PSI - Issue 84

Francesco Allegrezza et al. / Procedia Structural Integrity 84 (2026) 81–88

86

3.3. Case study 3: Post-tensioned concrete bridge A post- tensioned concrete highway bridge during normal traffic conditions was analysed as field test. The structure consists of three spans, each characterized by a deck consisting of a concrete slab resting on six simply supported prestressed reinforced concrete beams, with a span length of 32 m. The analysis focused on the western span, selected for its reduced distance from the ground, which allowed for easy installation of the reference measurement equipment, as illustrated in Micozzi et al. (2023, 2024, 2025). An aerial view of the bridge and the span under study is shown in Fig. 5a. The hardware for video acquisition was essentially the same described in the laboratory tests, with the only difference being the use of a 100 mm focal length C-mount lens (Kowa LM100JC1MS), more suitable for longer shooting distances. A target measuring 80 mm was positioned in the centre of the span, while a second reference target, measuring 160 mm, was placed on top of the pier cap. The latter was used to compensate for any rigid movements of the acquisition system due to vibrations transmitted from the ground to the camera tripod, as illustrated in Micozzi et al. (2025). The targets used had the same composition as described in the previous sections. A displacement transducer was also installed at midspan to measure the same vertical displacements as the target and provide a reference measurement for validating the results obtained through video analysis. A schematic representation of the experimental setup is shown in Fig. 5b.

a

b

Fig. 5. Aerial view of the monitored span (a) and schematic representation of the adopted setup (b).

The camera acquisition frequency was set to 60 FPS, while the transducer operated at a sampling frequency of 1200 Hz. During the experimental campaign, described by Micozzi et al. (2023, 2024, 2025), several ten-minute recordings were made under normal traffic conditions; the videos acquired were then analysed offline using template matching algorithms described in Section 2. The results of the analysis, displayed in Fig. 6, show that both methods are able to accurately reconstruct the vertical displacement of the span. The time histories obtained from the videos are in good agreement with the measurements provided by the displacement transducer, both during the passage of heavy vehicles and in the subsequent free vibration phase of the structure, highlighting a correct identification of the main dynamic components. Unlike what was observed in synthetic tests and laboratory tests under highly controlled conditions, the results obtained in the field do not show the presence of systematic error. This effect can be attributed to the fact that this kind of error is associated with sharp images, characterized by a clear separation of intensity levels. However, in real acquisitions there is generally a slight degree of blurring, caused by several factors such as inaccuracies in focusing or environmental noise, which introduces more gradual transitions between image regions. In this context, moderate blurring can therefore be beneficial for displacement estimation, helping to mitigate the systematic effects observed in artificially perfect conditions and making the method more robust in real-world applications.