Structural condition monitoring of a steel-concrete composite bridge based on statistical pattern recognition of dynamic measurements

Selected results collected in the continuous dynamic monitoring of a steel-concrete composite bridge are reported in the paper. The investigated bridge belongs to Autostrada Pedemontana Lombarda (APL), a motorway aimed at optimizing the road connection between Milan, Monza, Como and Varese (Lombardy, Northern Italy). After a concise description of the bridge and its monitoring system, details are given on the software tools developed to process the continuously acquired data within a statistical pattern recognition framework. Subsequently, the focus is moved on the data collected during the first year of monitoring, with particular attention to the influence of environmental parameters on the variations of natural frequencies and mode shapes. In the last part of the paper, the application of different novelty analysis procedures is exemplified: (a) the well-known control charts based on Mahalanobis distance and (b) the less diffused Cointegration technique, not requiring any removal of environmental/operational effects and involving the definition of a linear combination of non-stationary monitored features. It is further noticed that the strategy adopted for the health assessment of the bridge will be extended to various bridges managed by APL, as part of a project funded by the Italian Government.