Experimental capacity of <scp>TL</scp> ‐4 concrete barrier–deck connection using <scp>GFRP</scp> ‐bars with reduced‐radius 180° hooks and adhesive <scp>GFRP</scp> anchors

Abstract

Abstract This paper summarizes the experimental results performed on the TL‐4 bridge barrier, aiming to establish a cost‐effective detailing approach using glass fiber‐reinforced polymer (GFRP) bars for sustainable construction practices. The design used GFRP‐HM M13 bars as the main and secondary reinforcement for the TL‐4 barrier wall. Additionally, it featured newly developed GFRP bars with a 180° hook at the junction between the barrier and the deck. The experimental program was designed to validate the barrier's design and findings by comparing them with the factored applied moments outlined in the literature and the Canadian Highway Bridge Design Code (CHBDC) for the barrier–deck junction under simulated vehicle impact conditions. The program involved testing four full‐scale TL‐4 barrier specimens, each measuring 900 mm in length, to failure to assess their ultimate load‐carrying capacities and failure modes at the deck‐barrier joint. Two specimens represented barriers constructed for a slab‐on‐girder bridge system, while the other two were designed to rest on a non‐deformable deck slab, applicable to both new constructions and the replacement of deteriorated barriers in existing bridges. A correlation analysis was performed between the experimental results and the factored applied moments derived from CHBDC, corresponding to equivalent vehicle impact forces obtained through finite‐element modeling of the barrier–deck system. Recommendations for implementing the proposed design of the TL‐4 bridge barrier in highway bridges are provided. Pullout tests were conducted on #4 and #5 GFRP bars with reduced‐radius 180° hooks embedded in concrete blocks to determine the hook strength. Comparing the results with those obtained from the barrier testing showed that concrete breakout or diagonal tension cracking at the barrier–deck connection occurs before hook breakage.