced concrete specimen strengthened with NSM CFRP in monotonic and cyclic loads. A decrease in stiffness was noticed when the specimens were subjected to cyclic loading in comparison to the case under monotonic loading. The peak pull out force of NSM CFRP, however, was not influenced by cyc lic load. Badawi and Soudki (2009) tested reinforced concrete beams strengthened with NSM CFRP in high cycle fatigue over one million sinusoidal loading. Bond of NSM CFRP was satisfactory, thereby preserving ductility of the strengthened beams. The enduran ce limit of these fatigue beams was improved up to 24% when relative to that of an unstrengthened control beam. 2.2.6 Failure Modes of NSM Bonded Specimens Based on the conducted literature review, the failure modes of NSM bonded specimens are limited t o concrete crushing; adhesive splitting due to high concentrated stresses between the FRP and the adhesive resin interface; concrete splitting due to relatively no enough restrained on the concrete element or low tensile strength of concrete; and FRP

PAGE 33

32 raptu re which is a rare case and only happen if there is sufficient restrained that delay/prevent the debonding or concrete splitting or crushing to occurs. 2.3 High Temperature Effect 2.3.1 Effect of Elevated T emperature s on CFRP Materials CFRP composites demonstrate different therm omechanical properties compared to conventional structural materia ls such as steel reinforcement. For unidirectional CFRP, the coefficient of thermal expansion in the transverse direction is higher than that in the longitudinal d irection because the resin matrix expands more than the fibers do when heated. CFRP can exhibit a negative coefficient of thermal expansion (MBrace 2007). Thermally induced distress in a concrete structure having CFRP materials may thus be of technical con cern (Gentry and Hussain 1999). Glass transition temperature is one of the important parameters for the constituents of a CFRP composite. This critical temperature is defined as a temperature beyond which morphological change s in the polymeric resin of CFR P take place. It is worth noting that carbon fibers can withstand over 1 000 C [ 1 832F ] (Rostasy 1992) while most commercially available resins have a glass transition temperature of below 100 C [ 212F ] T hermomechanical properties of a polymeric resin r apidly decrease when the temperature applied exceeds its glass transition temperature (Dimitrienko 1999 ) The effect of elevated temperature s on the mechanical behavior of carbon fibers is not conclusive. Dimitrienko (1999) reported that

PAGE 34

33 the strength and s tiffness of carbon fibers tended to increase with temperature while the opposite w as presented by Sumida et al. (2001) 2.3.2 Behavior of Concrete Members Strengthened with NSM CFRP at H igh T emperature Although elevated temperatures are one of the crit ical parameter s for constructed reinforced concrete members (e.g., fire) limited research has been done on the thermal resistance of NSM CFRP. Palmieri et al. (2012) tested NSM CFRP strengthened reinforced concrete beams in fire associated with a service load The range of thermal exposure varied from room temperature to 900 C [ 1 652F ] for 2 hour s Fire protection systems were a dded to the strengthened beams using glass fiber cement and calcium silicate boards. Some beams exhibited bond failure of the CFR P during the fire test evidenced by a sudden increase in deflection of the strengthened beams. Residual strength tests showed that the strengthening system maintained sufficient interfac ial bond as long as the temperature applied was lower than 200 C [ 392 F] Burke et al. (2013) showed test data concerning one way slabs strengthened with NSM CFRP exposed to elevated temperatures. The strengthened slabs showed a noticeable decrease in load carrying capacity at a temperature of 100C [212F], including failu re of the adhesive concrete interface. Kodur and Yu (2013) developed a modeling approach to predict the behavior of reinforced concrete beams strengthened with NSM CFRP subjected to a fire. Temperature dependent material properties were taken into consider ation, including bond deterioration of the CFRP caused by thermal stress. The effect of temperature

PAGE 35

34 altered the failure mode of the se beams from concrete crushing to CFRP rupture. The location of CFRP was found to be a contributing factor to fire resistanc e of the strengthened beams. The need for adequate insulation was discussed. 2.4 Performance Based Fire Safety Design Prevention of heat, smoke, evacuation, and rescue, etc. in addition to the structural stability and integrity both locally and globally are the recent approac hes for fire safety performance based design. The intent, structural use, studying of fire scenarios over the structural life time, ensuring the soundness of the design, and durability d efines the level of performance based fire desi gn. Even though the exposure of heat, high temperatures and gases are non structural, they play an important factor to ensure the robustness and the functionality of the structure to allow for adequate rescue time. With this entire in mind, advanced simula tion and computational methods are used in quantitative assessment on structural fire resistance (Liew 2002). 2.4.1 History of Performance Based Codes Researchers and scholars have been publishing many docu ments on performance based design approach Thi s drive was initiated by the fire protection advancement in research and technology, prediction of fire risks and the complex restriction with the prescriptive goal orient ed designs to fire safety on some federal buildings. Wehrili et. al (1972),

PAGE 36

35 introduced new terms closely related to performance based codes. The author also standardized checklists to evaluate the fire safety designs for hospital bedrooms. Haviland (1978) discussed with the fire safety community the unconformity fire safety levels, the confusion between the codes and the applications, and the less focused on the (1978 ) performance base d objectives were to protec t the public welfare from the f ire. Boring et al. (1981) discussed the need to update the codes to take advantage of the new emerging technology and include the performance measures. He was worried about the ina pplicable of the restrictive codes. Boring et al. (1981) matched the safety of public welfare that Haviland discussed but added the ability of the structure to preserve its functionality. The Conseil International du Batiment (CIB, 1982) described the perf ormance guidelines and criteria approach in a report issued in 1982. The report detailed that the prescriptive requirements are not economical and costly despite the simplicity of working with this approach. The International Standard ISO 6241 (1984) provi des the guidelines for preparing performance standard approach in buildings. A research was conducted in Japan by Wakamatsu (1988) where he outlines the design method for fire safety building evaluation. In his outl ine, he stated that performance based cod es substitute the existing prescriptive codes though establishing equivalency. The author motivation was the problems behind implementing the prescriptive codes that are less efficient, difficult to execute with the available technology, and difficult to u nderstand the level of fire safety (Hdjisophocieous et al. 1998).

PAGE 37

36 Ferguson (1993) considered the prescriptive codes requirements may be simple to apply with less flexibility to allow innovation. The author states that the rewritten building regulations i n 1985 offered more guidance and alternative solutions using fire safety based on simple qualitative requirements (Hdjisophocieous et al. 1998). In 2001, the Canadian Commission on Building and Fire Codes (CCBFC, 1994) unveil a draft that develop ed a fully objective based code. Meacham and Custer (1995) provided a summary base d on the up to date performance based approaches for fire safety design. The authors state that there is a need for more guides related to fire engineering design (Hdjisophocieous et a l. 1998). On the other hand, Meacham and Custer (19 95) stated that the performance based design was based on specific cases with respect to the entire fire/building interaction (Hdjisophocieous et al. 1998). The main object ives of fire safety performance b ased design is to protect the building occupants from fire, retard its growth with minimum fire impacts that support the fire fighting operations, and to reduce the injuries if any structural loss occurs (Bukowski and Babrauskas 1994) (New Zealand Building Codes 1994). 2.4.2 Advantages an d Disadvantages of Prescriptive Based Design Fire safety design codes can be p rescriptive type or performance based type. Currently, most codes use prescriptive codes or combination of both prescriptive and performance ba sed codes. Th e disadvantages of prescriptive based method are that it does not clearly identify the factor against failure, because there is no consideration of the strength interaction and stability between the elements and the overall structural system ( Kim et

PAGE 38

37 al.). Despite the popularity and limitation of the prescriptive based design methods when dealing with less complex requirements, many countries around the world are making the shift between the pr escriptive based to performance based design (Hdjiso phocieous et al.). Almost all researchers agreed on the simplified evaluation of requirement for fire safety engineering with minimum engineering expertise for typical structures (Hdjisophocieous et al.). For special structures with non typical design for example large spaces, high ceiling may challenge the engineer to apply the prescriptive based design. The inapplicability of the accumulating restrictive regulations from code to code had caused higher cost design that may not consider the heat transfer fo r example in large volume spaces or the use of the intended spaces such as electronic of chemical cleaning rooms that determine the fire resistance degree needed, limited innovative approaches, and the belief that the only way to achieve fire safety design 2.4.3 Advantages a nd Disadvantages of Performance Based Design Fire sa fety engineering in performance based design is protection design based on (1) agreed upon fire safety goals, loss objectives and performance objecti ves, (2) deterministic and probabilistic evaluation of fire initiation, growth and development, (3) the physical and chemical properties of fire and fire effluents, and (4) quantitative assessment of design alternatives against loss and performance objecti ves (Meacham and Custer 1995) Recently, performance based codes are getting more and more recognized by allowing the design engineer to use more means to evaluate the safety of the structure under fire. It is mainly caused by the

PAGE 39

38 imposing new requirement s of the excessively restrictive prescriptive design methods. Yet, evaluating the overall structural performance is not easy. The degree of fire resistance is the key feature of implementing the performance based fire design codes. Performance based codes permits the use of the latest fire research, analysis, and models that will meet the code safety and needs of both the user and the design engineering (Hdjisophocieous et al. 1998). It establishes clear safety goals that the engineer has control on. Perfor mance base d approach helps harmonization of international guideline and the use of new technical applications at lower costs and more flexible designs (Hdjisophocieous et al. 1998). Other than expertise judgment of evaluating the risk assessment of fire, p erformance based design approach can be applied either probabilistic or deterministic. Typically, the end results are measured in terms of the level of risks that are imposed on the attendants and the structure if probabilistic methods were used, while the fire growth calculations, spreading of smoke and behavior of structure are the criteria if deterministic methods were used. The choice of the method used can be determined on the complexity of the problem. However, it is difficult to define the level of s afety quantitatively and to evaluate compliance with the current requirements (Hdjisophocieous et al. 1998). Usually, performance based approach needs of computer models to evaluate performance of the structure from the fire starts to the fire decay over t he components of the fire safety systems to check determine the level of success of the design. In summary, Fire safety engineering proposes a method to quantify and assess the performance of the structure with respect to the fire growth, and evacuation. With the of

PAGE 40

39 mo ving towards to the performance based design codes that predicts the structural system strength over the element strength itself (Kim et al. 2003). 2.4.4 Development and Evaluation of Design Fire The development of design fire was clearly illustrated thr ough a flow chart figure [Fig. 2.3 ] made by Custer (1995). Custer started with evaluating the situation that defines the client loss objective(s). The n he develop ed the performance criteria that suites the situation and then developed fire s cenarios where fire is the main variable component. If the performance exceeds the performance objectives then the design fire is selected. If the performance undermines the performance objectives then another objective(s) for t he client loss has to be def ined The process continues until all the performance objectives are met (Meacham and Custer 1995). The author also proposes a chart for evaluating alternative designs that meet the optimum requirement fo r fire safety as shown in Fig. 2.4 (Custer 1995). 2 .4.5 Importance Factor Approach Kodur and Naser (2013) proposed a method to evaluate bridges subjected to fire. The authors suggested an importance factor to evaluate the risks of fire over bridges. The importance factor is based on the susceptibility of bridges to fire. Many factors are involved in the calculations of the importance factor, some are related to size and

PAGE 41

40 materials used in the bridge, others are the chances of fire occurrence. Kudor and Naser (2013) proposes a weighted averages based on para meters and sub parameters determined from engineering judgment and design recommendatio ns from previous studies of Garlock et al. ( 2012) ; Elhag and Wang ( 2007) ; Dwaikat ( 2011) ; and Scheer ( 2010). Figure 2.5 and Table 2.2 shows the characteristics and the w eightage factors based on bridge features that influenced the of fire hazard in a bridge according to Kodur and Naser (2013). According to Kodur and Naser (2013) there is about 5% of the nationwide bridge population is considered to have a critical risk g rades to fire hazard. The importance factor classification listed in table 2.3 was based on the parameters and the sub parameters of the weightage factors shown in table 2.2 Using some equations stated in chapter 5, Kodur and Naser (2013) determine the co ntribution to overall influence factors from different classes as shown in Figure 2.6 2.5 Conclusion Based on the Author knowledge, it is very difficult to draw a conclusion from the literature reviews. Failure modes of those literature review tests w ere either debonding failure or concrete failure and they dependent on the certain conditions such as test setup, data analysis, strengthening method, etc. It is clear that the NSM FRP strengthening system shows better performance that the externally bonde d FRP. Because of the discrepancy between the NSM strip and NSM bar there is no clear conclusion, even

PAGE 42

41 though the performance FRP in the NSM strips are better from the NSM FRP bars and the FRP strips have the tendency rapture at high strength loads rather than debonding. With all these studies conducted, very minimum research has focused on the performance of NSM FRP at high temperature. The research presented in the following chapters is a one step further in developing the NSM FRP emerging technology an d ultimately developing a design method or recommendations for the NSM FRP strengthening applications.

