Evaluation of CFRP-Concrete Bond Behaviour in Field Conditions: A Champlain Bridge Case Study

Abstract

In 2019, the last Canadian infrastructure report card rated the state of the nation’s infrastructure to be at risk. One such structure was the Champlain bridge in Montreal, which exhibited signs of significant damage to structural components including the girders, deck, diaphragms, and foundation elements. Before it was closed in 2019, the Champlain Bridge was one of Canada's busiest bridges. Following a series of extensive but ultimately unsuccessful rehabilitative measures, the historic Champlain Bridge was eventually replaced at a cost to taxpayers of over $4 billion. Carbon fibre-reinforced polymer (CFRP) materials were extensively used to strengthen deteriorated components of the Champlain Bridge. Even though CFRP are commonly used to strengthen bridges globally, comprehensive studies on their long-term bond performance at service conditions are lacking from the literature. Realistic testing techniques and correlating laboratory and field data remains a significant challenge. In this research, a comprehensive study has been conducted to evaluate three CFRP-strengthened bridge diaphragms extracted from the recently deconstructed Champlain Bridge in Montreal that have been exposed to aggressive environmental conditions for several years. The main objectives are: 1) to develop a comprehensive procedure to assess the long-term performance of CFRP-concrete interfaces of field structures using non-destructive and destructive testing techniques, and establish a large database to correlate non-destructive test (NDT) findings with bond strength data; 2) to assess the integrity of the CFRP-concrete interface using advanced sensing and destructive testing techniques, including the development of new test configurations to evaluate the CFRP-concrete interface under distinct loading directions; 3) to predict the bond behaviour and failure mechanisms through finite element numerical modelling of intermediate crack-induced debonding behaviour under various loading conditions; and 4) develop recommendations to ensure bridge safety and improve future maintenance and repair protocols.

The first phase of work included a detailed visual inspection combined with a series of advanced NDT methods. Several issues were identified including material incompatibility, discolouration due to corrosion, and inter-fibre cracks as well as fundamental problems arising from construction of the bridge diaphragms and installation of the CFRP. The results of the NDT field testing techniques generally point to strong bond between the CFRP and concrete substrate with defects mainly concentrated around the anchorage strips, where four layers of CFRP strips were used in the manual wet lay-up installation.

In phase two, five different destructive/semi-destructive test methods were performed to investigate the effects of load orientation and test setup on bond behaviour and the degree to which bond may be compromised by field defects. Single mode loading protocols comprised of the direct-tension pull-off test on 490 samples, two single shear-lap test setups (newly designed and conventional), and a double shear-lap configuration. Mixed mode loading was achieved using flexural beam tests. The direct tension pull-off confirmed NDT results, showing that approximately 96% of the tested CFRP pull-off samples exhibited failure in the concrete substrate, indicating good bond durability. The debonding strains from the single mode and mixed mode tests were significantly lower than code predictions.

The efficiency of the novel designed test setups proposed and utilized in this thesis were calibrated through separate pilot tests. A newly designed and pilot tried variable angle peel test included in the appendix of this document is proposed to investigate the interfacial peel strength of the CFRP-concrete interface at 15 to 60 degrees peel angles (Fowai et al., 2022). During the various experimental stages, damage progression was monitored and characterized by incorporating advanced optical techniques such as 2-dimensional and 3-dimentional digital image correlation (DIC) and distributed fibre-optic sensors (DFOS), as well as strain gauges. Finally, experimental data obtained were used as input parameters for the development and calibration of a detailed finite element (FE) model. The FEM modelling approach was validated against existing literature and was used to verify the experimental results from phase two and analyze how key parameters affect the performance and integrity of the CFRP-concrete bond of the structural elements.