Structural health monitoring in end-of-life prediction for steel bridges subjected to fatigue cracking

This paper presents a monitoring and modelling methodology to assess the current and future conditions of steel bridges subjected to fatigue cracking. Steel bridges are subjected to fatigue cracking as a result of fluctuating stresses caused by the crossing of heavy vehicles. Specifically for orthotropic steel bridge decks, fatigue cracking is considered as one of the dominant failure mechanisms that drive maintenance interventions. These maintenance interventions, such as repairs, renovations or renewal, can have significant impact in terms of costs and traffic hindrance. Bridge owners would therefore benefit from clear indicators that enable sufficient reliable forecasts of the bridge conditions in terms of technical, functional and/or economic performance. This paper describes a pilot project in which Structural Health Monitoring is applied on an orthotropic steel deck of a bridge in the Netherlands. The focus of the pilot project was on the detection and monitoring of crack propagation in the deck plate. Firstly, the applied monitoring system and its purpose are explained. Subsequently, the paper discusses the prediction of the future condition via a three-step procedure: 1. The crack growth rate and crack size are evaluated by using a physical fracture mechanics model; 2. The monitoring data is combined with predictions of the crack growth development from the physical model of the monitored section of the bridge by using an updating procedure; 3. The end-of-life is predicted for the entire bridge (including the non-monitored part) by using a non-parametric belief net. The results of the three-step procedure demonstrate the value of combining monitoring techniques with modelling approaches by an increase of reliability in bridge performance predictions. Such an increased reliability can support bridge owners in their asset management decision making process.