PAGE 43

42 Table 2.1 (Adapted from ACI 440.2R 08) FRP Fiber Type Ultimate Tensile Strength Carbon 1,020 2,080 MPa [150 350 ksi] 100 140 GPa 15,000 21,000 ksi Glass 520 1,400 MPa [75 200 ksi] 20 40 Gpa [3,000 6,000 ksi] Aramid 700 1,720 MPa [100 250 ksi] 48 68 Gpa [7,000 10,000 ksi]

PAGE 44

43 Table 2.2 Bridge features weightage factors (Kodur and Naser 2013)

PAGE 45

44 Table 2.3 Risk grades w.r.t the overall class coefficient and the importance factor (Kodur and Naser 2013) Figure 2.1 Pre cured system aramid type FRP installation ( Carmichael and Barnes 2005) Figure 2.2 Groove dimension limitations (ACI 440.2R 08, 2008)

PAGE 46

45 Figure 2.3 Development of Design Fires (Custer 1995)

PAGE 47

46 Figure 2 4 Evaluation of Alternative Designs (Custer 1995) Figure 2.5 The influencing characteristics of fire hazard in bridges (Kodur and Naser 2013)

PAGE 48

47 Figure 2.6 The contribution to overall influence factors from different classes (Kodur and Naser 2013)

PAGE 49

48 3. Behavior of Near Surface Mounted CFRP C oncrete Interface in Thermal Distress 3.1 Fabrication and Specimen Design 3.1.1 General Overview Forty eight (48) non reinforced concrete blocks were made with compressive strength c of 40 MPa [5,800 psi] with each dimensioned as following (200 mm [8 in] long 100 mm [4 in] deep 100 mm [4 in] wide) having a groove of size (25 mm [1 in] deep 13 mm [0.5 in] wide) along the strengthening direction After concrete curing, the NSM Aslan 500 TM CFRP strip is insert ed and bonded Each twenty four (24) concrete blocks were bonded with the different type s of adhesives Power Fasteners T308 + (Power Fastener, 2011), a two components epoxy adhesive anchoring system which will be labeled as ordinary epoxy ; and high temperature laminated resin ( PTM&W Industries 2013 ) a high viscous adhesive epoxy resin with a hardener which will be labeled as high temperature epoxy B ased on the adhesive manufacturer data sheet recommendations, the NSM CFRP bonded strips were cured, and the concrete blocks were subjected to thermal distress from 25C [77F] to 200C [392F] for three hours.

PAGE 50

49 In summary the test setup had undergone three phases: Phase 1: Concrete casting and curing Phase 2: Bonding CFRP strips to concrete by two different types of adhesives, then curing at room temperature t o allow the adhesives to utilize the full strength. Phase 3: Subjecting the concrete blocks to thermal distresses up to 200C [392F] for three hours and then testing the specimens in tension Throughout the literature review, NSM CFRP strips have proven to utilize the FRP strength more than the CFRP bars. From the economical standing point, CFRP strips consume less adhesive than CFRP bars by using smaller groove dimensions. CFRP strips are also less labor intensive than the CFRP bars. The failure expec tancy of this test setup is either between the interface of the concrete NSM CFRP or between the adhesive NSM CFRP. The temperature exposure and the transition temperature, T g of the adhesive will determine the failure mode. 3.1.2 Preparation The concr ete blocks were casted in a wooden frame box that was prep ared by the author as shown in Fig. 3.1. The NSM CFRP concrete interface blocks were designed with the

PAGE 51

50 dimensions stated above and shown in Fig. 3.2. Two CFRP strip were inserted in the groove start ing from the middle of the concrete interface block to the other edge of the specimens. The CFRP strip extended 10 0 mm [ 4 in] outside the edges of the concrete block interface to allow for the MTS machine grips to clamp the specimen and apply tension force as shown in Fig. 3.3. 3.1.3 Materials Concrete was mixed in the laboratory for a specified compressive strength of 40 MPa [5,800 psi] The 28 day strength was measured to be 41. 8 MPa [6,0 6 0 psi] based on the average capacity of test c ylinders 8 in.]) T he CFRP strip used is comprised of carbon fibers (4 626 MPa [700 ksi]) impregnated with a bisphenol epoxy vinyl ester resin (Hughes Brothers 2008). The strip surface is treated to improve bond wh en used with an adhesive. Table 3. 1 summarizes typical engineering properties of the CFRP. Two kinds of bonding agents were utilized to install the CFRP in to a concrete element : ordinary and high temperature epoxy adhesives. The ordinary epoxy is a two par t compound consisting of a hard en er and a resin and is injected into a narrow groove using a specially designed nozzle Th is mercaptan free epoxy is viscous and has a capacity of at least 57 MPa [ 8,267 psi ] The high temperature epoxy includes low viscosit y and demonstrates good compatibility with structural fibers It is particularly suitable for structural bonding when significant thermal stress is expected T he bonding agent will maintain its geometric stability until a temperature of 180C [ 356 F] is re ached without showing premature distortion and shrinkage ( PTM&W Industries 2013 ). A

PAGE 52

51 recommendation which had favorable characteristics for wetting fabrics up to a n intermediat e temperature range between 149C [ 300 F ] and 177C [ 350 F ] The glass transition temperature specified by the manufacturer is at 170C [338F]. The high temperature epoxy requires three stage curing ( PTM&W Industries 2013 ) : i) a prepared mixture is appli ed at room temperature and allow s curing for a minimum of 18 hours ; ii) it is exposed to gradually increasing temperature s up to 177C [350F] for 12 hours ; and iii) the heated adhesive is cooled down to room temperature. 3.1.4 Test Specimen To examine the behavior of an NSM CFRP strengthening system an isolated interface c onfigura tion was used as shown in Fig. 3.2 S uch a test protocol can readily represent NSM CFRP strips installed in a reinforced concrete beam with emphasis on the effective tensile z one of the strengthened beam. A total of forty eight ( 48 ) specimens were prepared ( Table s 3 .2 ) Each specimen included one concrete block ( 200 mm [ 8 in ] long 100 mm [4 in] deep 100 mm [4 in] wide ) having a groove along the strengthening direction. The groove size ( 25 mm [1 in] deep 13 mm [0.5 in] wide ) was designed as per the recommendation of ACI.440.2R 08 (ACI 2008): the width and depth of the groove were greater than 3 and 1.5 times the CF RP thickness and width, respectively. Upon complete curing of the concrete specimens, CFRP strips were bonded with the adhesives mentioned earlier. The groove was first cleansed with an air brush to eliminate surface dirt that could degrade bond between th e adhes ive and concrete surface [Fig. 3.4 (a )].

PAGE 53

52 Two precut CFRP strips ( 16 mm [0.63 in] wide 2 mm [ 0.08 in ] thick 200 mm [8 in] long, each) were positioned with each bonding agent injected into the groove [Fig. 3 .4 (b )]. Both ends of the groove were un bonded using styroform pieces to prevent the premature failure of the interface caused by stress concentrations when mechanically loaded The assemble d test specimens were cured according to recomm endation. 3.1.5 Thermal Exposure After adequate curing of the adhesive ly bonded NSM CFRP strips all specimens were exposed to elevated temperatures ranging from 25 C [77F] to 200 C [ 392 F] at a typical interval of 25C [77F] for three hours using a d igital control electric furnace [Fig. 3.5 (a)] It should be noted that the furnace was preheated to a designated temperature before the thermal exposure of the NSM CFRP concrete interface commenced The temperature of the specimens was monitored using a l aser the rmometer gun, as shown in Fig. 3.5 (a), to ensure the temperature inside the furnace. The thermometer gun confirms the thermocouple reading as shown in the Fig 3.6 When the planed thermal exposure was completed each specimen was cooled down to ro om temperature for one day [Fig. 3.5 (b)] The variation of concrete strength within the temperature range studied in this experimental program was previously examined by Kim et al. (2012) and thus additional material test ing was not conducted.

PAGE 54

53 3.1.6 Test Setup and Instrumentation The specimens were positioned to an MTS machine for mechanical testing, as shown in Fig. 3.5 (c). The CFRP strips were gripped and monotonic tension was applied at a loading rate of 0.1 mm/sec [ 0.004 in/sec ] until interfac ial fa ilure took place. A clamping system was used [Fig. 3.5 (b) and (c)] to preclude tensile splitting failure of the concrete in the vicinity of the gap between the two CFRP strips (i.e., mid length of the specimen) The load applied was measured with the buil t in load cell and corresponding displacement was recorded by the stroke of the machine 3.2 Material Level Testing 3.2.1 D ynamic Mechanical Analysis Dynamic Mechanical Analysis (DMA ) [fig.3.7 ] was used to measure the C FRP and adhesive epoxy response. D MA uses an imposed oscillating stress or strain. Some level of deformation will be caused by the dissipation of stress imposed on the samples. Since most adhesives are viscoelastic in nature, some level of deformation will be recovered when stress is relea sed. stress strain. The computerized DMA instrument measures the storage and recovered modulus and the loss energy as well. The ratio of the loss modulus over the storage and recovered modulus will de verify ing the transition temperature ( T g ) of the different type of adhesives and the CFRP

PAGE 55

54 strip. The test samples were conducted at the mechanical department according to ASTM E1356 03. The resul ts are shown in Fig. 3.8 for the CFRP strip [Fig. 3.8 (a)] the high temperature epoxy [Fig. 3. 8 (b)] and the ordinary epoxy [Fig. 3.8 (c )] The high temperature and the ordinary epoxy were specified as a thermoset resin. A sudden drop in the storage modulus was shown by the ordinary epoxy which may indicated a melting possibility and thus losing it properties. The glass transition temperature ( T g ) of the high temperature epoxy is at 163 C [32 5 F] and the ordinary epoxy appeared to be approximately 65C [149 F] 3.2.2 Concrete Testing under Thermal D istress 200 8 in.] diameter cylinders. The results were discussed earlier under (3.1.3). The re s ults are summarized in Fig. 3.9 (a). From the results the temperature specified seemed to have minimal effect on the concrete strength. 3.2.3 CFRP St r ip Testing under Thermal D istress The performance of CFRP strips under tension load was consistent over the specified temperatures. The failure mode was along the fiber line and might be caused by a CFRP slip between the grips. The results are summarized in Fig. 3.9 (b). From the results and literature review the specified temperature has no effect on the CFRP tension capacity.

PAGE 56

55 3.2.4 Ordinary Epoxy and High Temperature Coupons The tensile testing of the ordinary epoxy and the high temperature epoxy were (2010) to verify the effect of thermal distress on the epoxy resin. The results are summarized in Figs. 3.9(c) and 3.9 (d). The results does not fully confirm with the manufacturer specification data sheet, mainly due to the residual tensile testing where s ome of the properties were recovered after thermal curing and cooling process. A further investigation is needed in this field. The failure mode of the adhesive epoxy coupons and the CFRP strip is shown in Figure 3.10. 3.3 Experimental Results 3.3 .1 Anal ysis of Variance for Concrete at Elevated Temperatures T he strength variation of concrete in compression is summarized in Table 3. 3 depending upon the degree of temperature exposure. Although the strength tended to decrease with an increasing temperature a n insignificant difference was observed in all test categories within a range between 4 1.8 MPa [ 6,060 psi ] to 39 .2 MPa [ 5,6 90 psi ] To clarify the temperature dependent strength issue of the concrete, analysis of variance (ANOVA) was performed at a leve l of significance = 0.05 Hypotheses were proposed whether the

PAGE 57

56 temperature range studied would have any effect on the variation of concrete compressive strength. T he following was tested using Eq. 1 : H 0 : A ll the means of the tested concrete strength are equal H a : Not al l the means of the tested concrete strength are equal ( 1) where F is the F distribution of the tested concrete; m is the sample size per test group; is the sample variance; is the sample variance per group; and k is the number of the means. The degree of freedom is defined as DOF = ( k 1, n k ) to dete rmine the critical F distribution value (e.g., F .05 for a level of confidence of 0.05). The calculated F distribution value was 0.87, which was not in the critical region F .05 of 2.85. The H 0 hypothesis, therefore, was not rejected. It implies that insuffi cient statistical evidence has existed to conclude the thermal exposure varying from 25C [77F] to 200C [392F] influenced the strength of the concrete 3.3 .2 Interfacial Strength of NSM CFRP Table 3. 2 list s the failure load of the specimens bonded w ith the ordinary epoxy adhesive and high temperature epoxy The average capacity of the NSM CFRP concrete interface

PAGE 58

57 without experiencing thermal distress for ordinary epoxy (the OE25 series) was 21. 5 kN [4.8 kip] with a standard deviation of 2.5 kN [0.6 ki p] The interfacial capacity was preserved up to a temperature of 75C [ 167 F ] This illustrates that the glass transition temperature ( T g ) of the ordinary epoxy appeared to be approximately 75C [167F] and hence the performance of the NSM CFRP strengthen ing system was not influenced by the degree of t hermal exposure within this temperature boundary The specimens subjected to a range between 100C [ 212F ] and 150C [ 302F ] exhibited some degradation in interfacial capacity including an average decrease o f 20.8 % in comparison to the capacity of the 25 C [ 77F ] category The rate of strength decrease was significant beyond a temperature of 150C [302F]. The capacity of the specimens exposed to temperatures of 175 C [ 347 F] and 200 C [ 392 F ] was 38.6 % and 8 9.8 % lower than that of the specimens tested at 25C [77F] T he strength of the specimens subjected to 200C [392F] was almost negligible. The NSM CFRP strengthening system bonded with the ordinary epoxy therefore, will not function as designed when the surrounding temperature exceeds 75 C [ 167 F] and will demonstrate a noticeable decrease in its interfacial capacity over 150 C [302F] even though an insulation material covers the strengthening system. It means that the fire endurance time of NSM CFRP s trengthened members bonded with such an adhesive material needs to be estimated until critical temperature s are reached for design and practice The interfacial strength of the specimens bonded with the high temperature epoxy is given in Table 3. 2 The o nes without thermal exposure (the HE 25 series) demonstrated an average interfacial strength of 17.2 kN [3.9 kip], which was 20% lower than the

PAGE 59

58 strength of the OE series at 25C [77F] The interfacial capacity of the specimens tended to decrease with an i ncreasing temperature, while the degree of strength reduction was not as obvious as the case of the ordinary epoxy bond. For instance, the strength drop of the HE series was 28.5 % when a temperature changed from 25 C [77F] to 200 C [392F] which was a su bstantially low rate when relative to the 89.8% of the OE series under the same condition. According to the test results of the HE specimens shown in Table 3. 2 the glass transition temperature of the high tem perature epoxy seemed to be at around 175C [34 7F] This value well conform s to t he glass transition temperature reported by the manufact ure 170 C [ 338 F ]. A graphical comparison on the thermal performance of the NSM CFRP concrete interface bonded with the ordinary or high temperature epoxy is made i n Fig. 3 .11 The strength gap between these two cases was by and large, maintained up to a temperature of 150C [302F] beyond which the specimens bonded with the high temperature epoxy displayed superior behavior to those with the ordinary epoxy 3.3 .3 Load D isplacement Response Figure 3.1 2 shows the load displacement behavior of the specimens bonded with the ordinary epoxy. The OE25 series exhibited a linear response until an abrupt load drop was associated due to bond failure [Fig. 3.1 2 (a)] The sti ffness of the respective specimens was more or less similar to each other, while their strength was slightly different. This can be explained by the fact that the amount of epoxy inside the groove was not the same because of experimental randomness to a ce rtain extent The specimens

