PSI - Issue 78

Andrea Pozè Falet et al. / Procedia Structural Integrity 78 (2026) 325–332

326

1. Introduction Due to the ageing civil engineering heritage and the rising frequency of extreme weather events like floods, hydraulic and hydrogeological phenomena stand out as the primary cause of partial or total collapse for in-river bridges, with a very large margin over the other collapse mechanisms - more than 16 times the number of events related to material issues, the second most frequent cause (D’Angelo et al., 2025) . Notably, on its own, foundation scour is considered responsible for 50-60% of bridge collapses in the United States (Arneson et al., 2012). In fact, the deterioration caused by scour alters the constraint conditions of piers and abutments, leading to a progressive reduction in local and global structural stiffness, particularly with respect to horizontal loads (Ciancimino et al., 2022). Within this context, this brief contribution focuses on the Strambino Bridge, a road bridge near Turin, which experienced a significant scour event at Pier P2 in 2003 (see Figure 1), promptly consolidated and retrofitted afterwards. To assess the bridge's structural health status and the effectiveness of the interventions, three campaigns of Ambient Vibration Tests (AVTs) were analysed, focusing on output-only dynamic measurements from the deck. The campaigns covered three conditions pre-intervention (2003), post-intervention (2004), and recently in 2024, aimed at identifying the bridge dynamic behaviour twenty years after retrofitting. In addition, as visual inspections highlighted the possible presence of scour at Pier P3 (Figure 1d), dynamic data were analysed to investigate variations in behaviour related to scouring. The 2003 and 2004 campaigns were both initially studied using ARMAV (Autoregressive Moving-Average Model Vector) in (Foti & Sabia, 2011) and are here re-analysed with SSI-DATA (Stochastic Subspace Identification Data Driven). After identification, the modal parameters (natural frequencies, damping ratios, and mode shapes) have been investigated, using them as damage-sensitive features (DSFs) both spatially between spans and temporally between subsequent campaigns. Thus, this contribution presents some preliminary findings based on these comparisons. The remainder of this document is structured as follows. In Section 2, the field case study is briefly introduced. The most recent experimental campaign is described in Section 3, highlighting similarities and differences with the procedures followed 20 years ago. This ensures comparability and allows for the application of more recent advances in System Identification. Furthermore, the outcomes of this 3-day-long campaign are discussed, interpreted, and compared to the benchmark values of 2003 and 2004 in the same Section. Finally. Section 4 concludes this short discussion, especially highlighting current and future research on these three datasets. 2. The Strambino Bridge The case study (Figure 1) is located in a rural area of Piedmont, northwest Italy, in the Metropolitan City of Turin area. Specifically, it allows the SP56 road to cross the Dora Baltea river between the small towns of Strambino (to the West) and Caravino and Vestignè (to the East). It is a five-span, simply supported bridge, with each span composed of a 30-metre-long prestressed, reinforced concrete girder deck. The pier numbering begins from the left riverbank when looking downstream, thus, from P1 on the Caravino/Vestignè side to P4 close to the right bank (Strambino side). The spans are numbered following the same order: Span C1 is between Abutment A1 (East) and P1, C2 between P1 and the now-retrofitted P2, and so on, till Span C5 between P4 and Abutment A2 (Strambino side). Figures 1.b and 1.c show P2 before and after the retrofitting, respectively; more details can be found in (Foti & Sabia, 2011).

(a)