PSI - Issue 64

Vera Rillo et al. / Procedia Structural Integrity 64 (2024) 700–707 Rillo et al./ Structural Integrity Procedia 00 (2019) 000 – 000

703

4

On the uphill side of the roadway deck, at the level of the lower flange of the main beams, there is a walkway which widens in the central zone, increasing from a 2.15m projection measured from the outer edge of the flanges to 3.20m. The roadway deck, constructed with in-situ reinforced concrete C32/40, features an overall width of 9.50m, divided into two driving lanes of 3.50m each one and two lateral curbs of 1.00m width. The thickness of the lateral curbs is 33cm while it varies between 25 and 34cm for the driving lanes. The entire deck, for both vehicular and pedestrian traffic, is seismically isolated from the underlying structures by employing elastomeric seismic isolators arranged centrifugally with respect to the viaduct axis. Consequently, two seismic isolators are located at the outer supports, while two multidirectional devices are installed at the inner supports. 3.2. Preliminary numerical finite element model In order to identify the dynamic behavior of the bridge under examination, a preliminary numerical finite element model of the deck only was developed with the commercial software SAP2000 (Computers and Structures, 2016). The various components of the structure were modeled considering their real geometry with mono-dimensional (frame) and bi-dimensional (shell) finite elements. In particular, the steel beams were modeled with frames while the deck was modeled by thin "shell" elements, with 4 nodes and a formulation that combines membrane and plate bending behavior. A modal analysis was carried out to identify the main modes of vibration in terms of frequencies and modal shapes. The subsequent analysis focuses on the first six identified modes whose frequencies are presented in Table 1.

Table 1. Modal Analysis Results

ID MODE f [Hz] 1 1.785 2 2.619 3 3.098 4 3.409 5 4.750 6 4.784

The first mode exhibits translational motion along the z-axis and primarily affects the central span. The second mode is torsional followed by a third mode that is characterized by its predominant influence on a lateral span, with spans alternating in motion (if one moves upward, the next moves downward). The fourth mode involves all spans oscillating along the z-axis on the same side while the fifth mode also involves translational motion along the z-axis. Finally the sixth mode examined is torsional. The modal deformations in 3D are shown in Fig. 2. Then the developed model was used to numerically simulate the ambient vibration test. Specifically, the model was excited with white noise, produced as a sequence of normally distributed random variables using the software Matlab (2022), simulating the environmental actions. The accelerometers that will be arranged on the bridge deck for the dynamic test were numerically simulated as joints in the same locations. Linear time history analyses were performed reading the histories of accelerations at the joints were the virtual accelerometers are located. The results are exported to a text file for further processing in the software ARTeMIS (SVS, 2019). The first sensors configuration assumes the position of nodes at the end of the beams, skipping one beam at a time for a total of 24 nodes, i.e., 24 virtual mono-axial accelerometers. It is worth to underline that, since the modes predominantly affect the z-component, the derived sensor configuration refers only to z-component.