PAGE 60

59 subjected to elevated temperatures demonstrated similar behavior [Fig. 3.1 2 (b) to (g)] while a local load drop was noticed in some cases Such an observation illustrates that the interface was partially damage d due to the ther mal exposure. The specimens at 200C [392F] showed somewhat ductile failure with a substantially low capacity It should be noted that Specimen OE200 3 failed in a premature manner when tested and thus its load displacement response was not provided in Fi g. 3.1 2 (h). A summary of load displacement over the range of different temperature for ordinary epoxy is shown in Fig. 3.14(a) Figure 3.13 reveals the behavior of the specimens bonded with the high temperature epoxy. The response of these cases was virtua lly the same as that of the former cases discussed in Fig. 3.1 2 It may be of interest to note that the cured high temperature epoxy looked like brittle that was different from the ordinary epoxy and thus shows adhesive failure at low temperatures and bond failure at high temperatures 3.3 .4 Failure Mode The interfac e failure of selected specimens is presented in Fig. 3.1 5 Although infinitesimal deformation was recorded with an increasing mechanical load (Figs. 3.1 2 and 3.1 3 ) a morphological c hange along the CFRP concrete interface w as not apparent until a peak load was achieved. When the failure of the interface was imminent several hairline cracks formed along the bond line and further developed as shown in Fig. 3. 15 (a) At this stage, s tre ss redistribution was expected between the CFRP system and surrounding concrete. The failure of the CFRP was localized along the groove [Fig. 3.1 5 (b)] which indicated the existence of an effective failure zone in the vicinity of the

PAGE 61

60 installed NSM CFRP A relationship among this effective failure zone the configuration of the CFRP system and the properties of the surrounding concrete appears to be an interesting topic for future research. A secondary concrete crack was observed near the mid length of the specimen because of the spatial gap between the two CFRP strips The failure mode of the HE specimens bonded with the high temperature epoxy [Fig. 3.15 (c)] was analogous to that of the OE categories. Additional failure modes of ordinary epoxy and high temp erature epoxy at 50 C [ 122 F] and 200 C [ 392 F] are shown in Fig. 3.16 (a) (d). 3.4 S ummary and Conclusion This chapter has discussed the residual behavior of NSM CFRP strips embedded in a concrete substrate when subjected to elevated temperatures rangi ng from 25 C [ 77F ] to 200 C [ 392F ] Two types of adhesives were used to bond the strips : ordinary and high temperature epoxies. Test specimens were conditioned in the predefined thermal environment for three hours, cooled down to room temperature, and me chanically loaded to failure. Interfacial strength between the CFRP and concrete was measured and corresponding failure mode was studied. The se technical results are part of an ongoing research program examining the thermal performance of an NSM CFRP stren gthening system for concrete structures with focus on material and structure level investigations. Some preliminary conclusions are drawn as follows.

PAGE 62

61 According to the statistical analysis, the variation of concrete strength in compression was not influen ced by a temperature range from 25C [77F] to 200C [392F]. The capacity of the CFRP concrete interface was, however, affected by thermal exposure within this boundary The specimens bonded with the ordinary epoxy preserved interfacial strength until 75 C [167F] beyond which a noticeable strength decrease was observed, in particular over 175C [347F]. The ones with the high temperature epoxy, by and large, maintained their interfacial strength up to 200C [392F] even though a trend of strength reductio n was associated with an increasing temperature. The interfacial capacity of the ordinary epoxy was higher than the high temperature epoxy, but the high temperature epoxy showed a consistent capacity over the temperature increase.

PAGE 63

62 Table 3. 1 P roperties of CFRP strip from the manufacturer (Hughes Brothers 2008) Property Value Width 16 mm [0.63 in] Thickness 2 mm [0.079 in] Cross sectional area 31.2 mm 2 [0.49 in 2 ] Tensile strength 2068 MPa [300 ksi] Tensile modulus 124 GPa [18,000 ksi] Ult imate strain 0.017 Coefficient of thermal expansion (transverse) 74 to 104 10 6 (/ o C) [41 to 58 10 6 (/ o F)] Coefficient of thermal expansion (longitudinal) 9 to 0 10 6 (/ o C) [4 to 0 10 6 (/ o F)]

PAGE 64

63 Table 3. 2 Test specimens load and failure mode of ordinary epoxy and high temperature epoxy N/A: premature failure during test

PAGE 65

64 Table 3. 3 Compressive concrete strength depending upon temperature (Kim et al. 2012) ID Test strength (MPa [psi]) ID Test strength (MPa [psi]) ID Test strength (MPa [psi]) Each Average Each Average Each Average C25 1 41.7 [6,050] 41.7 [6,050] C100 1 40.8 [5,920] 40.8 [5,920] C200 1 36.6 [5,310] 39.2 [5,690] C25 2 41.2 [5,980] C100 2 42.1 [ 6,110] C200 2 41.2 [5,980] C25 3 42.1 [6,110] C100 3 39.1 [5,670] C200 3 38.8 [5,630] C50 1 41.2 [5,980] 41.8 [6,060] C125 1 38.4 [5,570] 40.4 [5,860] ANOVA: = 0.05 F = 0.87 F .05 = 2.85 nclusion: F < F .05 C50 2 44.1 [6,400] C125 2 41.2 [5,980] C50 3 40.1 [5,820] C125 3 41.7 [6,050] C75 1 42.6 [6,180] 41.2 [5,980] C150 1 37.1 [5,380] 40.0 [5,800] C75 2 39.1 [5,670] C150 2 42.8 [6,210] C75 3 41.7 [6,050] C150 3 39.3 [5,700]

PAGE 66

65 Figure 3.1 Wooden mold in which concrete blocks were formed Figure 3.2 NSM CFRP concrete interface test specimen dimension (not to scale) Figure 3.3 Interface test specimens after appl ying an epoxy adhesive

PAGE 67

66 (a ) (b ) Figure 3.4 Test specimen: ( a ) air blasting before CFRP bonding ; (b ) CFRP installation (a) (b) (c) Figure 3.5 Test setup: (a) high temperature exposure; (b) conditioned specimen; ( c ) tension test Groove CFRP

PAGE 68

67 (a) (b) (c) (d) (e) (f)

PAGE 69

68 (g) (h) Figure 3.6 Ther m ocouple temperature readings over a three hour range acquired from the data logger: (a) at 50C [122F]; (b) at 75C [167F]; (c) at 100C [212F]; (d) at 125C [257F]; (e) at 150C [ 302F]; (f) at 175C [347F]; (g) at 200C [392F] ; (h) combined temperature (a) (b) (c) Figure 3.7 Dynamic mechanical analysis (DMA): (a) testing machine ; (b) DMA clamps; (c) DMA specimens 200 o C 175 o C 5 0 o C 125 o C 1 00 o C 75 o C 150 o C

PAGE 70

69 (a) (b) (c) Figure 3.8 Dynamic Mechanical Analysis (DMA) results: (a) CFRP strip; (b) high temperature epoxy ; (c) ordinary epoxy

PAGE 71

70 (a) (b) (b) (d) (e) (f) Figure 3.9 Temperature dependent residual strength measured: (a) concrete in compression; (b) CFRP strip in tension; (c) ordinary epoxy in tension; (d) high temperature epoxy in tension ; (e) normalized strength of OE; (f) normalized strength of HE

PAGE 72

71 (a ) ( b ) (c) Figure 3.10 Failur e mode : (a) HE epoxy cou pons at 200C [392F]; (b) OE epoxy coupons at 200C [392F]; (c) CFRP strip at 200C [392F] (a) (b) (c) Figure 3 .1 1 Tem perature dependent interfacial strength: (a) specimens bonded with ordinary epoxy; (b) specimens bonded with high temperature epoxy; (c) t emperature dependent average interfacial strength Ordinary epoxy High temperature epoxy

PAGE 73

72 (a) (b) (c) (d) (e) (f)

PAGE 74

73 (g) (h) Figure 3.12 Load displacement of specimens bonded with ordinary epoxy : (a) at 25C [77F] ; (b) at 50C [ 122 F] ; (c) at 75C [ 167 F] ; (d) at 100C [ 212 F] ; (e) at 125C [ 257 F] ; (f) at 150C [ 302 F] ; (g) at 175C [ 347 F] ; (h) at 200C [ 392 F]

PAGE 75

74 (a) (b) (c) (d) (e) (f)

PAGE 76

75 (g) (h) Figure 3.13 Load displacement of specimens bonded with high temperature epoxy: (a) at 25C [77F]; (b) a t 50C [122F]; (c) at 75C [167F]; (d) at 100C [212F]; (e) at 125C [257F]; (f) at 150C [302F]; (g) at 175C [347F]; (h) at 200C [392F ] (a ) (b ) Figure 3.14 Load displacement comparison of exposure temperatures: (a) bonded with ordinary epoxy; (b) bonded with high temperature epoxy (a) (b) (c) Figure 3.1 5 Failure mode : (a) interface failure of ordinary epoxy ; (b) interfacial failure of ordinary epoxy after test; (c) interfacial f ailure of high temperature epoxy after test Interface failure Interface failure Secondary concr ete failure Interface failure Secondary concrete failure 200C 100C 150C 50C 200C 100C 150C 50C

PAGE 77

76 (a) (b) (c) (d) Fig ur e 3.1 6 Failure mode: (a) ordinary epoxy at 50C [122 F] ; (b) ordinary epoxy at 200C [392 F] ; (c) high temperature epoxy at 50C [122 F] ; (d) high temperature epoxy at 200C [392 F] Failure along concrete substrate Failure along bond line Failure along concrete substrate Failure along bond line

PAGE 78

77 4. Bond Performance of Concrete Adhesive Interface at Elevated Temperatu res 4.1 Fabrication and Specimen Design 4.1.1 General Overview Seventy two (72) non reinforced concrete prisms were made with compressive strength, c of 40 MPa [5,800 psi] with each dimensioned as following (400 mm [16 in] long 100 mm [4 in] wide 80 mm [3.15 in] deep). The prisms were split in half by using a wood divider sheet as shown in Fig. 4.1. A fter concrete curing, e very twenty four (24) concrete prism with the different type of adhesive Power Fast eners T308 + (Power Fastener, 2011), a two components epoxy adhesive anchoring system w hich will be labeled as high viscosity adhesive epoxy; B ASF MBrace Saturant (MBrace, 2007), also a two components epoxy adhesive anchoring system will be labeled as Low v iscosity adhesive epoxy; and high temperature laminated resin ( PTM&W Industries 2013 ), a high viscous adhesive epoxy resin with a hardener which will be labeled as high temperature epoxy. T he two pieces prism were bonded together to form one and then cur ed based on the adhesive manufacturer data sheet recommendations. After adhesive curing, a small groove was created by a stationary electric saw, dimensioning 10 mm [ 0. 4 in] deep. T he concrete prism were subjected to thermal distress from 25C [77F] to 20 0C [392F] for three hours.

PAGE 79

78 The failure expectancy of this test setup is at the bonding location of the two prism pieces, t he interface of the concrete adhesive or between the adhesive adhesive The temperature exposure and the transition temperature, T g of the adhesive, and the adhesive thickness will determine the failure mode. 4.1.2 Concrete and Adhesive Material Properties Concrete was mixed in the laboratory under the same conditions of the previous experiment. The specified compressive strength was 40 MPa [5,800 psi] after 28 days. Dog bone shaped coupons were made from the three types of adhesive. Twenty four (24) coupons were made of each adhesive type. The coupons are 100 mm [4 in] long 5 mm [0.2 in] wide 10 mm [0.4 in] deep as show in Fig. 4.2. The final shape of different adhesive types testing coupon are shown in Fig. 4.3. The coupons were then exposed to temperature distress from 25C [77F] to 200C [392F] for three hours. The high viscosity adhesive is a two part compound consisting of a harder and a resin and is injected onto one side of the prism and then bonded together. This mercaptan free adhesive is viscous and has a capacity of at least 57 MPa [8,267 psi]. The low viscosity adhesive i s a two part (Part A to Part B) mixed with a 100:30 ratio by weight applied after manufacturer recommended mixing on one side of the prism and bonded the other prism piece. The adhesive compression capacity is 86.2 MPa [12,500 psi]. The low viscosity adhes ive has a transition temperature, T g of 71 C [163F] (Mbrace, 2007) Material properties of this bonding agent are provided in Table 4.1. The high temperature epoxy is particularly suitable for structural bonding when significant thermal

PAGE 80

79 stress is expect ed. The transition temperature, T g of this epoxy is 180C [356F] ( PTM&W Industries 2013 ). a har dener for this bonding material mixed over a ratio of 100:22 by weight ( PTM&W Industries 2013 ) 4.1.3 Testing Specimen and Ther mal Exposure The test was conducted per ASTM C78, a standard testing method with third point loading (ASTM, 2009) The test set up investigates the prism bond strength of 72 concrete prism specimens After concrete curing, each 24 prism were bonded with d ifferent type of adhesive and cured according to the manufacturer recommendations The concrete prisms are all identical and mixed with the same concrete mix design. The test set up is shown in Fig. 4.5 (b) T o the author knowledge, t he re was no particular test preparation found that provide the bondin g performance of concrete adhesive interface at elevated temperatures. After proper curing of the adhesive bonded prisms, specimens were exposed to elevated temperature over the range from 25C [77F] to 200 C [392F] over a typical interval of 25C [77F] for three hours using a digital control electric furnace and verified by using a temperature data logger [Fig. 4.6 ]. The data logger shows consistent oven temperature over the tes ting temperature range [Fig. 4.7 (a) (e)]. It should be noted that the furnace was preheated to a designated temperature before the thermal exposure of the concrete prisms and coupons

PAGE 81

80 4.1.4 Instrumentation The concrete prisms were positioned over a steel fabricated three point sup ports that are 350mm [13.8 in] apart, then placed in an MTS machine as shown in Fig. 4.4 (a) (b) The compression loading rate was of 0.1 mm/sec [0.004 in/sec] until interfacial failure happened. Before flexural concrete prism test took place, the specimens were cut with notches over the bonding location of the specimen by a diamond saw with high stability and accuracy [Fig 4.8 ]. The depth of the notches was 10mm [0.4 in]. This notch was placed to keep the failure mode at the bonding location and prevent fro m having a concrete failure between the supports. The load applied was measured with the built in load cell and corresponding displacement was recorded by the stroke of the machine. The dynamic modulus test was performed as per ASTM C215 ( 2008 ) using the DK 4000 machine [Fig. 4. 10 (a) ] The transverse frequency was measured by using the impact test method An a ccelerometer with an output signal was attached to one end of the prism. The concrete prism was supported by a standard two point aluminum pad specif ic for this test and impacted by a standard hammer at consistent specified locations over all the concrete prisms. The response frequency domain wave was recorded by DK 4000 [Fig. 4. 10 (b)]. Please note that the dynamic modulus test was performed after the no tch in the concrete was made, the sample was fully concr ete and adhesive cured, and the concrete prisms were subjected to thermal distresses.

