PSI - Issue 84

Laura Dieci et al. / Procedia Structural Integrity 84 (2026) 591–598

592

1. Introduction Civil infrastructure monitoring is essential for maintaining safety, functionality, and long-term performance. Among the available assessment tools, dynamic identification of bridges has proven particularly effective, as it allows engineers to determine key modal parameters, such as natural frequencies, mode shapes, and damping ratios, that provide insight into the structural behavior under operational or ambient conditions (Ranieri et al. (2020); Dong et al. (2020); Magalhães et al. (2012); Romanazzi et al. (2023)). These modal parameters are sensitive indicators of structural changes, including anomalies or damage, often detectable before any visual signs appear (Comanducci et al. (2016); Ponsi et al. (2023)). Consequently, dynamic testing and long-term vibration monitoring have become an essential component of modern Structural Health Monitoring (SHM) systems (Castagnetti et al. (2016); Poluzzi et al. (2019); Zona (2020); Maes and Lombaert (2021)). Alongside experimental observations, numerical modelling is a critical tool for understanding and predicting bridge behavior. Finite Element (FE) models allow detailed analysis of stress distribution, load transfer, and both dynamic and static structural responses. The reliability of these models depends strongly on accurate representation of geometry, materials, and boundary conditions. Since discrepancies between observed and modeled behavior are inevitably present, due to the unavoidable approximations inherent in the model, the FE model must be calibrated using experimental data (Ponsi et al. (2021); Vincenzi et al. (2019)). When static measurements such as load-test deflections are available alongside modal data, integrating both types of information enhances the robustness of the FE model and increases confidence in its predictive capabilities. This paper focuses on the “Ponte delle Grazie”, a reinforced concrete Gerber-type bridge located in Faenza, Italy. Constructed in 1950, the bridge has undergone considerable deterioration over time, particularly in its bearings, Gerber saddles, and lateral girders. Based on the available design documentation and a laser scanner survey, a detailed finite element model (FE) of the bridge has been developed, representing the deck with its five main girders, transverse beams, and slab, with particular attention to the Gerber saddle. Dynamic testing carried out within the DIGI-BRIDGE research project provided experimental modal parameters (Scocciolini et al., 2025) and the first modes are used as a reference for the calibration procedure. Additionally, a static load test was performed in 2017, during which deflections at key locations, such as the bridge midspan and the Gerber saddle, were measured. Based on these experimental observations, the model is then calibrated to reproduce both its dynamic and static behavior by using an improved surrogate-assisted evolutionary algorithm, known as DE-S (Vincenzi et al. (2017)). This algorithm combines the robustness of the Differential Evolution (DE) algorithm with the computational efficiency of a second order surrogate approximation of the objective function and of an effective infill sampling strategy (S). The structure of the paper is as follows. Section 2 introduces the case study. Section 3 describes the experimental campaigns, including both dynamic and static tests. Section 4 presents the FE modelling strategy. Section 5 is about the calibration method and the comparison of numerical predictions with experimental results in both dynamic and static conditions. Finally, Section 6 draws conclusions. 2. Case study: “Ponte delle Grazie” The case study focuses on the “Ponte delle Grazie”, a reinforced concrete Gerber-type bridge located in Faenza, northern Italy, spanning the Lamone River (Fig. 1). Constructed in 1950, the bridge has a total length of 72 meters and a width of approximately 13 meters, and it is composed of three spans: the two end spans measure 21 meters each, while the central span is 30 meters long and includes a section supported by Gerber saddles. The deck consists of a slab, transverse beams, and five main girders, spaced at 2.40 meters, with a height of 2.40 meters at the pile supports and varying along the bridge length. The abutments and piers are founded on reinforced concrete posts. Recent inspections, conducted in 2017 and 2022, revealed that the actual dimensions did not match those reported in the original design drawings. Moreover, the bridge also exhibited significant deterioration, particularly at the bearings, lateral girders, and Gerber saddles. Urgent interventions were carried out to restore structural safety, including the installation of safety devices beside the existing bearings, steel tie-rods to secure the Gerber saddles, and the strengthening of the concrete bearings and lateral girders. Given its structural condition and importance, the bridge was then equipped with a permanent monitoring system as part of the DIGI-BRIDGE project.