Abstract

According to Transportation Association of Canada (TAC), the rough estimate of number of bridges in Canada is 80,000 with the replacement value of $35 billion. A large number of bridges will need replacement during 2005 to 2015 which will result in 50% annual increase in replacement cost. Recent alarming incidents of the Laval De la Concorde Overpass collapse (2006), Canada and the I-35W Mississippi River bridge collapse (2007), USA show the gravity of the situation. One of the main factors responsible for this situation is the present available techniques of the bridge condition monitoring and rehabilitation are not able to cope up with the drastic deterioration and ageing of the bridges. The widely employed method for bridge inspection is visual inspection, and it lacks the reliability-based assessment of bridge and its components. The instrumentation of the bridge with Structural Health Monitoring (SHM) systems and assessment of the bridge condition and behaviour based on the information obtained from SHM systems is one of the promising solutions of the present problem. The main focus of the current research is to integrate SHM data with traditional information (e.g. visual inspection), develop a reliability based structural condition index using the updated information on a structures operational performance, and assessing the value of information for SHM in regard to the overall lifecycle cost of a structure. This study develops a methodology for a reliability based assessment of the bridge components using SHM system information, and information updating by fusing SHM data with traditional information for precise evaluation of expected life cycle cost. The methods developed herein have been demonstrated through a case study based on an existing bridge namely, the Crowchild Bridge in Calgary, Alberta. A finite element model of the bridge has been developed and validated against the field data. This validated model has been used to simulate the static load test on the bridge, deterioration in the bridge and to study the bridge response under the different loading conditions. The artificial neural network (ANN) technique has been used for the diagnosis of the SHM data, and then the reliability index of the bridge deck has been calculated using the Monte Carlo Simulation technique. A method for updating the bridge deck repair strategy is introduced based on the reliability index calculation. The maintenance and rehabilitation strategy is updated based on the hypothetical results. The results of updated strategy are compared with un-updated one using the Bayesian Theorem. The expected life cycle cost is evaluated considering the capital cost, maintenance and rehabilitation cost, user cost, and failure cost. Capital cost is treated as deterministic while maintenance and rehabilitation cost, and user cost are considered probabilistic. Each individual cost and then total cost is calculated per m 2 . The value of information is also discussed.