PAGE 82

81 4.2 M aterial Level Testing 4.2.1 Stress strain Response of Adhesive Resin Table 4. 2 lists the failure tensile load of the coupons made with different adhesives that was specified earlier at elevated temperature. Even though the residual tensile capacity of the adhesive coupons did not reflect failure around the glass transient temperature, the results remains fai rly consist e nt throughout. The average residual tensile capacity of high viscosity adhesive was 25.4 MPa [3.68 ksi]. The residual tensile capacity of the low viscosity adhesive epoxy coupons did show a trend drop in the tensile capacity at T g around 75C [167F] but the ultimate capacity was 58.8 MPa [8.53 ksi]. The tensile ultimate strength capacity of the low viscosity adhesive epoxy coupon was 55.2 MPa [8,000 psi] as per the manufacturer data sheet [table 4.1] The High temperature adhesive epoxy coupon s did not perform as it wa s specified in the data sheet This behavior is due to the complex curing requirement that this adhesive requires. As per the data sheet, the curing requires a 18 hours set after mixing the adhesives then gradually heat to 65.5 o C [150 o F] for 3 4 hours, then the manufacturer also specifies to slowly raise the temperature to 121 o C [250 o F] and hold for 3 4 hours. Lastly the manufacturer recommends turning off the curing oven and letting it cool down to room temperature before service The glass transient temperature was not clear in this test, which was examined earlier and found to be at 175C [347F] from the NSM CFRP concrete block tests.

PAGE 83

82 4.2. 2 Storage Modulus of Adhesive Resin The storage modulus of the adhesive resin was determ ined by usi ng the dynamic mechanical analysis (DMA) The DMA is a is a computerized instrument that measures the storage modulus under a heat flow difference and was used to verify the transition temperature ( T g ) of the different type of adhesives. The tes ts were performed according to ASTM E1356 03. The results shown in Fig. 4.10 (a) identify a T g of 76C [169F] for low vi scosity adhesive, while Fig. 4.9(b ) shows a T g of 65C [149 F] for high viscosity, and Fig. 4.9(c) shows a T g 163C [325F] for high tem perature adhesive 4.2.3 Frequency Test The frequency test was performed on the concrete prisms bonded w ith the three different types of adhesives discussed earlier and subjected to different thermal distresses. The frequency test was p erformed according to ASTM C215 (20 08 ) The prisms were marked and had the same conditions throughout the testing period. The results will be are shown in Fig. 4.11 (a)

PAGE 84

83 4.3 E xperimental Results 4.3.1 Dynamic Elastic Modulus of Concrete Prisms The dynamic elastic modul i of concrete prisms were determined by the standard test method for fundamental transverse, longitudinal, and torsional frequencies of con crete specimens (ASTM C215, 2008 ) Table 4.3 list the dynamic elastic modulus over the specified temperature range. T he high viscosity specimens that was not subjected to thermal distresses HV 25 shows a dynamic elastic modulus of 20.44 GPa [2964.6 ksi]. The high viscosity specimens subjected to 200C [392F] temperature shows 51.8% drop with a dynamic elastic modulus of 9.85 GPa [1428.6]. The low viscosity (LV) dynamic elastic modulus starts with 21.09 GPa [3058.85 ksi] at room temperature and dropped to 7.36 GPa [1067.48 ksi] at 200C [392F]. The LV dynamic elastic modulus was preserved up to 75C [167F] that confir ms the glass transition temperature ( T g ) specified by the manufacturer at 71C [163F]. The dynamic elastic modulus of the LV loses 65.1% of its original capacity. The high temperature (HT) dynamic elastic modulus starts with 21.09 GPa [3058.85 ksi] with out being subjected to any thermal distresses. The dynamic elastic modulus was fairly consistent up to 125C [257F] with a drop of 3%. The dynamic elastic modulus dropped 24.9% at 175C [347F]. The overall drop in the dynamic elastic modulus at 200C [39 2F] was 51.7%. Figure 4. 11 (a) shows the frequency readings with the impact

PAGE 85

84 test method. By using the frequency from the impact testing, the dynamic modulus of elasticity was found th r ough the ASTM C215 (2008 ) equations [Eq. 2 and 3 ] The overall dynamic e lastic modulus is presented in figure 4. 11 (b). E: Dynamic modulus of elasticity (Pa) n: Transverse frequency M: Mass of the specimen C: Dimensions of the specimen L: Length of specimen t: Thickness of specimen b: Width of specimen T: Correction factor dep endent on the radius of gyration to length of specimen and the (2) (3) 4.3.2 Load D isplacement Response of the Bond Test Concrete Prisms Figure 4.1 2 (a) (b) shows the u ltimate load and capacity of the concrete prism specimens bonded with high viscosity adhesive after subjected to thermal distresses As it is clear, there is no specific trend in the results even when utilizing the bonded area in the concrete pr isms. Since, it was observed that the thickness of the adhesive varies

PAGE 86

85 between one type to another depending on many factors e.g. the quantity, application technique, and human error. It was proposed by the author to use the thickness as a way to normalize the capacity of the concrete prisms. With this being in mind the normalized capacity utilized the bonded volume r ather t han the bonded area. Figure 4.12 (c) shows a decreasing trend although the glass transition temperature ( T g ) is not clear. Similarly, the low viscosity adhesive ultimate load and capacity [Fig. 4.1 3 (a) (b)] shows no clear trend. After normalizing the results over the thickness of the adhesive, it was found a decreasing trend [Fig. 4.1 3 (c)]. As for high temperature adhesive concrete pris m, the ultimate load and capacity w ere not clear as well [Fig. 4.1 4 (a) (b)]. The normalized ultimate capacity shows a consistent decreasing trend [Fig. 4.1 4 (c)]. Table 4.4 a lists the temperature dependent adhesive concrete interfacial capacity of the conc rete prism specimens bonded with high viscosity adhesive, Table 4.4b list the adhesive concrete interfacial capacity of the concrete prism specimens bonded with low viscosity adhesive at the specified temperature range, and Table 4.4c lists the high temper ature adhesive concrete interfacial capacity of the temperature dependent concrete prisms. The average capacity of the bond performance of concrete prisms bonded with high viscosity adhesive without being subjected to thermal distresses was 0.31 MPa/mm [11 42 psi/in]. At 200C [392F], the normalized bond capacity dropped by 68.5%. Similarly the low viscosity normalized bond capacity dropped by 68.2% between the

PAGE 87

86 specimens not subjected to thermal distresses and specimens subjected to 200C [392F] The concr ete prisms bonded with high temperature adhesive was 0.47 MPa/mm [1731.5 psi/in] at room temperature and dropped to 0.24 MPa/mm [884.2 psi/in] at 200C [392F]. The normalized capacity drop was the least compared to the other types of adhesive with 48.8%. The concrete prisms bonded with high temperature adhesive have the highest capacity [Fig. 4.1 5 ]. Even the glass transition temperature ( T g ) was not clear in this test, the re s ults follows the same trend of the frequency [Fig. 4.11 (a)] and the dynamic elast ic modulus [Fig. 4.11 (b)]. 4.3.3 Failure Mode of Bonded Concrete P risms The interface failure of the concrete prisms bonded with different types of adhesive is shown in Fig. 4.1 6 The failure mode of the high viscosity specimen was typically bond failure where the adhesive fails at both temperatures 5 0 C [ 122 F] and 200C [392F] as shown in Fig. 4.16 (a) (b) The other two types of adhesives, low viscosity and high temperature failure was within the concrete bonding. The concrete prisms bonded with high v iscosity adhesive have a thickness almost double the concrete prisms bonded with high temperature. A relationship can be drawn from the thickness of the adhesive and the normalized bond failure that the thickness of the adhesive subjected to thermal distre sses can be an interesting topic for future research.

PAGE 88

87 4.4 S ummary and Conclusion This Chapter has discussed the bond performance of the concrete adhesive interface at elevated temperatures. The concrete prisms were subjected to three hours of a tempera ture ranging from 25C [77F] to 200C [392F]. Three different types of adhesive were used to bond the non reinforced concrete prisms: (1) high viscosity adhesive; (2) low viscosity adhesive; (3) high temperature adhesive. The author used the normalized c apacity that utilized the ultimate capacity over the thickness of the adhesive, in other words the adhesive bonded volume was used to normalize the capacity not the bonded area due to thickness variation among the specimens between different types of adhes ive and between the same adhesive type as well. Since the glass transient temperature was not reflected by the concrete prisms, additional material testing was performed to check the T g using the coupon testing and the differential scanning calorimetry. A ccording to the temperature dependent concrete prisms load displacement results, the specimen bonded with high temperature maintained the highest bond performance capacity. The load displacement graph was linearly decreasing with temperature. The thickness of the adhesive was used to normalize the ultimate capacity instead of using the bonded area alone due to a non uniform quantity of adhesive used to bond the concrete prisms and other mentioned factors.

PAGE 89

88 Table 4.1 Properties of low viscosity adhesive fr om the manufacturer ( MBrace, 2007 ) Property Tensile Value Compressive Value Flexural Value Yield Strength 54 MPa [7,900 psi] 86.2 MPa [12,500 psi] 138 MPa [20,000 psi] Strain at Yeild 2.50% 5.00% 3.80% Elastic Modulus 3034 MPa [440 ksi] 2620 MPa [380 k si] 3724 MPa [540 ksi] Ultimate Strength 55.2 MPa [8,000 psi] 86.2 MPa [12,500 psi] 138 MPa [20,000 psi] Rupture Strain 3.50% 5.00% 5.00% Table 4.2 Temperature dependent residual strength of adhesive coupons: (HV) high viscosity adhesive coupons; (LV ) low viscosity adhesive coupons; (HT) high temperature adhesive coupons HV LV HT Temp Stress Stress Stress o C MPa MPa MPa 25 22.2 57.8 21.7 25 15.8 55.4 20.0 25 N/A 55.4 22.6 50 14.5 50.4 28.7 50 22.2 65.9 32.7 50 19.5 69.2 20.7 75 34.8 60.1 12.0 75 26.6 70.8 12.2 75 31.2 73.0 29.4 100 28.5 67.2 35.0 100 16.1 71.9 24.9 100 29.6 64.7 24.4 125 28.8 75.5 28.6 125 20.0 79.6 15.7 125 36.6 50.8 14.0 150 N/A 60.7 26.8 150 23.0 47.8 20.8 150 34.8 76.9 38.2 175 38.2 45.5 28.0 175 35.0 19.6 14.5 175 15.6 49.3 18.1 200 14.8 47.8 36.0 200 32.9 23.2 23.2 200 17.8 72.7 22.3

PAGE 90

89 Table 4.3 A verage d ynamic elastic mod ulus response at elevated temperature s Temp Mbrace T 308 High temp LV HV HT DEM (GPa) Average DEM (GPa) Average DEM (GPa) Averag e 25 22.55 21.09 20.06 20.44 21.89 20.44 25 21.32 20.06 19.96 25 19.36 18.21 18.12 25 20.12 22.00 21.89 25 22.74 20.65 18.93 25 22.74 22.22 22.42 25 19.20 20.65 18.93 25 21.93 21.43 21.62 25 21.63 20.48 20.26 25 20.84 20. 48 20.26 25 22.04 20.87 18.76 25 18.55 18.23 22.22 50 21.42 19.47 19.47 18.83 20.23 19.64 50 21.81 19.10 19.85 50 20.24 19.47 19.47 50 18.36 19.10 19.10 50 20.73 19.55 19.60 50 20.34 19.18 19.60 50 19.96 19.18 18.51 50 1 8.83 18.07 19.60 50 17.81 18.58 20.22 50 18.18 18.21 20.22 50 18.18 17.84 19.84 50 17.81 18.21 19.46 75 19.39 19.55 18.06 17.98 21.25 20.94 75 19.39 17.70 20.47 75 19.01 18.06 20.86 75 18.63 18.06 20.47 75 18.80 17.78 20. 82 75 18.07 17.78 20.42 75 18.07 18.14 20.42 75 17.71 17.78 20.82 75 21.46 18.19 21.75 75 21.07 18.19 21.35 75 21.46 18.19 21.35 75 21.48 17.83 21.34

