PSI - Issue 78

Andrea Dhima et al. / Procedia Structural Integrity 78 (2026) 1366–1373

1368

As no information on the structure under investigation was available, a simulated design analysis was carried out by combining data from on-site inspections with typical material properties and design criteria used at the time of construction. The numerical model of the viaduct (Figure 2) was developed using the Finite Element (FE) analysis software Midas Civil (Midas, 2024). Piers, main girders, and crossbeams were modelled using general beam elements, the deck was instead modelled using four-node plate elements, with a nominal size of 0.5 m × 0.5 m. Subsequently, slabs and girders were connected via rigid links. At the substructure, a characteristic cylindrical compressive strength of 35 MPa was assumed for the concrete together with a characteristic yield strength of 355 MPa for the non prestressed rebars. Regarding the girders, a characteristic cylindrical compressive strength of 40 MPa was adopted for the concrete together with a characteristic tensile strength of 1860 MPa and a characteristic proof stress at 0.1 % offset of 1670 MPa for the prestressing steel. According to the simulated design, the double- T girders contain fifty 0.6’ inches diameter prestressing strands, whereas the rectangular-section girders comprise thirty- five 0.6’ inches diameter strands. A linear-elastic approach was used for the model, with a Young's modulus, E , equal to 34625 MPa for the substructure and 35547 MPa for the superstructure. The self-weight of the structural elements was calculated automatically by the software, whereas the permanent non-structural loads were imposed as distributed loads. The traffic loads were applied according to the configurations specified in the Italian standard (NTC 2018).

Fig. 2. Axonometric view of the numerical model of the viaduct.

The assessment of the safety level, as proposed by the guidelines (ANFISA, 2022; MIT, 2020), involves the verification of the structure against both traffic loads and seismic action. Regarding traffic loads, the standard defines three performance levels for the verification: adequate, operative and transitable conditions. For a detailed description of these performance levels, the interested reader is referred to (Gara et al., 2025). The results of the safety assessment under traffic loads, based on the comparison between the bending moment demand M, Ed and the corresponding resistance M, Rd (as reported in Table 1), indicate that the 42-metre span beams fail to meet the requirements for any of the three specified performance levels. However, it is noted that, for the transitable conditions, the safety index ζv is close to or greater than unit y.

Table 1. Beams assessment under traffic load.

Beam-1 (T1)

ζv

Beam-2 (T2)

ζv

Beam-3 (T3)

ζv

LEVEL 4

M, Ed [kNm] M, Rd [kNm] M, Ed [kNm] M, Rd [kNm] M, Ed [kNm] M, Rd [kNm]

20901.41

19259.59

19162.80

0.75

0.81

0,81

Adequate

15582

15582

15582

19159.51

17703.71

17614.10

0.85

0.92

0.93

Operative

16364

16364

16364 11874 16364

16964.15

12214.2

0.96

1.34

1.38

Transitable

16364

16364

The seismic assessment, according to the guidelines, was conducted using a linear dynamic analysis, based on site specific seismic hazard parameters and assuming a behaviour factor q equal to 1.5. The results of the linear dynamic analysis indicate that the deck beams meet the verification requirements for both bending and shear. In contrast, the assessment of the substructure shows that whilst the piers meet the bending capacity requirements, their shear capacity is found to be insufficient. This result is consistent with the typical seismic behaviour of bridges, in which the piers