PSI - Issue 84

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

1021

5. Synthetic performance indicators Based on the previous comparison between numerical predictions and field measurements, two key load thresholds can be identified. These thresholds can be used to screen the vehicles crossing the bridge for potential overload conditions and, if necessary, enable preventive traffic-control actions through the dynamic weighing platforms: - a first limit is associated with the end of elastic branch (approximately for 28 kN/m), corresponding to cracking of the concrete due to flexure; - a second limit is associated with the ultimate capacity (approximately for 90 kN/m), corresponding to the reaching of the ultimate compressive deformation of concrete. By determining the ratios between these thresholds and the equivalent loads extracted from the SHM data (i.e., cracking / eq and failure / eq , shown in Fig. 9), all values are greater than 1, indicating adequate structural safety under the considered conditions. The first index is directly related to durability, since cracking may expose the reinforcement to environmental actions, and should therefore be monitored over time. Notably, the lowest ratios (closest to 1) are associated with the heaviest trucks, indicating that these vehicles are the most likely to approach or exceed the cracking threshold.

(a) (b) Fig. 9 (a) Ratio between numerical prediction of load at cracking and equivalent uniform load and (b) ratio between numerical prediction of load at failure and equivalent uniform load. 6. Concluding remarks Leveraging the traffic information provided by the WIM platform, the measurements acquired by the SHM system, and the numerical models developed for the bridge, a robust and scalable methodology was established for assessing existing prestressed concrete bridges. The main outcomes of this work are summarized in the following: - The correlation between vehicles total weight and mid-span deflection shows a trend similar to linear load– response relationship, indicating that the bridge still behaves in the elastic range despite its age (50 years). Equivalent distributed load modelling by using the data algorithm with FEM analyses confirmed that even multi-axle heavy vehicles produce responses compatible with elastic structural performance, with a reduced dispersion of the data as compared to the previous one, and with more simplified parameters to check. - The advanced numerical simulations helped to better understand how the non-linear properties and the viscoelastic ones could affect the global response of the bridge. As previously shown, the contribution of asphalt layers (in addition to the others stiffening effects, like diffused supports or transversal beams) could bring an increase of the elastic stiffness by 20-40%, and the ultimate load reached by 40-50%. - Numerical analyses give the opportunity to set some safety indices related to the loss of the elastic behaviour of the structure and the ultimate load. By comparing the ratio between these two threshold values and all the equivalent extracted loads, it’s possible to see that all these values are higher than 1, despite that some of them are close to this minimum value (1.15), representing a possible issue of overloading and local damages on the artifact. Between all of these 3˙000 vehicles monitored, the val ue linked to the 5 th percentile and related