PAGE 91

90 Table 4.3 Average dynamic elastic modulus response at elevated temperatures (cont.) Temp Mbrace T 308 High temp LV HV HT DEM (GPa) Average DEM (GPa) Average DEM (GPa) Average 100 17.87 18.12 15.23 15.30 20.43 20.22 100 17.52 17.31 21.21 100 17.17 13.61 20.82 100 17.52 14.25 18.19 100 16.43 14.70 20.65 100 16. 43 12.52 20.26 100 16.10 16.70 21.04 100 16.10 17.39 18.75 100 20.38 14.45 20.33 100 20.77 17.55 20.71 100 20.38 15.45 19.95 100 20.77 14.45 20.33 125 15.80 16.23 17.67 15.72 17.77 19.73 125 19.38 16.59 19.96 125 14.15 15. 55 20.34 125 18.64 14.54 19.58 125 17.13 15.07 20.78 125 15.07 15.07 20.39 125 14.10 14.10 18.86 125 15.74 15.74 18.86 125 18.85 16.38 20.91 125 16.71 17.11 20.53 125 13.76 15.06 18.29 125 15.36 15.73 20.53 150 16.68 11.78 12.12 12.71 15.59 16.37 150 17.03 12.12 15.59 150 16.68 11.84 14.63 150 16.68 12.12 15.91 150 8.84 13.27 16.17 150 9.09 13.57 15.48 150 8.59 13.57 15.48 150 8.10 13.57 16.17 150 10.27 12.28 17.95 150 10.27 13.48 17. 95 150 10.53 12.28 17.95 150 8.53 12.28 17.60

PAGE 92

91 Table 4.3 Average dynamic elastic modulus response at elevated temperatures (cont.) Temp Mbrace T 308 High temp LV HV HT DEM (GPa) Average DEM (GPa) Average DEM (GPa) Average 175 11.42 12.36 13.0 4 12.77 14.06 15.35 175 11.14 13.35 13.15 175 10.87 13.04 13.15 175 11.14 13.66 13.15 175 14.36 14.71 16.24 175 14.03 15.04 16.24 175 13.71 14.39 16.58 175 14.03 14.71 16.24 175 12.11 10.53 16.59 175 11.82 10.26 16.25 1 75 11.53 9.99 16.25 175 12.11 10.53 16.25 200 7.52 7.36 8.94 9.85 9.29 9.88 200 7.07 9.69 9.54 200 7.29 9.19 9.54 200 7.52 12.41 9.54 200 7.32 9.00 9.25 200 6.89 8.76 10.28 200 7.32 9.00 10.54 200 6.89 8.76 9.00 200 7.8 1 10.68 10.52 200 7.35 10.40 10.26 200 7.58 10.68 10.52 200 7.81 10.68 10.26

PAGE 93

92 Table 4.4 a Temperature dependent adhesive concrete interfacial capacity of high viscosity adhesive High Viscosity Adhesive Temp Pu (kN) Average (KN) MPa Average (MPa) Thickness Average Thickness Normalized Normalized Average o C mm MPa/mm 25 2.92 3.08 0.50 0.52 1.60 1.67 0.31 0.31 25 3.34 0.57 1.90 0.30 25 2.98 0.50 1.50 0.33 50 5.81 4.32 0.95 0.73 2.20 1.90 0.43 0.38 50 4.25 0.73 1.80 0.41 50 2.90 0.51 1.70 0.30 75 4.04 2.97 0.66 0.50 2.10 1.73 0.32 0.28 75 3.19 0.54 1.80 0.30 75 1.68 0.30 1.30 0.23 100 0.14 1.96 0.02 0.32 1.40 1.80 0.02 0.16 100 2.75 0.45 2.00 0.23 100 2.99 0.48 2.00 0.24 125 2.34 2. 41 0.39 0.41 2.10 2.10 0.19 0.19 125 2.19 0.38 2.20 0.17 125 2.72 0.45 2.00 0.22 150 2.61 1.93 0.43 0.32 2.00 1.97 0.21 0.16 150 0.98 0.16 1.90 0.08 150 2.18 0.36 2.00 0.18 175 3.16 1.77 0.51 0.29 2.10 2.00 0.24 0.14 175 0.92 0.16 1.90 0.08 175 1.23 0.20 2.00 0.10 200 1.41 1.26 0.24 0.21 2.10 2.13 0.11 0.10 200 1.55 0.26 2.30 0.11 200 0.84 0.14 2.00 0.07

PAGE 94

93 Table 4.4 b Temperature dependent adhesive concrete interfacial capacity of low viscosity adhesive Low Viscos ity Adhesive Temp Pu (kN) Average (KN) MPa Average (MPa) Thickness Average Thickness Normalized Normalized Average o C mm MPa/mm 25 4.83 3.81 0.77 0.62 1.40 1.33 0.55 0.46 25 2.98 0.50 1.20 0.41 25 3.63 0.59 1.40 0.42 50 5.60 4.92 0.95 0 .83 1.70 1.67 0.56 0.49 50 5.54 0.94 1.80 0.52 50 3.62 0.60 1.50 0.40 75 4.04 2.97 0.67 0.50 1.60 1.47 0.42 0.33 75 3.19 0.54 1.50 0.36 75 1.68 0.27 1.30 0.21 100 1.69 2.54 0.28 0.42 1.50 1.60 0.19 0.26 100 3.99 0.64 1.70 0.38 1 00 1.94 0.32 1.60 0.20 125 4.42 2.62 0.74 0.44 1.70 1.60 0.43 0.27 125 1.50 0.25 1.50 0.17 125 1.95 0.33 1.60 0.21 150 2.31 1.81 0.38 0.30 1.60 1.60 0.24 0.19 150 1.10 0.19 1.60 0.12 150 2.04 0.34 1.60 0.21 175 1.85 2.15 0.31 0.3 5 1.80 1.83 0.17 0.19 175 2.75 0.44 1.90 0.23 175 1.87 0.30 1.80 0.17 200 3.30 1.72 0.55 0.29 2.00 1.93 0.27 0.15 200 0.46 0.08 1.90 0.04 200 1.41 0.24 1.90 0.12

PAGE 95

94 Table 4.4 c Temperature dependent adhesive concrete inter facial capacity of high temperature adhesive High Temperature Adhesive Temp Pu (kN) Average (KN) MPa Average (MPa) Thickness Average Thickness Normalized Normalized Average o C mm MPa/mm 25 2.10 2.01 0.42 0.37 0.90 0.77 0.47 0.47 25 2.99 0.50 0. 80 0.62 25 0.93 0.19 0.60 0.32 50 2.77 3.09 0.47 0.51 1.00 1.03 0.47 0.49 50 2.21 0.38 0.90 0.42 50 4.31 0.69 1.20 0.58 75 4.00 3.96 0.67 0.66 1.60 1.47 0.42 0.45 75 3.40 0.58 1.30 0.44 75 4.50 0.73 1.50 0.48 100 2.36 2.22 0.3 9 0.37 1.20 1.10 0.33 0.34 100 2.31 0.39 1.00 0.39 100 1.99 0.33 1.10 0.30 125 2.36 2.36 0.39 0.39 1.10 1.10 0.36 0.37 125 2.35 0.39 0.90 0.44 125 2.36 0.39 1.30 0.30 150 1.91 1.26 0.31 0.21 0.90 0.70 0.34 0.29 150 0.96 0.16 0.60 0.27 150 0.93 0.15 0.60 0.26 175 1.06 1.52 0.18 0.25 0.90 0.97 0.20 0.25 175 0.92 0.15 0.90 0.17 175 2.57 0.41 1.10 0.38 200 0.11 1.01 0.02 0.17 0.50 0.83 0.04 0.17 200 1.06 0.17 0.90 0.19 200 1.86 0.32 1.10 0.29

PAGE 96

95 Figure 4.1 Concrete beam mold being prepared for concrete casting (a) (b) (c) Figure 4.2 Adhesive coupon preparation and casting: (a) mold shape; (b) high viscosity epoxy; (c) low viscosity epoxy Figure 4.3 A dhesive coupon s : (a) during testing; (b) before and after thermal exposure HT 25 o C HV 25 o C LV 25 o C HT 200 o C HV 200 o C LV 200 o C

PAGE 97

96 (a) (b) (c) Fig ure 4.4 Tempera ture dependent residual capacity of adhesive coupons: (a) high viscosity adhesive coupons; (b) low viscosity adhesive coupons; (c ) high temperature adhesive coupons (a) (b) Figure 4.5 Test setup : (a) specimen details (not to scale) ; (b) t est set up configuration using the MTS machine

PAGE 98

97 Figu re 4.6 Temperature data logger

PAGE 99

98 (a) (b) (c) (d) (e) (f)

PAGE 100

99 (g) (h) Figure 4.7 Temperature recording acquired from the data logger: (a ) at 50C [122F] ; (b ) at 75C [167F] ; (c ) at 100 C [212F] ; (d ) at 125C [257F] ; (e ) at 150C [302F] ; (f ) at 175C [347F]; ( g ) at 200C [392F] ; (h) combined temperatures Figure 4.8 Electric stationary diamond saw 200 o C 175 o C 5 0 o C 1 00 o C 75 o C 150 o C

PAGE 101

100 (a) (b) (c) Figure 4.9 DMA results of adhesive: (a) low viscosity; (b) high viscosity; (c) high temperature Figure 4. 10 Frequency test: (a) frequency test set up; (b) frequency test reading

PAGE 102

101 (a) (b) Figure 4.11 Residual impact resonant frequency of interface test specimen (dot = individual; line = average): (a) frequency response; (b) temperature de pendent dynamic e lastic modulus (a) (b) (c) Figure 4.1 2 High viscosity adhesive concrete prism specimen s: (line = individual; dot = average): (a) ul timate load; (b) ultimate capacity; (c) normalized capacity HV epoxy LV epoxy H T epoxy

PAGE 103

102 (a) (b) (c) Figure 4.1 3 Low viscosity adhesive concrete prism sp ecimens: (line = individual; dot = average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity

PAGE 104

103 (a) (b) ( c) Figure 4.1 4 High temperature adhesive concrete prism specimens: (line = individual; dot = average): (a) ultimate load; (b) ultimate capacity; (c) normalized capacity Figure 4.1 5 Normalized bond capacity of concrete prism specimens

PAGE 105

104 (a) (b) Figure 4.16 Failure mode of concrete prisms bonded with different types of adhesive: (a) at 50C [122F]; (b) at 200C [392F] HT 50 o C HT 200 o C HV 50 o C LV 50 o C LV 200 o C HV 200 o C

PAGE 106

105 5. Structural Concrete Performance under Fire 5 .1 Introduction Conserving the structural integrity is the last defense lin e to protect the public from fire. The codes and specification of fire safety resistant describes th e ability of structural element to conserve its functionality under standard fire conditions. Maintaining the structural integrity, temperature transfer, and stability over time is a mean to rate fire resistance structures. The degree of structural fire re sistance depends on the material used, geometry, load integrity, and degree of fire exposure. Traditionally, laboratory tests are carried out on materials that may be exposed to fire to check their performance under the ASTM E119 (ISO834) (2000) Now adays sophisticated modeling based on numerical and analytical analysis is more commonly used. Over the past years, finite element analysis is determining the structural performance under extreme conditions. Finite element analysis is an accurate but yet time c onsuming method to analyze the structural performance under fire. Codes now address structural performance under fire though a simplified empirical design tools just like snow, wind, and earthquake design equations. The Department of T ransportation of Ne w York State conducted a survey showed that 52 out of 1746 surveyed bridges [2.9%] collapsed due to fire (Bronson 2004). In another study, 16 of 503 bridges [3.18%] surveyed bridge collapsed due to fire related issues (Wardhaua and Hadipriono 2003). Furthe r, 26 out of 536 bridges [4.9%] showed bridge collapse due to fire (Scheer 2010).

PAGE 107

106 Fire protection r equirement can be satisfied by two methods: (1) Conventional prescripti ve approach; or (2) Performance based design method approach. The utilization of con ventional prescriptive approach can be through formulated equations and tabulated factors available in building codes and standards that are mostly based on the tests of structural elements in buildings and performed under the ASTM E119 (2000) standard sce nario. The fire temperature and behavior between the buildings and bridges can vary because there is a difference between the cellulosic base fire fuel type over buildings and hydrocarbon base fire fuel type over bridges, which makes the ASTM E119 (2000) f ire scenarios not applicable to bridges The performance based design fire approach is created on engineering principles that find a unique solution to the fire susceptible bridge. Applying the performance based design approach can provide the flexible, e fficient, and economical design. This chapter discusses the importance factor based on the Kodur and Naser (2013) method to evaluate the experimental performance of NSM CFRP concrete blocks interface presented in chapter 3 and the bond performance of concr ete adhesive interface presented in chapter 4 at high temperature. 5.2 Importance Factor Evaluation Approach Kodur and Naser (2013) proposed a method to evaluate the importance factor of a bridge that may be subjected to fire hazard. With fire being a significant threat on the integrity

PAGE 108

107 of the bridge, it must be clear that designing bridges to resist a little risk of fire threat is not economical. Similar to the importance factor of vertical structural design that incorporate the wind and snow loading w hen desired, an importance factor to evaluate fire risks over bridges should be useful (Kodur and Naser 2013). 5.2.1 Importance Factor Calculation The susceptibility of bridge elements to fire is the key element to define the importance factor (IF) of a bridge. The traffic flow has a big influence on the importance factor due to the rapid growth in transporting flammable and highly combustible materials. Based on previous researches over bridges subjected to fire incidents done by the Guide to Highway Vul nerability A ssessment (2002), Garlock et al. (2012), and Elhag and Wang (2007), t he degree of fire exposure of a certain bridge is controlled by the geometric dimensions, material characteristics being used within its design features and the probability of fire occurrence around area surrounding the bridge (Kodur and Naser 2013) Kodur and Naser (2013) classified the importance of a bridge to five classes based on the vulnerability of bridge to fire and critical nature of the bridge. They stated that each parameter is associated with an importance fac tor based on the contribution. Th e y also sub divided the parameters into sub parameters to determine the condition of specific bridge Kodur and Naser (2013) based his parameters and sub parameters weighted ave rages on engineering judgment and design recommendations from previous studies

