PSI - Issue 62

4

Mattia Zizi et al. / Procedia Structural Integrity 62 (2024) 430–437 M. Zizi et al. / Structural Integrity Procedia 00 (2019) 000 – 000

433

between 30 and 50% were forced between two consecutive frames. Drone velocity was kept constant and the frames were acquired in an automatic way by imposing a shooting interval of 0.5 frame per second. Given the hard environmental conditions due to the presence of vegetation, the aerial acquisitions were aimed at surveying the geometrical characteristics of the central span, only. The other geometric characteristics for obtaining a reliable geometrical model were acquired by means of a direct survey (i.e. by using laser meter Leica DISTO X3). The manual measurements confirmed a certain degree of repetitiveness of the geometrical characteristics of the structure. Thanks to this survey program, it was possible to obtain a reliable geometrical model, which was adopted for the analyses described in the following. 3. Dynamical characterization 3.1. Experimental setup A series of Ambient Vibration tests were performed on the bridge. Four tri-axial and two bi-axial (in the horizontal x-y plane) force-balanced accelerometers with 2.5 V/g sensitivity and bandwidth 0-200 Hz were placed on the bridge roadway in multiple setups (Fig. 3). Two sensors (T1 and T2) on both sides in the middle of the central span were kept in the same position during all tests, while sensors T3-4 and B5-6 were moved in order to investigate the modal properties of the bridge throughout its whole length. Due to technical problems, setup IV failed and no data were recorded. Acceleration data were acquired at 100 Hz sampling rate, and each setup lasted 20 minutes.

Fig. 3. AVT setups: accelerometers are labelled as T (triaxial) or B (biaxial); the Roman letter indicates the setup number.

3.2. Results

In Fig. 4 acceleration measurements taken in setup I are displayed after baseline correction. Peaks are well distributed in time, which is an indication that considerable traffic passed on the bridge on the day of the acquisition. Moreover, the general level of accelerations is significantly high, with peaks in sensors T1 and T2 (placed in the longest span) generally in the range of 1- 2‧10 -3 g. Dynamic properties (frequencies and mode shapes) were obtained by means of Frequency Domain Decomposition (FDD) (Brincker et al., 2001) applying the procedure proposed by Amador and Brincker (2021) to merge multiple setups having reference and roving sensors. The first four frequencies are clearly identifiable as peaks of the first Singular Value plot Fig. 5 at 1.66 Hz, 2.54 Hz, 3.52 Hz and 5.47 Hz. The corresponding mode shapes are displayed in Fig. 6 as interpolation of the modal displacements at the sensor positions over a simplified 3D model of the bridge for a clearer visualization. It is possible to verify that the first three modes are bending modes in the plane of the roadway, while the fourth mode is a vertical mode.