Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Civil and Environmental Engineering

Major Professor

Rajan Sen, Ph.D.

Co-Major Professor

Gray Mullins, Ph.D.

Committee Member

Autar Kaw, Ph.D.

Committee Member

Kandethody Ramachandran, Ph.D.

Committee Member

Abla Zayed, Ph.D.


Composite, Durability, Pile, Porosity, Testing


This dissertation presents findings from three separate investigations, a laboratory study and two field studies that evaluated the durability of the Fiber Reinforced Polymer (FRP)-concrete bond. The laboratory study explored the role of porosity on CFRP-concrete bond following immersion in warm water. Two disparate field studies measured residual bond after 20 years outdoor exposure of FRP repairs of full-size masonry walls and after 12 years for partially submerged piles supporting the Friendship Trail Bridge, Tampa Bay.

The ACI 440 code requires the same surface preparation for all externally bonded FRP concrete repairs. This disregards the role of porosity that is a function of the water / cementitious (w/c) ratio. Concretes with high w/c ratios are low strength concretes, have large voids and a more elaborate capillary pore network compared to low w/c, high strength concretes. Epoxies will therefore penetrate deeper into high porosity concretes. As a result, the performance of low strength, high porosity concrete under moisture exposure can be anticipated to be superior. The laboratory study was intended to determine whether this hypothesis was correct or not.

Three different concrete mixes with water / cementitious ratios of 0.73, 0.44 and 0.25 representing high, medium and low porosities were used for the study. The corresponding target compressive strengths were 2,500 psi, 5,000 psi and 7,500 psi respectively. A total of eighteen, 9 in. x 9 in. x 2.5 in. thick slabs, three for each concrete porosity were tested. Slabs were allowed to cure for over 90 days before surfaces were lightly sand blasted to provide the required concrete surface profile (CSP 3). Specimens were then pre-conditioned in an oven for 48 hours to ensure uniform drying.

Concrete porosity was characterized using mercury porosimetry, SEM, 3D surface scanning and images obtained using a portable microscope. Two commercially available CFRP materials were bonded to the oven-dried prepared slab surfaces and the epoxy allowed to cure at room temperature for 4 weeks. Twelve FRP bonded slabs were completely submerged in potable water at 30 oC (86 oF) as part of the aging program. The six remaining slabs were used for establishing baseline bond values through destructive pull-off tests. The twelve exposed slabs were similarly tested following 15 weeks of exposure.

Results showed minimal degradation in the high porosity, low strength concrete but over 20% reduction in the low porosity, higher strength concrete. Analysis of the failure plane indicated that the lower porosity of the high strength concrete had limited the depth to which the epoxy could penetrate. This was confirmed from magnified images of the bond line taken using a microscope and from a careful assessment of the failure mode. Findings also suggest that the CSP 3 surface profile (light sand blasting) may be adequate for lower strength concrete but not so for higher strength concrete. For applications where FRP concrete repairs of higher strength concrete are permanently or intermittently exposed to moisture, alternative surface preparation may be needed to allow epoxy to penetrate deeper into the concrete substrate. The viscosity of the resin hitherto not considered may be a critical parameter.

In 1995, two full-scale concrete masonry walls were repaired using three horizontally aligned 20 in. (508 mm) wide uni-directional carbon fiber sheets using different commercially available epoxies. Twenty years later the CFRP-CMU bond was determined through selective pull-off tests that were preceded by detailed non-destructive evaluation. Results showed that despite superficial damage to the top epoxy coating and debonding along masonry joints, the residual CFRP-CMU bond was largely unaffected by prolonged exposure to Florida’s harsh environment.

Therein, 99% of samples exhibited in cohesive failure of the CMU or mortar. Pull-off strength was poorer at mortar joints but because the CFRP was well bonded to the masonry surface, its impact on structural performance of the repair was expected to be minimal. Overall, the repairs proved to be durable with both epoxy systems performing well.

The Friendship Trail Bridge linking St. Petersburg to Tampa FL was demolished in 2016. This was the site of three disparate demonstration projects in which 13 corroding reinforced concrete piles were repaired using fiber reinforced polymer (FRP) in 2003-04, 2006, and 2008. The repairs were undertaken using combinations of carbon and glass fiber, pre-preg and wet layup, epoxy and polyurethane resin, and were installed using either shrink wrap or pressure bagging. Residual FRP-concrete bond was evaluated after up to 12 years of exposure through 120 pull-off tests conducted on 10 representative repaired piles. Results showed a wide variation in the measured pull-off strength depending on the type of resin, the number of FRP layers, the prevailing conditions at the time the epoxy was mixed and the method of installation. Epoxy-based systems were found to be sensitive to ambient conditions at installation. Pressure bagging improved performance. The highest residual bond was recorded in pressure bagged piles repaired in 2008. The findings suggest that in marine environments epoxy-based systems installed using pressure bagging can lead to durable repairs.