PAGE 109

108 (Garlock et al. 2012) (Elhag and Wang 2007) (Dwaikat 2011) (Scheer 2010) and discussed in literature review The weightage factors assigned a scale from 1 to 5 A class fa c t or for the bridge various parameters was calculated according to the following equation: max : The maximum weightage factor total : The total maximum weightage factors : Class type factor i : The sub parameter weightage factor : Overall class coe fficient : The product summation of class coefficient x : class parameter (4 ) The ratio of the weightage factors and the overall class coefficient are calculated according to the following Eq (5 ) and Eq. (6 ) respectively ( 5 ) (6 )

PAGE 110

109 5.2.2 Different Type Contribution to the Overall Influence Factor Using Kodur and Nase r (2013) approach, several factors were considered to determine the vulnerability of the NSM CFRP concrete interface interaction and the bond performance of concrete adhesive interface at elevated temperatures. The weightage factors were based on several p arameters that the author though they have contribution on the performance of the specimens over a range of variable elevated temperature. In general, the weightage factor reflects the testing results meaning that as the specimens reaches its critical perf ormance behavior; the weightage factors increases to reflect the highest risk [Table 5.1 5.2] 5.2.2.1 Parameters Type Contribution to the Performance Based approach for NSM CFRP Concrete Interface Interaction Material Properties The author knows the i mportance of strength of materials and the contribution to the overall performance of the test results. With this in mind, the concrete strength and the CFRP tensile capacity will be not included in the weightage factor because they showed no significant c hange in results with increase in temperature (up to what specified in this research)

PAGE 111

110 According to the author the temperature exposure, failure mode, resin type, and capacity drop has played significant role in determining the overall class coefficient ( ) and thus identifying the importance factor that Kodur and Naser (2013) based their performance based risk grades on. Temperature Exposure It should be noted that the temperature exposure influence that overall integrity of the structure under fire ha zard. Hence, from testing results discussed in chapter 3, the temperature range contribute significantly to the performance of the NSM CFRP concrete .27) to the overall performance based criteria. The temperatur e exposure has been identified as type I as shown in Table 5.1 The temperature exposure then was divided to four sub parameters with a maximum temperature to be 200 C [392F] over a range of 50 C [122F] in which some capacity drop were noticed from previ ous tests explained in chapter 3 and 4. Failure Mode F ailure mode is the type II parameter that determines the NSM CFRP blocks interfacial capacity. The failure mode contributes with 20% to the overall performance of the NSM CFRP concrete b According to statistical and experimental analysis, elevated temperature has no effect on the FRP failure where FRP maintain it capacity

PAGE 112

111 over a wide temperature variation range thus it was given a weightage factor of 1 meaning that the FRP failure occurrence will have the least contribution to the interfacial strength of the NSM CFRP concrete blocks. On the other hand, resin failure occurred when the resin deteriorate under thermal stresses or approaching the maximum capacity of the resi n itself. Resin failure has the value of 2 as for the weightage average since it is the second most occurring failure mode determined from the test results compared to the concrete adhesive interface failure which has a weightage factor of 3. Resin Type Two types of adhesive resin were used to bond the NSM CFRP to the concrete blocks The mechanical properties of the High temperature epoxy and ordinary epoxy were discussed in chapter 3 and 4. Based on test results, the high temperature adhesive resin is less susceptible to thermal stresses and thus it was given a weightage factor of 1. With contrast to high temperature epoxy, ordinary epoxy was more vulnerable to temperature changes in which the weightage average reflects the highest risk to fire hazards and therefore was given a weightage average of 2 It is good to note that the type of resin = 0.20).

PAGE 113

112 Capacity Drop The capacity drop is the last parameter type t hat wa s considered in the performance NSM CFRP concrete blocks with 5 sub parameters divided over a 20% drop in strength compared to the control specimens in which t hey were cured at room temperature. The interfacial capacity drops with the increase in temperature. The type of resin plays an important role in controlling the interfacial capacity as well as the thermal stresses. 5.2.2.2 Parameters Type Contribution to the Bond Performance of Concrete adhesive Interface at Elevated Temperatures T he author used NSM CFRP concrete blocks parameters type to evaluate the performance based of concrete adhesive interface at elevated temperature. The only difference was the ad hesive resin thickness that from test results shows flexural capacity variation. The adhesive thickness replaced the failure mode of the NSM CFRP. The adhesive thickness of the resin in between the prism was dependent on the viscosity of the adhesive resin The prism details used to utilize the concrete adhesive interface at el evated temperature were discussed in chapter 4. In addition to the adhesive thickness the low viscosity adhesive was assigned a weightage factor of 2 and the high viscosity adhesive w as assigned a weightage factor of 3 based on previous tests conducted in chapter 3 and 4.

PAGE 114

113 5.3 Importance Factor A nalysis for NSM CFRP Concrete Interface Interaction Based on the same overall class coefficient proposed by Kodur and Naser (2013), Figure 5 .1 shows the overall class coefficient to assign fire risk grade of NSM CFRP concrete blocks bonded with ordinary epoxy and high temperature epoxy The class category was identifi ed in Table 5.1 based on Kodur and Naser (2013) approach to determine the imp or tance factor of the performance base design. The overall class coefficient of the NSM CFRP concrete interface blocks without experiencing thermal distress for ordinary epoxy was 0.45, this will indicate a medium risk grade and importance fa ctor of 1 acco rding to table 5.3 The failure mode of these NSM CFRP concrete interface blocks was within the resi n failure as shown and discussed in chapter 3 The overall class coefficient was preserved up to a temperature of 75C [167F]. This reflects what was state d earlier in chapter 3 that the glass transition temperature ( T g ) of the ordinary epoxy appeared to be approximately between 50C [122 F] and 75C [167F] and hence the overall class coefficient of the NSM CFRP strengthening system was preserved up to 75 C [167F]. The NSM CFRP concrete blocks subjected to a range between 100C [212F] and 150C [302F] exhibited constant overall coefficient. Even if the test results of the interfacial capacity shows some degradation the failure mode of the 100C [212F] and 150C [302F] NSM CFRP concrete blocks was within the interface and the interfacial capacity drop was within the boundary assigned class as shown in table 5.1 thus maintaining a high risk grade and importance factor of 1.2. The overall class coefficie nt and the importance factor tended to increase with an increasing temperature as shown in Fig. 5.2 (a) The degree of strength reduction was 89.8% at 200C [392F ],

PAGE 115

114 which raised the overall class coeff icient to greater than 0.95 and changed the risk grade from critical to high with an importance factor of 1.5 [Fig.5.2] Similar to ordinary epoxy, the same class type was used to high temperature epoxy The overall class coefficient of the high temperature epoxy was lower than the ordinary epoxy as shown i n Fig 5.1 The total interfacial strength drop of high temperature epoxy was 28.5 % over a temperature range from 25C [77F] to 200C [392F], which was a substantially low rate relative to the ordinary epoxy From 75C [167F] to 150C [30 2F] the failu re mode was within the resin for the NSM CFRP concrete blocks specimens with high temperature epoxy. This failure mode led to an increase in the overall coefficient and the ri sk grade from medium to high and thus the importance factor from 1 to 1.2 respect ively [Fig. 5.2(b)] From 150C [302F] to 200C [392F], the failure mode was within the interface. At 200C [392F], the risk grade is critical with high temperature epoxy. 5.4 Importance Factor Analysis for the Bond Performance Concrete adhesive Inter face Similar to NSM CFRP concrete blocks, the high temper ature epoxy remains constant over the temperature range from 25C [77F] to 75C [167F] as shown in Figure 5. 3 The overall class coefficient was 0.33 yielding to a medium risk grade and to import ance factor of 1. From 100C [212F] to 150C [302F], the only type factor that change is the strength reduction from 27.8% to 38.5% respectively, thus led to increase in the overall

PAGE 116

115 class coefficient to 0.48. The risk grade remains preserved at medium an d the importance factor at 1. At 175C [347F], the strength reduction increases to 47% with respect to the control samples at room temperature. The strength reduction was due to the glass transition temperature ( T g ) of the high temperature epoxy that was found to be at 175C [347F]. The risk grade changes from medium to high with an importance factor of 1.2. Similarly, the importance factor remains at 1.2 for the specimens at 200C [392F]. The importance factor is shown in Fig. 5.4(a). The low viscosity epoxy starts with a medium risk grade at 25C [77F] then increases to high risk grade at 50C [122F] and remains constant to reach the glass transition temperature ( T g ) at 75C [167F]. The overall class coefficient kept increasing after 75C [167F] ev en though the risk grade remains at high and the importance factor of 1 as shown in Fig. 5.4(b) The high viscosity epoxy has the lowest normalized strength capacity and the highest overall class coefficient compared to the previous epoxies and thus the i mportance factor was at 1.2 over the temperature range [Fig. 5.4(c)] The controlling factor type over this higher overall class coefficient was the thickness of the adhesive epoxy. The adhesive epoxy thickness was around double the high temperature epoxy. A detailed calculation of the overall class coefficients were shown in the appendix B.

PAGE 117

116 5.5 Utilization of the Importance Factor into Performance based Approach In the case of fire, many factors determine the complex physicochemical transformations of the concrete behavior ( Khoury 2000) The structural element should satisfy load bearing capacity without failure for specific time period in case of fire. The integrity of the structure should be preserved to allow sufficient time for escape and firefighti ng operation to take place as well as safety of the neighboring structures. Fire resistance is defined as the time that the structure to failure under a fire scenario. Design codes aims to guarantee (1) the overall dimensions of the element to preserve acc eptable limits of the heat transfer ; (2) stability failure to maintain acceptable load bearing capacity; (3) integrity failure where separation characteristics under fire (Buchanan 2001) Thus researchers consider the concrete cover as one of the essential fire resistance design to ensure that the reinforcement did not reach critical values. Bisby L. and Stratfor d T. (2013) defines performance based design as it competent structural designers to take any design approach that can be shown to meet the quantified performance objectives for a structure rather than taking a prescriptive Thus performance based design allows the design engineer to meet goals with any rational approa ch that ensure satisfactory structural integrity performance and life safety in the event of fire. The structural performance of FRP reinforced system during fire can be expected to perform similarly to conventional steel reinforced c oncrete in mild heatin g conditions and pre flashover stages ( Buchanan 2001).

PAGE 118

117 The behavior of NSM CFRP concrete interface and bond performance of concrete adhesive interface at elevated temperatures shows a significant loss in strength around the glass transition temperature ( T g ) of the epoxy adhesives. The epoxy adhesives at temperature close to T g results in loss of stiffness transferring to concrete or CFRP strips. The tensile bond of NSM CFRP and the flexural capacity of the concrete adhesive interface were reduced with the i ncrease of temperature as discussed in chapter 3 and 4. According to Khoury (2000), pe rformance based methods can be classified into three categories; (1 ) using simplified calculations based on limit state analysis; (2 ) finite element analysis using ther mo me chanics; (3) advanced method of using thermo hydro mechanical finite element analysis. The author utilized the proposed importance fa ctor in LRFD design that which to the author knowledge will improve the performance of the overall structure. The impo rtance factor proposed earlier will be used as an operational importance factor in the limit state design. (7 ) : Ductility factor : Redundancy factor : Operational importance factor (Importance Factor) : Load Factor : force effect : Resistance factor : Nominal Resistance

PAGE 119

118 5.6 Conclusion ACI 440 ( 2006 2008), Concrete Society (2012), and CSA (2012) provided introductory guidance on how to design FRP strengthened structures. Nevertheless, the guidance pr ovided fall short to be used in a performance design approach (Khoury 2000). Fire resistance design of NSM FRP strengthened concrete elements is not standardized anywhere, according to the author knowledge. Researchers (kodur et al. 2005; Nigro et al. 2010 ; Bisby a nd Kodur 2007 ) con duct numerous fire resistance tests and developed computational models to simulate the be havior of the FRP. With that being said, the FRP communit y did not embrace a performance based approach (Bisby L. and Stratford T. 2013). Ex perienced d esign fire engineers can now support or implement the use of NSM CFRP strengthening systems in fire rated applications under the p erformance based design approach; however, there is a lot of uncertainty and complexity. Further analysis and compu tational finite element modeling of the interface is needed. The tensile strength importance factor of the NSM CFRP proposed can allow additional operational importance factors similar to highly important structures like hospitals.

