PSI - Issue 84

F. Foria et al. / Procedia Structural Integrity 84 (2026) 645–652

650

The MAC shows very good correspondence between both the frequencies and the exact, identified modal shapes. 4. Study case This section deals with the application of the algorithm described in section 3 to the case study of the steel-concrete composite bridge located in central Italy. The same case study was discussed in [7], where the results obtained using the primitive version of the dynamic identification algorithm were shown. The geometrical and mechanical characteristics have been deducted from the original design documents and the on-site survey results. The bridge consists of 9 arches (7 central full-core arches and 2 lateral lattice arches) hinged at the ends with variable spacing between 1.75 m (between the central full-core arches) and 1.5 m (between full-core arches and lattice arches). Each arch has a span of 25.47 m, a rise of 1.68 m, and a length of 25.76 m. The central arches and the two outer arches are reticular structures connected to each other by lattice diaphragms and braces placed at the level of the lower flanges of the main elements. The overall transversal width is 14.5 m. The supports, which create a hinge system, are made of cast steel. The deck is constructed with a slab poured on Zores Steel and is supported by a series of struts that connect it to the arches. The masonry abutments are inclined at about 16° relative to the roadway axis, causing the 9 arches to be staggered. The bridge currently carries a three-lane municipal road.

Figure 2 3D Point cloud laser scanner survey

An accurate finite element model was built based on geometric information obtained through laser scanner surveying. The mechanical characterisation of the various structural elements was carried out on the basis of information acquired from the in-situ survey campaign conducted on the structure. Modal analysis of the model discussed in [7] provided a first frequency of 2.36 Hz, corresponding to a transverse vibration mode relative to the bridge axis, with a mass participation of 82%. The first global vertical mode is at 5.14 Hz with a mass participation of 63%. The bridge, being dated and in need of repair, is constantly monitored for vibrations caused by environmental actions and vehicle loads. All the necessary information to understand the current conditions of the bridge are processed in real time to be aware of its structural behavior under static and dynamic actions, in order to ensure an adequate level of safety for public safety. Several sets of recordings acquired at different times and under different load conditions were analysed, following the algorithmic steps described in section 3. Preliminary steps 1, 2, and 3 were performed. Figure 3 shows the stabilization diagrams of frequency varying the model order. The algorithm automatically returns the alignments of stable poles and the multiplicity of frequencies identified (Step 4 and 5).

(a) (c) Figure 3 Stabilization diagram (number of poles versus frequency) for various model order values and fixed time lag equal to 150: (a) model order 30, (b) model order 100, and (c) model order 200 (b)