Bond and Flexural Behaviour of Self Consolidating Concrete Beams Reinforced and Prestressed with FRP Bars

Abstract

Self consolidating concrete (SCC) is widely used in the construction industry. SCC is a high performance concrete with high workability and consistency allowing it to flow under its own weight without vibration and making the construction of heavily congested structural elements and narrow sections easier. Fiber reinforced polymer (FRP) reinforcement, with its excellent mechanical properties and non-corrosive characteristic, is being used as a replacement for conventional steel reinforcement. In spite of the wide spread of SCC applications, bond and flexural behaviour of SCC beams reinforced or prestressed with FRP bars has not been fully studied. Furthermore, the ACI 440.1R-06 equation for determining the development length of FRP bars is based on Glass FRP (GFRP) bars and may not be applicable for Carbon FRP (CFRP) bars. This research program included an experimental and analytical study to investigate the flexural and bond behaviour of SCC beams reinforced with FRP bars and SCC beams prestressed with CFRP bars. In the experimental phase, fifty-six beams were fabricated and tested. Sixteen of these beams were prestressed with CFRP bars and forty beams were reinforced with non-prestressed GFRP or CFRP bars. Four concrete batches were used to fabricate all the specimens. Three mixes were of self consolidating concrete (SCC) and one mix was of normal vibrated concrete (NVC). The test parameters for the non-prestressed beams were the concrete type, bar type and bar diameter, concrete cover thickness and embedment length while the test parameters for the prestressed beams were the concrete type and the prestressing level (30%, 45% and 60%). The transfer length of the prestressed CFRP bars was determined by means of longitudinal concrete strain profile and draw-in methods. All beams were tested in four-point bending to failure. Measurements of load, midspan deflection, bar slip if any at the beam ends, strain in reinforcing FRP bar at various locations, and strain in concrete at the beam midspan were collected during the flexural test. The concrete compressive strength at flexural tests of SCC mix-1, mix-2, and mix-3 were 62.1MPa, 49.6MPa and 70.9MPa, respectively and for the NVC mix was 64.5MPa. The material test results showed that SCC mixes had lower modulus of elasticity mechanical properties than the NVC mix. The modulus of elasticity of the SCC mixes ranged between 65% and 82% of the NVC mix. The modulus of rupture of the SCC mixes was 86% of the NVC mixes. The test results for beams prestressed with CFRP bars revealed that the variation of transfer length of CFRP bars in SCC versus their prestressing level was nonlinear. The average measured transfer lengths of 12.7mm diameter CFRP bars prestressed to 30%, 45% and 60% was found to be 25db, 40db, 54db, respectively. Measured transfer lengths of the 12.7mm diameter CFRP bar prestressed to 30% in SCC met the ACI440.4 prediction. However, as the prestressing level increased, the predicted transfer length became unconservative. At a 60% prestress level, the measured/prediction ratio was 1.25. Beams prestressed with CFRP bars and subjected to flexural testing with shear spans less than the minimum development length had local bar slippage within the transmission zone. Beams that experienced local bond slip, their stiffness was significantly decreased. A modification to the existing model used to calculate the transfer and development lengths of CFRP bars in NVC beams was proposed to account for the SCC. The test results for beams reinforced with FRP bars indicated that the average bond strength of CFRP bars in NVC concrete is about 15% higher than that of GFRP bars in NVC. The ACI 440.1R-06 equation overestimated the development length of the CFRP bars by about 40%, while CAN/CSA-S6-06 equation was unconservative by about 50%. A new factor of (1/1.35) was proposed to estimate the development length of the CFRP bars in NVC when the ACI440.1R-06 equation is used. Beams made from SCC showed closer flexural crack spacing than similar beams made from NVC at a similar loading. The deflection of beams made from SCC and reinforced with CFRP bars was found to be slightly larger than those made from NVC. The average bond stresses of GFRP and CFRP bars in SCC were comparable to those in NVC. However, FRP bars embedded in SCC beams had higher bond stresses within the uncracked region of the beams than those embedded in NVC beams. In contrast, FRP bars in SCC had lower bond stresses than FRP bars in NVC within the cracked region. The average bond strength of GFRP in SCC was increased by 15% when the concrete cover thickness increased from 1.0db to 3.0db. Cover thicknesses of 2db and 3db were found to be sufficient to prevent bond splitting failure of GFRP and CFRP bars in SCC, respectively. Bond splitting failure was recorded when the cover thickness dropped to 1.5db for the GRP bars and to 2.0db for the CFRP bars. An insignificant increase in average bond stress was found when the bar diameter decreased from 12.7mm to 6.3mm for the CFRP bars, and a similar increase occurred in GFRP bars when the bar diameter decreased from 15.9mm to 9.5mm. New models to calculate the development length of GFRP and CFRP bars embedded in SCC were proposed based on the experimental results. These models capture the average bond stress profile along the embedment length. A good agreement was found between the proposed model and the experimental results. Analytical modeling of the load-deflection response based on the effective moment of inertia (ISIS Canada M5) was unconservative for SCC beams reinforced with CFRP bars by 25% at ultimate loading. A new model for bond stress versus Ma/Mcr (applied moment to cracking moment) ratio was developed for GFRP and CFRP bars in SCC and for CFRP bars in NVC. These bond stress models were incorporated in a new rigorous model to predict the load-deflection response based on the curvature approach. The FRP bar extension and bond stress models were used to calculate the load-deflection response. With these models 90% of the calculated deflections were found to be within ± 15% of the experimental measured results for SCC beams reinforced with FRP bars. Analytical modeling of the load-deflection for NVC and SCC beams prestressed with CFRP bars are proposed done. The moment resistance was calculated using Sectional Analysis approach. The deflection was calculated using simplified and detailed methods. The simplified method was based on the effective moment of inertia while the detailed method was based on effective moment of inertia and effective centroid. The experimental results correlated well with the detailed method at higher loads range. This study provided an understanding of the mechanism of bond and flexural behaviour of FRP reinforced and prestressed SCC beams. The information presented in this thesis is valuable for designers using FRP bars as flexural reinforcement and also for the development of design guidelines for SCC structures.