Testing A Full-Scale Reinforced Concrete Bridge Deck with GFRP and Steel Reinforcement Using Cyclic Pulsating and Rolling Loads

Abstract

The repair and replacement of bridge decks represents a large portion of infrastructure spending. Gaining a better understanding of bridge deck behaviour through more representative vehicular loading can lead to a reduction in the costs, time and environmental impacts associated with construction. Past studies have most often used fixed pulsating loads to represent cyclic loading on a bridge deck. However, it has been shown that these types of loading do not induce the most critical conditions when compared with moving loads. This thesis is the first part of a long-term study intended to compare the loading effects of fixed pulsating cyclic loads against moving wheel loads, up to 3000 load cycles. The study also compares different types of deck reinforcement and construction methods, namely conventional steel reinforcing bar, glass fibre reinforced polymer (GFRP) rebar and GFRP stay-in-place (SIP) structural forms for rapid construction. A slab-on-girder reinforced concrete bridge deck with dimensions of 15.24 m x 3.89 m x 210 mm and a girder spacing of 3.05 m was designed according to the Canadian Highway Bridge Design Code (CHBDC) and constructed in the laboratory. Detailing and construction procedures have also been proposed for the new GFRP SIP structural form system. The deck was conceptually divided into four sections but all were monolithically cast. Sections 1 and 4, at either ends of the deck, were identical and reinforced with GFRP bars, Section 2 incorporated the novel GFRP SIP form system and Section 3 was reinforced with conventional steel rebar. Section 1 was subjected to pulsating loads and Sections 2-4 were subjected to moving wheel loads. Monotonic load tests were performed at various cycling intervals to establish stiffness degradation. After 3000 load cycles, the maximum values of reinforcement and concrete strain were well within their ultimate limits, signifying the deck was not at risk of failure. Section 4 experienced a larger reduction in stiffness than Section 1, suggesting that the moving loads were more damaging than the pulsating loads. It is important to note that observations could change at higher levels of cyclic loading, which is beyond the scope of this thesis.