Moment Connections of Concrete-Filled Fibre Reinforced Polymer Tubes to Reinforced Concrete Footings

Abstract

Fiber reinforced polymers (FRPs) are increasingly being accepted in structural engineering applications. One promising system involves the use of concrete-filled FRP tubes (CFFTs) as bridge piers, columns or piles. While CFFT members have been extensively studied under various loading conditions, very little attention has been given to their connections to other structural components such as footings and beams. This study explores two different moment connections for CFFT members, using 13 medium-scale specimens and seven ancillary tests. The first connection involves embedment of the FRP tube into the concrete foundations during casting. Five-219 mm diameter (D) precast CFFTs were embedded into 500x500x500 mm concrete foundation each, at different embedment lengths ranging from 0.3D to 1.5D and tested in flexure as cantilevers with 1100 mm spans. The study showed that the optimal embedment length was 0.73D. This was essentially the minimum embedment length necessary to produce tension failure of the CFFT member outside the footing, rather than premature bond failure that would otherwise occur at lower loads. Additionally, six push-through tests were conducted on CFFT stubs embedded into footings. The average bond strength was found to be 0.75 MPa. The second connection involved adhesive bonding of hollow FRP tubes to short reinforced concrete circular stubs protruding from concrete footings. The remainder of the tube was then filled with concrete, without the need for shoring. Four-169 mm diameter FRP tubes were first adhesively bonded onto footings with heavily steel-reinforced concrete stubs varying in length from 0.5D to 2.0D, and tested as cantilevers with 1300 mm spans. The optimal bond length that would lead to flexural failure of the tube just outside the stub, rather than bond failure, was about 1.1D. Based on this, two additional specimens with 1.5D stubs having varying steel reinforcement ratio (ρ) in the stubs were tested. It was shown that the optimal ρ was 2.5%. Finally, the effect of low cycle reversed bending fatigue was studied using two additional specimens, including one with a sustained axial load of 15-19% of the CFFT axial capacity. Remarkable levels of ductility associated with the plastic hinge forming in the stub were observed.