Designing monitoring systems for long-term structural response

Designing structural health monitoring (SHM) is logically equivalent to designing a civil structure. The capacity must be greater than the demand to achieve the required performance. Monitoring capacity and demand are the counterparts of structural capacity and demand in the semi-probabilistic structural design. They are defined as the uncertainty of key-parameters that represent the structure behaviour and will be estimated through the monitoring system: the capacity is the uncertainty resulting from the estimation, the demand is the design target. As far as concrete and prestressed concrete bridges are concerned, important key-parameters are long-term temperature-compensated responses, such as strain trend, displacement trend, and rotation trend. Their estimation as well as the estimation of their uncertainty can be easily performed a posteriori through Bayesian inference, once monitoring data are available. However, in the design phase measurements are not yet available. We propose an approach for designing a structural health monitoring system accounting for temperature compensation, which allow to quantify the uncertainty of structural response trends a pre-posteriori, before monitoring data are available. We analyse the impact of sensors' accuracy, monitoring duration, and seasonal temperature variation on the expected uncertainty. Finally, we test our framework on a real-life case study, the Colle Isarco viaduct, one of the longest prestressed concrete highway bridges in the European Alpine region.