Strengthening reinforced concrete beams using FRP composite fabrics : full-scale experimental studies and design concept verification

Graduation date: 2001The Horsetail Creek Bridge (HCB), constructed in 1914, is located along the Historic Columbia River Highway in Oregon. The original cross beams from the HCB were substantially deficient in shear strength, particularly for the projected increase in traffic loads. One control beam and three beams with varying configurations of a carbon fiber reinforced polymer (CFRP) and a glass fiber reinforced polymer (GFRP) were constructed to simulate the retrofit of the actual HCB cross beams. CFRP unidirectional fabrics were applied to increase flexural capacity and GFRP unidirectional sheets to mitigate shear failure. Thirdpoint bending tests were conducted, and load, deflection and strain data were collected. Fiber optic sensors and conventional resistive gauges were placed to provide an overall behavioral understanding of the unstrengthened and strengthened beams. Results revealed that the FRP composite strengthening provided static (total applied third-point load) capacity increases of 45% for the addition of either CFRP or GFRP when compared to the unstrengthened beam. The addition of both CFRP and GFRP increased the moment capacity by 100%. Post cracking stiffness of all beams was increased primarily due to the flexural CFRP. Results suggest that the experimental beams retrofit with both the designed GFRP and CFRP should well exceed the bridge design load of 530 ft-kips, sustaining up to 640 ft-kips applied moment. The addition of GFRP alone for shear was sufficient to offset the lack of steel stirrups and allow for a conventional reinforced concrete beam failure by yielding of the tension steel followed by crushing of the concrete. The resulting ultimate deflections of the shear GFRP reinforced beam were nearly twice those of the pre-existing shear deficient beam. Experimental beams retrofit with only the designed CFRP still failed as a result of diagonal tension cracks, albeit at a 45% greater load than for the unstrengthened beam. The experimental beam retrofit with only the designed shear GFRP failed in flexure at the midspan at a 45% higher load than the control specimen, with the failure mechanism in this case being yielding of the tension steel followed by concrete crushing. A design method for flexure and shear was proposed before the onset of this experimental study. The design procedure for flexure was refined and allows for predicting the response of the beam at any applied moment. The flexural design procedure includes provisions for non-crushing failure modes, and was shown to be slightly conservative using the design material properties