PSI - Issue 84

Ivan Beltracchi et al. / Procedia Structural Integrity 84 (2026) 1015–1022

1016

1. Introduction The development of the national transport system requires not only new infrastructure, such as bridges and tunnels, but also sustained investment in maintaining the existing network, which is far larger than the portion built in recent years. A key challenge is the management of ageing structures, typically reinforced concrete (RC) and prestressed reinforced concrete (PRC) bridges, many of which are more than 50 years old and were designed for traffic demands significantly lower than those experienced today. In addition, a substantial share of the network was not designed to meet modern seismic requirements, which are now standards for new construction. In this context, structural health monitoring (SHM) systems are increasingly being implemented to improve understanding of service behaviour and to support the definition of risk indices, which can guide the prioritization of temporary safety measures, detailed assessments, and rehabilitation interventions. Statistical analyses of bridge collapses worldwide indicate that vehicular traffic, particularly impacts and overloads, is a major hazard to bridge safety, second only to hydraulic risk (Proske, 2018). Italy has more than 120,000 bridges, unevenly distributed across the country; over 18,000 of them are located in the highly industrialized region of Lombardy (Santarsiero et al., 2021). The size of this asset base, combined with its age, increases vulnerability, especially where structures are exposed to high traffic volumes and heavy vehicle loads. A notable Italian bridge collapse associated with heavy loading is the failure of the Annone overpass in 2016 (Province of Lecco), located at km 42+000 along State Road SS36. The collapse occurred while an eight-axle truck with a gross weight of 107.6 tons crossed the structure at low speed. In this case, a single exceptionally heavy vehicle was sufficient to trigger failure in a bridge already affected by advanced material deterioration. These issues were investigated within a PRIN research project coordinated by the University of Brescia in collaboration with the University of Parma, focusing on the behaviour of PRC bridges subjected to repeated heavy loads under high traffic volumes. Monthly reports from the SHM system installed on a case-study bridge along the southern expressway of Brescia show consistently high traffic, often exceeding 800˙000 vehicles per month ( acquired along 2 of the total 4 lanes). A significant percentage , approximately 80˙000 vehicles, consists of heavy trucks with gross weights g reater than 44 tons. Given its location on one of the busiest roads in the province, this bridge was selected as a representative case study for evaluating SHM-based monitoring strategies and developing predictive models to assess its safety and performance over time. The maintenance of existing bridges begins with the collection of detailed information on their current condition, enabling defects to be identified and classified in accordance with Italian Guidelines (MIT, 2022). Two main investigative approaches are commonly adopted. The first is structural health monitoring (SHM), which provides continuous assessment of structural integrity through sensor networks, signal processing, and statistical analysis. The second is system identification (SID), typically performed as a one- off campaign aimed at estimating a structure’s actual physical and mechanical properties. The range of SHM and SID techniques is broad, spanning dynamic and static methods, passive observation and active excitation, and both remote and on-site investigations. Common tools include acoustic emission systems, fiber-optic sensing networks, piezoelectric sensors, accelerometers, guided-wave techniques, tomography, and Light Detection and Ranging Technology (LiDAR). SHM often relies on non-destructive testing techniques, such as visual inspection, galvanostatic pulse testing, half cell potential measurements, ultrasonic pulse velocity, and acoustic emission, that are effective for identifying localized deterioration but provide limited information on the overall structural capacity. Vibration-based monitoring, by contrast, can support damage detection and residual-life assessment by tracking changes in modal properties, particularly natural frequencies, which may decrease as stiffness is reduced (e.g., due to corrosion-related degradation). Other modal parameters, such as damping ratios and mode shapes, are generally less sensitive; they tend to become informative mainly in the presence of localized damage and typically require dense sensor layouts as well as dedicated tools for comparison (e.g., the Modal Assurance Criterion). Beyond direct response measurements, data from dynamic weighing platforms can quantify actual traffic loads, supporting the estimation of lifetime load effects through analytical models and finite element models (FEM). These can simulate both short- and long-term structural behaviour, incorporating appropriate material degradation laws. Accordingly, given the large volumes of data generated by SHM systems, artificial intelligence methods are increasingly used to support data processing, interpretation, and prediction of structural performance.