PAGE 120

119 Table 5.1 Parameters used to predict the overall class coefficient for NSM CFRP interface concrete Class Parameter Sub parameters Weightage Factor Maximum Weightage Factor Ratio Type I Te m perature exposure 25C 75C 1 4 0.3 76C 125C 2 126C 175C 3 176C 200C 4 Type II Failure Mode FRP 1 3 0.2 Resin failure 2 Interface 3 Type III Resin Type High Temperature Epoxy 1 2 0.1 Ordinary High Viscosity Epoxy 2 Type IV Capacity Drop 0% 20% 1 5 0.4 20% 40% 2 40% 60% 3 60% 80% 4 80% 100% 5 14 1.00

PAGE 121

120 Table 5.2 Parameters used to predict the overall class coefficient for bond performance of adhesive concrete interface Class Parameter Sub parameters Weightage Factor Maximum Weightage Factor Ratio Type I Temperature exposure 25 C 75C 1 4 0.3 76C 125C 2 126C 175C 3 176C 200C 4 Type II Adhesive Thickness 0.2mm 0.75mm 1 3 0.2 0.75mm 1.5mm 2 1.5mm 2.25mm 3 Type III Resin Type High Temperature Epoxy 1 3 0.2 Ordinary Low Viscosity Epoxy 2 Ordina ry High Viscosity Epoxy 3 Type IV Capacity Drop 0% 20% 1 5 0.3 20% 40% 2 40% 60% 3 60% 80% 4 80% 100% 5 15 1.00 Table 5.3 Importance factor table criteria based on Kodur and Naser (2013) Risk Grade Overall Class Coefficient Importance Facto (IF) Critical 1.5 High 0.51 0.94 1.2 Medium 0.20 0.50 1.0 Low <0.20 0.8

PAGE 122

121 Figure 5. 1 O verall class coefficient to assign fire risk g rade of NSM CFRP concrete interface (a) (b) Figure 5.2 Importance Factor based on the overall class coefficient of NSM CFRP concrete interface : (a) for HE epoxy; (b) for OE epoxy

PAGE 123

122 Figure 5. 2 O verall class coefficient to assign fire risk gr ade of bond performan ce adhesive concrete interface (a) (b) (c) Figure 5.4 Importance Factor based on the ov er all class coefficient of bond performance adhesiv e concrete interface: (a) for HT epoxy; (b) for L V epoxy ; (c) for H V epoxy

PAGE 124

123 6. Conclusion and Recommendation The thesis presents the testing and analysis the high temperature behavior of near surface mounted carbon fiber reinforced polymer and the bond performance of concrete adhesive interface. The experimental program was performed with several adhesive types that vary in the mechanical properties and glass transient temperature. The additional material tes ting includes the effect of temperature on concrete, CFRP, and adhesive types. A dynamic mechanical analysis was also performed on CFRP, and adhesives. A frequency test and dynamic elastic modulus was performed on the prism specimens. A n importance factor was suggeste d to be used in the performance based approach and then tied to the limit state design. A literature review was showed to present the NSM CFRP application and performance at different temperatures range. Based on the literature review a conclu sion can be drawn: 1. NSM FRP concrete strengthening performs better than the externally bonded FRP which was able to utilize fully the bonding strength. 2. Different failure modes make it difficult to develop a design procedure. 3. Perfor mance based design approa ch reduces costs, and gave the freedom to the design engineer to focus on the goals and not the procedures to maintain an acceptable serviceability. 4. Performance base d helps harmonization of international guideline and the use of new technical applications

PAGE 125

124 5. Performance based design can be applied either probabilistic or deterministic. In probabilistic approach, the end results are measured in terms of the level of risks that are imposed on the attendants and the structure, while the fire growth calculation s, spreading of smoke and behavior of structure are the criteria if deterministic methods were used. 6. Fire safety engineering proposes a method to quantify and assess the performance of the structure with respect to the fire growth, and evacuation. From the result shown in chapters 3 a conclusion can be drawn from this study: 7. The interfacial strength of the NSM CFRP bonded with ordinary epoxy started higher than the high temperature epoxy. 8. The interfacial strength of NSM CFRP bonded with ordinary epoxy and high temperature epoxy dropped around the glass transient temperature. 9. Even the interfacial strength of the NSM CFRP bonded with high temperature was lower; it remains consistent with a slight drop in the overall strength. 10. The compressive conc rete str ength did not drop when subjected to thermal distress ( from 25C [77F] to 200C [392F] ) 11. The CFRP tensile strength testing remains consistent throughout the temperature range from 25C [77F] to 200C [392F]. 12. The NSM CFRP concrete blocks bonded with ord inary epoxy conserved interfacial strength up to 75C [167F] beyond which a noticeable strength decrease was observed

PAGE 126

125 13. Similarly the NSM CFRP concrete blocks bonded with high temperature epoxy maintained the interfacial strength up to 200C [392F] even though a trend of strength reduction was associated with an increasing temperature. 14. A linear load displacement response was noticed in all test specimens until their peak load was achieved, regardless of the type of adhesive, followed by a sudden load drop because of interfacial failure. Hairline cracks formed when the failure of the specimens was impending and further propagated. The failure zone of the interface was localized and several factors seemed to be engaged with such a failure zone, including CFR P configurations and concrete properties. With respect to the chapter 4 where bond performance of concrete adhesive interface was studied at high temperature a summary can be drawn: 15. The thickness of the adhesive controls the ultimate capacity and was us ed to normalize the load displacement results instead of using the bonded area alone due to a non uniform quantity of adhesive used to bond the concrete prisms and other mentioned factors. 16. The high temperature adhesive perform the best with the temperature range from 25C [77F] to 200C [392F] over the low viscosity adhesive and the high viscosity adhesive. 17. A strength reduction were associated with the three types of adhesives over the temperature range from 25C [77F] to 200C [392F]. The load displace ment

PAGE 127

126 response was linear until the maximum capacity and then a sudden load drop were realized. 18. The failure mode of the adhesives ranges between concrete adhesive failure and bond failure what was independent to temperature range. With respect to the cha pter 5 where performance based design approach was suggested a summary can be drawn: 19. The performance based design approach was proposed to ensure the structural fire safety of C FRP strengthe ned concrete members 20. Engineers may use the performance based des ign method approach to quantify adequate FRP performance in fire design. 21. The proposed imp ortance factor can increase safety and can be utilized in the limit state design approach. 22. The overall class coefficient of the high temperature performs the best ove r the high viscosity and the low viscosity epoxy /ordinary epoxy when considering the parameters specified in Table 5.2 and Table 5.3.

PAGE 128

127 6.1 Recommendation for Future Work 1. Develop a proper thermally isolated design method for NSM CFRP reinforcement duri ng fire; 2. Determine a proper minimum depth of concrete cover to ensure that deferential thermal expansion of NSM CFRP during fire will not occur; 3. Generalize in codes the minimum mechanical properties of epoxy adhesives to be used in NSM CFRP repair procedu res; 4. Determine if cementitious adhesive ba sed epoxy will improve the fire performance; 5. Reduce the complexity and uncertainty of the prescriptive desi gn method and propose a performance based design approach during fire; 6. Perform a finite element analysis to confirm the proposed method performance based design approach. 7. Study the effect of adhesive thickness on the overall interfacial adhesive concrete strength.

PAGE 129

128 R EFERENCES ACI, 2006. Guide for the design and construction of structural concrete reinforced with FRP bars (ACI.440.1 06), American Concrete Institute, Farmington Hills, MI. ACI. 2008. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures (ACI.440.2R 08), American Concrete Institute, Farming ton Hills, MI. Aidoo, J., Harries, K., and Petrou, M. 2006. Full scale experimental investigation of repair of reinforced concrete interstate bridge using CFRP materials, Journal of Bridge Engineering, 11(3), 350 358. Alkhrdaji, T., Nanni, A., Chen, G., and Baker, M. 1999. Upgrading the transportation infrastructure: soild RC decks strengthened with FRP, Concrete International, 21(10), 37 41. ASTM, 2010, ASTM D638 10 Standard t est Method for Tensile Properties of Plastics ASTM International. ASTM, 2009 ASTM C78 09 Standard test method for flexural strength of concrete (using simple beam with third point loading ASTM International ASTM, 2007. ASTM E1356 03 Standard test met hod for the assignment of glass transition temperatures by di fferential scanni ng calorimetry ASTM International p 4. AS TM, 2008 ASTM C215 02 Standard test method for fundamental transverse, longitudinal, and torsional frequencies of concrete spe cimens, ASTM International ASTM, 2000. ASTM E119 00a Standard Methods of Fire tests of Building Construction and Materials, ASTM International. Badawi, M. and Soudki, K. 2009. Fatigue behavior of RC beams strengthened with NSM CFRP rods, Journal of Composites for Construction, 13(5), 415 421.

PAGE 130

129 Barros, J. and Fortes, A. 2005. Flexural st rengthening of concrete beams with CFRP laminates bonded into slits, Cement and Concrete Composites, 27, 471 480. Bertolotti, E. A., 2012. Near Surface Mounted Fiber Reinforced Polymer Strips for Strengthening of Reinforced Concrete Girders, PhD Thesis, A uburn Unversity, AL. Bisby, L. A., and Kodur, V.K.R. 2007. Evaluating the fire endurance of concrete slabs reinforced with FRP Bars: Considerations for a holistic approach. Composites Part B. 38: 547 559.. doi: 10.1016/j.compositesb.2006.07.013. Bisby L. A., Statford, T. 2013. Design for fire of concrete elements strengthened or reinforced with fibre reinforced polymer: state of the art and opportunities from performance based approaches, Canadian Journal of Civil Engineering, 40:1 10. Blaschko, M. 2003. Bond behavior of CFRP strips glued into slits, Proceedings of Fiber Reinforced Polymer Reinforcement for Concrete Structures (FRPRCS 6), 205 214. Boring, D.F., Spence, J.C., and Wells, W.G. 1981. Fire Protection Through Modern Building Codes, American Iro n and Steel Institute, 5 th Edition, Washington, DC. Buchanan, A. 2001. Structural design for fire safety. John Wiley & Sons, New York. Bukowski, R. W. and Babrauskas, V., 1994. Developing Rational, Performance Based Fire Safety Requirements in Model Buil ding Codes, Fire and Materials, 18(3), 173 191. Building Industry Authority, Approved Documents C1, C2, C3, C4, New Zealand Building Code, New Zealand, 1994. Burke, P .J. 2008. Low and high temperature performance of near surface mounted FRP strengthened Burke, P.J., Bisby, L.A., and Green, M.F. 2013. Effects of elevated temperature on near surface mounted and externally bonded FRP strengthening systems for concrete, Cement and Conc rete Composites, 35, 190 199. Canadian Commission on Building and Fire Codes, 1994. Draft Strategic Plan Final Report of CCBFC Strategic Planning Task Group, National Research Council of Canada.

PAGE 131

130 Concrete Society, 2012. Thenical report 55: Design guidanc e for strengthening concrete structures using fibre composite materials, 3/e. The Concrete Society, Camberley UK. Carmichael, B. M., and R. W. Barnes. 2005. Repair of the Uphapee Creek Bridge with FRP Laminates Auburn, AL: Auburn University Highway Resea rch Center, Final Report RP 930 466 2. Case Histories and Use of FRP for Prestressing Applications SP 245CD eds. Raafat El Hacha and Sami H. Rizkalla, Farmington Hills, MI: American Concrete Institute, 143 164. CD ROM. Chaallal, O., Mofidi, A., Benmokra ne, B,. and Neale, K. 2011. Embedded through section FRP rod method for shear strengthening of RC beams: performance and comparison with existing techniques, Journal of Composites for Construction, 15(3), 374 383. CIB Report, 1982. Working with the Perfor mance Approach in Building, Publication 64, International Council for Building Research Studies and Documentation, Rotterdam. CSA, 2012. S806 12: Design and construction of building structures with fibre reinforced polymers. Canadian Standards Association Mississauga, Canada. Custer, P.L.P., 1995. Introduction to the Use of Fire Dynamics in Performance Based Design, Proceedings of the SFPE Technical Symposium: Applications of Fire Dynamics, University of British Columbia, Vancouver, British Columbia. D e Lorenzis, L. Nanni, A., and La Tegola, A. 2000. Strengthening of reinforced concrete structures wit h near surface mounted FRP rods bibl. International Meeting on Composite Masterials, PLAST 20 00, Milan, Italy, May 9 11 p 8. De Lorenzis, L. and Teng, J.G. 2007. Near surface mounted FRP reinforcement: an emerging technique for strengthening structures, Composites Part B, 38, 119 143. Dimitrienko, Y.I. 1999. Thermomechanics of composites under high temperatures, Klewer Academic Publishers, London, UK.

PAGE 132

131 El Hacha, R. and Rizkalla, S. 2004. Near surface mounted fiber reinforced polymer reinforcements for flexural strengthening of concrete structures, ACI Structural Journal, 101(5), 717 7 6. Ferguson, A.P.S., CIB W14 Symposum, 1993. Fire Safety Equivalence Who Needs it?, International Symposium and Workshops on Engineering Fire Safety in the Process of Design: Demonstrationg Equivalency, Proceedings Part 2, CIB W14: Fire Safety Engineering, University of Ulster at Jordanstown, Antrim, Northern Ireland, 1 12. Foster, S.K. and Bisby, L.A. 2008. Fire survivability of externally bonded FRP strengthening systems, Journal of Composites for Construction, 12(5), 553 561. Garlock, M., Paya Zaforteza, I., Kodur, V., and Gu, L. 2012. Fire hazard in bridges: review, as sessment and repair strategies, Engineering Structures, 35, 89 98. Gentry, T.R. and Hussain, M. 1999. Thermal compatibility of concrete and composite reinforcements, Journal of Composites for Construction, 3(2), 82 86. Hassan, T. and Rizkalla, S. 2004. B ond mechanism of near surface mounted fiber reinforced polymer bars for flexural strengthening of concrete structures, ACI Structural Journal, 101(6), 830 839. Hadjisophocieous, G.V., Benichou, N., and Tamim, A. S. 1998. Literature Review of Performance B ased Fire Codes and Design Environment, Journal of Fire Protection Engineering 9:12. Haviland, D. S. 1978. Toward a Performance Approach to Life Safety from Fire in Building Codes and Regulations, Center for Fire Research, National Bureau of Standards, US Department of Commerce, NBS GCR 78 118. Hughes Brothers. 2008. Carbon fiber reinforced polymer (CFRP) tape Aslan 500, Seward, NE. International Standard ISO 6241, 1984. Performance standards in buildings principles for their preparation and factors to be considered, First Edition, International Organization for Standardization, Switzerland.

