Behaviour of Axially Loaded Short Columns With Damaged FRP Jackets and Their Repair Using FRP Patches

Thumbnail Image
Hutcheson, Zachary Peter
CFFT , Concrete-filled , FRP tube , Adhesive , Mechanical fastener , Eccentric loading , Condition assessment , Residual strength
Extensive research has been conducted on concrete confined by fibre-reinforced polymer (FRP) jackets. However, there is a gap in the research concerning the behavior of columns with damaged FRP jackets subjected to eccentric loading and the effectiveness of different repair methods in restoring lost strength. The first study varied bond length of FRP repair patches, number of layers, number of fasteners, and compared wet layup to prefabricated shells. It was shown that with a vertical cut of a length equal to 0.6 the diameter, axial strength (P) dropped to 0.57 the original strength (Po). As the bond length of the Glass-FRP (GFRP) repair patch increased up to one half of the perimeter from either side of the cut, P/Po increased from 0.57 to 0.68. When the ends of the GFRP patch included a 100 mm overlap, a full-strength recovery was achieved. The prefabricated GFRP shell patch (cut from the same tube) showed the highest strength with P/Po reaching 0.75, 9% and 15% higher than Carbon-FRP and GFRP wraps, respectively. Adding mechanical fasteners enhanced strength recovery. For a GFRP patch with 150 mm bond length on either side of the cut, increasing the number of fasteners from zero to 15 on either side increased (P/Po) from 0.65 to 0.84 (i.e., by up to 29%). The second study aims to develop a simple empirical tool based on experimental results to aid in condition assessment and establish the reduced axial strength ratio (P/Po) of short columns loaded at small eccentricities. It was shown that jacket cuts affect eccentrically loaded columns more severely than concentrically loaded ones. As (e) increased from zero to 0.1D, (P/Po) reduced from 1.0 to 0.81 for intact samples and from 0.56 to 0.4 for samples with jacket cut. As cut location travelled around the circumference starting from extreme compression (θ=0o), less strength reduction was observed where (P/Po) increased, and at (θ=180o) it even exceeded (P/Po) at zero eccentricity. This suggests a reduction of eccentricity due to a shift of cross-sectional centroid due to the disturbed confinement on one side. An empirical model accounting for (e/D), (θ), (x/D) and inclined cuts, has been developed and calibrated for short columns of up to 22 slenderness ratio. A simple design model is proposed and solved for both chapters of this thesis.
External DOI