PSI - Issue 62

Valentina Giglioni et al. / Procedia Structural Integrity 62 (2024) 887–894 Giglioni et al. / Structural Integrity Procedia 00 (2019) 000–000

891

5

seizing of the bearings is carried out in a certain number of piers by locking longitudinal and/or rotation movements. Since the main goal of this paper is to transfer damage labels across spans, a restricted portion of simulated scenarios is here investigated. The corresponding damage classes are illustrated in Fig. 3b, where M1, M2 and M3, M4 indicate the stiffness reduction by applying a 64 g weight mass on the side and centre line, in Span 2 and Span 1, respectively. Examples of M3 and M1 damages are depicted in Fig. 3a and Fig. 3c. Given that each damage is independently applied, the corresponding data are not affected by the previous simulations.

Fig. 3. (a) example of M3 damage; (b) illustration of the simulated scenario; (c) example of M1 damage.

3.1. System identification and features extraction The acquired accelerations are fed into MOSS, a software presented in García-Macías et al. (García-Macías, 2020), to perform automated system identification, via the covariance-based Stochastic Subspace Identification (SSI) technique, to continuously extract modal parameters for the bridge configuration of interest. The identified mode shapes from a single healthy acquisition and the corresponding natural frequencies, considered as references for frequency tracking, are shown in Fig. 4, which describes a first symmetric bending mode (Mode 1), a torsional mode mainly involving the main span (Mode 2), a second non-symmetric bending mode (Mode 3) and a third symmetric bending mode, especially affecting the lateral spans (Mode 4).

Fig. 4. Identified mode shapes for B4 bridge configuration.