PAGE 133

132 Kim, Y.J., Hmidan, A., Choi, K., and Yazdani, S. 2012. Fracture characteristics of notched concrete beams shear strengthened with CFRP sheets subjected to high tem perature, ACI Special Publication (SP 286): A Fracture Approach for FRP concrete Structures, American Concrete Institute Farmington Hills, MI. Kim, S.E., Choi, S., and Ma S. 2003 P erformance based design of steel arch bridges using practical inelastic n onlinear analysis, Journal of Constructional Steet Research 59, 91 108. Khoury, G. A., 2000, Effect of fire on concrete and concrete structures, Prog. Struct. Engng Master John Wiley & Sons, Imperial Collage, London, 2:429 447. Kodur, V.K.R. and Yu, B. 2 013. Evaluating the fire response of concrete beams strengthened with near surface mounted FRP reinforcement, Journal of Composites for Construction, 17(4), 517 529. Kodur, VK.R., Bisby, L.A. and Foo, S. 2005. Thermal behavior fire exposed concrete slabs reinforced with fiber reinforced polymer bars. ACI Materials Journal, 102(6): 799 807. Liew, J. Y. R. Tang, L. K., and Choo, Y. S. 2002. Advanced analysis for performance based des ign of steel structures exposed to fires Journal of Structural Engineeri ng ASCE, USA, Vol. 128, No. 12, pp. 1584 1593. MBrace. 2007. Wabo MBrace CF130 unidirectional high strength carbon fiber fabric, Watson Bowman Acme, Amherst, MA. Meacham, B. J., and Custer, R.LP., 1995. Performance Based Fire Safety Engineering: An Int roduction to Basic Concepts, Journal of Fire Protection Engineering, 7( 2 ), 35 54. the NFPA Annual Meeting, Philadelphia. Nigro, E., Cefarelli, G., Bilotta, A., Manfredi, G ., and Cosenza E. 2010. Mechanical Behavior of Concrete Slabs Reinforced with FRP Bars in Case of Fire: Experimental Investigation and Numerical Simulation. In 3 rd fib International Congress, Washington DC, 29 th May to 2 nd June. P. 15.

PAGE 134

133 Oehlers, D. J., M. Haskett, C. Wu, and R. Seracino. 2008. Embedding NSM FRP Plates for Improved IC Debonding Resistance. Journal of Composites for Construction 12, no. 6: 635 642. Palmieri, A., Matthys, S., and Taerwe, L. 2012. Experimental investigation on fire endurance o f insulated concrete beams strengthened with near surface mounted FRP bar reinforcement, Composites Part B, 43, 885 895. PTM&W Industries. 2013. High temp epoxy laminating resin (PT2846), PTM&W Industries, Inc. Santa Fe Springs, CA. Rashid, R., D. J. Oeh lers, and R. Seracino. 2008. IC Debonding of FRP NSM and EB Retrofitted Concrete: Plate and Cover Interaction Tests. Journal of Composites for Construction 12, no. 2: 160 167. Rostasy, F. 1992. Fiber composite elements and techniques as non metallic reinf orcement of concrete, Brite project 4142/BREU CT910515, Evaluation of Potential and Production Technologies of FRP, Technical Report Task 1. Sena Cruz, J.M., Barros, J. A.O., Gettu, R., and Azevedo, A.F.M. 2006. Bond behavior of near surface mounted CFRP laminate strips under monotonic and cyclic loading, Journal of Composites for Construction, 10(4), 295 303. Sena Cruz, J. and Barros, J. 2004. Bond between near surface mounted carbon fiber reinforced polymer laminate strips and concrete, Journal of Comp osites for Construction, 8(6), 519 527. Seracino, R., Jones, N., Ali, M., Page, M., and Oehlers, D. 2007. Bond strength of near surface mounted FRP strip to concrete joints, Journal of Composites for Construction, 11(4), 401 409. Stone, D., Tumialan, G., and Nanni, A. 2002. Near surface mounted FRP reinforcement: application of an emerging technology, Concrete, 36(5), 42 44.

PAGE 135

134 Sumida, A., Fujisaki, T., Watanabe, K., and Kato, T. 2001. Heat resistance of continuous fiber reinforced plastic rods, 5 th Interna tional Conference on Fibre reinforced Plastics for Reinforced Concrete Structures (FRPRCS 5), Cambridge, UK. T ljsten, B., and H. Nordin. 2007. Concrete Beams Strengthened with External Prestressing Using External Tendons and Near Surface Mounted Reinfor cement (NSMR). In SP 245CD Tljsten, B ., Carolin, A., and Nordin, H. 2003. Concrete structures strengthened with near surface mounted reinforceme nt of CFRP. Advances in Structural Engineering 6(3), 201 213. Teng, J.G., De Lorenzis, L., Wang, B., Li, R ., Wong, T., and Lam, L. 2006. Debonding failures of RC beams strengthened with near surface mounted CFRP strips, Journal of Composites for Construction, 10(2), 92 105. Vasquez Rayo, D. L. 2008. Plate End Debonding of Longitudinal Near Surface Mounted Fib er Reinforced Polymer Strips on Reinforced Concrete Flexural Members. M.S. thesis, North Carolina State University. Wakamatsu, T., 1988. Development of Design System for Building Fire Safety, Fire Safety Science Proceedings of Second International Symposi um, Wakamatsu T. et al. Edis., International Association for Fire Safety Science, Tokyo, Japan, 881 895. Wehrili, R. and Kapash, R. 1972. Hospital Bedrooms and Nursing Units. A Systems Approach for Building Technology, NMS Report 10972, Center for Buildin g Technology, National Bureau of Standards, US Department of Commerce.

PAGE 136

135 Appendix A Additional detailed figures of the process for the NSM CFRP concrete block testing with different failure mode

PAGE 137

136

PAGE 138

137

PAGE 139

138

PAGE 140

139

PAGE 141

140 Ordinary Epoxy 25C Delta lamda 50C Delta lamda Type I 1 0.250 0.071 Type I 1 0.250 0.071 Type II 2 0.667 0.143 Type II 2 0.667 0.143 Type III 2 1.000 0.143 Type III 2 1.000 0.143 Type IV 1 0.200 0.071 Type IV 1 0.200 0.071 Sum 0.429 Sum 0.429 75C Delta lamda 100C Delta lamda Type I 1 0.250 0.071 Type I 2 0 .500 0.143 Type II 2 0.667 0.143 Type II 3 1.000 0.214 Type III 2 1.000 0.143 Type III 2 1.000 0.143 Type IV 1 0.200 0.071 Type IV 2 0.400 0.143 Sum 0.429 Sum 0.643 125C Delta lamda 150C Delta lamda Type I 2 0.500 0.143 Type I 3 0.750 0.214 Type II 3 1.000 0.214 Type II 3 1.000 0.214 Type III 2 1.000 0.143 Type III 2 1.000 0.143 Type IV 2 0.400 0.143 Type IV 1 0.200 0.071 Sum 0.643 Sum 0.643 175C Delta lamda 200C Delta lamda Type I 3 0.750 0.214 Type I 4 1.000 0.2 86 Type II 3 1.000 0.214 Type II 3 1.000 0.214 Type III 2 1.000 0.143 Type III 2 1.000 0.143 Type IV 2 0.400 0.143 Type IV 5 1.000 0.357 Sum 0.714 Sum 1.000

PAGE 142

141

PAGE 143

142

PAGE 144

143

PAGE 145

144

PAGE 146

145

PAGE 147

146

PAGE 148

147

PAGE 149

148

PAGE 150

149

PAGE 151

150

PAGE 152

151 High Temperature Epoxy 25C Delta lamda 50C Delta lamda Type I 1 0.250 0.071 Type I 1 0.250 0.071 Type II 2 0.667 0.143 Type II 2 0.667 0.143 Type III 1 0.500 0.071 Type III 1 0.500 0.071 Type IV 1 0.200 0.071 Type IV 1 0.200 0.071 Sum 0.357 Sum 0.357 75C Delta lamda 100C Delta lamda Type I 1 0.250 0.071 Type I 2 0.500 0.143 Type II 2 0.667 0.143 Type II 2 0.6 67 0.143 Type III 1 0.500 0.071 Type III 1 0.500 0.071 Type IV 1 0.200 0.071 Type IV 1 0.200 0.071 Sum 0.357 Sum 0.429 125C Delta lamda 150C Delta lamda Type I 2 0.500 0.143 Type I 3 0.750 0.214 Type II 2 0.667 0.143 Type II 3 1.000 0.214 Type III 1 0.500 0.071 Type III 1 0.500 0.071 Type IV 2 0.400 0.143 Type IV 1 0.200 0.071 Sum 0.500 Sum 0.571 175C Delta lamda 200C Delta lamda Type I 3 0.750 0.214 Type I 4 1.000 0.286 Type II 3 1.000 0.214 Type II 3 1. 000 0.214 Type III 1 0.500 0.071 Type III 1 0.500 0.071 Type IV 2 0.400 0.143 Type IV 2 0.400 0.143 Sum 0.643 Sum 0.714

PAGE 153

152 Appendix B The figures of the bond performance of concrete adhesive prisms that shows the failure mode.

PAGE 154

153

PAGE 155

154

PAGE 156

155

PAGE 157

156 Low Viscosity Adhesive 25C Delta lamda 50C Delta lamda Type I 1 0.3 0.067 Type I 1 0.3 0.067 Type II 2 0.7 0.133 Type II 3 1.0 0.200 Type III 2 0.7 0.133 Type III 2 0.7 0.133 Type IV 1 0.2 0.067 Type IV 2 0.4 0.133 Sum 0.400 Su m 0.533 75C Delta lamda 100C Delta lamda Type I 1 0.3 0.067 Type I 2 0.5 0.133 Type II 2 0.7 0.133 Type II 3 1.0 0.200 Type III 2 0.7 0.133 Type III 2 0.7 0.133 Type IV 3 0.6 0.200 Type IV 3 0.6 0.200 Sum 0.533 Sum 0.667 125C Delta lamda 150C Delta lamda Type I 2 0.5 0.133 Type I 3 0.8 0.200 Type II 3 1.0 0.200 Type II 3 1.0 0.200 Type III 2 0.7 0.133 Type III 2 0.7 0.133 Type IV 3 0.6 0.200 Type IV 3 0.6 0.200 Sum 0.667 Sum 0.733 175C Delta lam da 200C Delta lamda Type I 3 0.8 0.200 Type I 4 1.0 0.267 Type II 3 1.0 0.200 Type II 3 1.0 0.200 Type III 2 0.7 0.133 Type III 2 0.7 0.133 Type IV 3 0.6 0.200 Type IV 4 0.8 0.267 Sum 0.733 Sum 0.867

PAGE 158

157 High Viscosity Adhesive 25C Delta lamda 50C Delta lamda Type I 1 0.3 0.067 Type I 1 0.3 0.067 Type II 3 1.0 0.200 Type II 3 1.0 0.200 Type III 3 1.0 0.200 Type III 3 1.0 0.200 Type IV 1 0.2 0.067 Type IV 2 0.4 0.133 Sum 0.533 Sum 0.600 75C Delta lamda 1 00C Delta lamda Type I 1 0.3 0.067 Type I 2 0.5 0.133 Type II 3 1.0 0.200 Type II 3 1.0 0.200 Type III 3 1.0 0.200 Type III 3 1.0 0.200 Type IV 1 0.2 0.067 Type IV 3 0.6 0.200 Sum 0.533 Sum 0.733 125C Delta lamda 150C Delta lamda Type I 2 0.5 0.133 Type I 3 0.8 0.200 Type II 3 1.0 0.200 Type II 3 1.0 0.200 Type III 3 1.0 0.200 Type III 3 1.0 0.200 Type IV 3 0.6 0.200 Type IV 3 0.6 0.200 Sum 0.733 Sum 0.800 175C Delta lamda 200C Delta lamda Type I 3 0.8 0. 200 Type I 4 1.0 0.267 Type II 3 1.0 0.200 Type II 3 1.0 0.200 Type III 3 1.0 0.200 Type III 3 1.0 0.200 Type IV 3 0.6 0.200 Type IV 4 0.8 0.267 Sum 0.800 Sum 0.933

PAGE 159

158 High Temperature 25C Delta lamda 50C Delta lamda Type I 1 0 .3 0.067 Type I 1 0.3 0.067 Type II 2 0.7 0.133 Type II 2 0.7 0.133 Type III 1 0.3 0.067 Type III 1 0.3 0.067 Type IV 1 0.2 0.067 Type IV 1 0.2 0.067 Sum 0.333 Sum 0.333 75C Delta lamda 100C Delta lamda Type I 1 0.3 0.067 Type I 2 0.5 0.133 Type II 2 0.7 0.133 Type II 2 0.7 0.133 Type III 1 0.3 0.067 Type III 1 0.3 0.067 Type IV 1 0.2 0.067 Type IV 2 0.4 0.133 Sum 0.333 Sum 0.467 125C Delta lamda 150C Delta lamda Type I 2 0.5 0.133 Type I 3 0.8 0.200 Typ e II 2 0.7 0.133 Type II 1 0.3 0.067 Type III 1 0.3 0.067 Type III 1 0.3 0.067 Type IV 2 0.4 0.133 Type IV 2 0.4 0.133 Sum 0.467 Sum 0.467 175C Delta lamda 200C Delta lamda Type I 3 0.8 0.200 Type I 4 1.0 0.267 Type II 2 0.7 0.133 Type II 2 0.7 0.133 Type III 1 0.3 0.067 Type III 1 0.3 0.067 Type IV 3 0.6 0.200 Type IV 3 0.6 0.200 Sum 0.600 Sum 0.667

PAGE 160

159

PAGE 161

160

PAGE 162

161

PAGE 163

162

PAGE 164

163

PAGE 165

164

PAGE 166

165

PAGE 167

166

PAGE 168

167

PAGE 169

168

PAGE 170

169

PAGE 171

170 Appendix C The dynamic mechanical analysis test specimens and set up.

PAGE 172

171

PAGE 173

172 Appendix D Frequency tests performed on concrete prisms and the dynamic elastic modulus in pictures.

PAGE 174

173 Appendix E Three Types of Epoxy Adhesive Testing over the Temperature Range along with Str ess Strain Figures.

PAGE 175

174

PAGE 176

175

PAGE 177

176

PAGE 178

177

PAGE 179

178

PAGE 180

179

PAGE 181

180

PAGE 182

181 Appendix F CFRP Testing over the specified temperature r ange

PAGE 183

182