2019 CSCE Annual Conference - Laval (Greater Montreal) Conference
Dr. Douglas Tomlinson, University of Alberta
Bridge decks and barriers reinforced with Glass Fibre Reinforced Polymer (GFRP) bars are becoming more popular in bridge construction as their corrosion resisting properties can extend the lives of bridges significantly. GFRP bars have comparable price and structural performance compared to other non-corrosive materials used in bridge construction. Although several procedures are in place for repairing steel-reinforced barriers after they are damaged by a significant vehicle impact there is limited information available regarding the repair of damaged GFRP-reinforced barriers. For this reason, GFRP-reinforced barriers are not currently being specified by some transportation ministries in Canada.
The performance of two proposed repair techniques for GFRP-reinforced single and double-sloped PL-4 type barriers was investigated using the 2D finite element program VecTor2. The model incorporates barriers and deck overhangs. For both techniques, the damaged barrier is removed and replaced with a new GFRP-reinforced barrier. In the first technique, the barrier is anchored to the existing bridge deck using single headed GFRP bars while in the second technique, the barrier is anchored using Near Surface Mounted (NSM) GFRP bars. The spacing, diameter, and embedment depths of the anchorage systems are the main parameters investigated in this study. Embedment is modelled using bond-slip data from previously completed experiments on NSM and headed GFRP bars. For each technique, the influence of deck GFRP reinforcement spacing (varied between 100 and 300 mm) and deck overhang length (0 to 4 m) was also investigated. The model results are calibrated using models and tests on undamaged GFRP-reinforced bridge barriers.
The results show that, for small overhangs, failure occurs at the bridge deck/barrier interface. For small embedment depths, failure occurs by pullout of the headed or NSM bars while for longer embedment depths, failure occurs by concrete crushing. Sections that failed by concrete crushing performed equivalently to undamaged barriers, indicating that the repair was effective in those cases. The NSM bars generally performed better than headed bars with the same total anchorage area but are more difficult to implement in construction due to the labour required. As overhang length increased and deck reinforcement ratio decreased, failure was more likely to occur at the support of the overhang. There was insignificant change when comparing the single and double-sloped barriers. The results of this investigation are being compared to an experimental program that will test repaired barriers under static loads. Future work will incorporate two-way and impact loading behaviour.