PSI - Issue 78

Samuele Faini et al. / Procedia Structural Integrity 78 (2026) 718–725

720

Notably, severe reinforcement corrosion and spalling of concrete cover layer were observed at the girders’ bottom edges (see Fig. 1-b) and at columns (see Fig. 1-c). Vegetation within half-joints hinders proper thermal expansion and contraction of the deck, making also questionable their reliability in seismic conditions. Additionally, the mid-span thermal joint, with a small gap of about 20 mm, may cause hammering under dynamic loads. The mechanical properties of both concrete and steel rebars were evaluated through experimental tests on specimens extracted directly from the structure. The compressive strength of concrete ( , ) was determined according to EN 12390-13 [CEN, 2021], for testing procedures, and to EN 12504-1 [CEN, 2019], for core extraction: (i) , = 60.1 MPa for the piers; (ii) , = 74.6 MPa for the arch; (iii) , =43.9MPa for the girders. The elastic modulus of concrete ( ) was not measured directly on specimens but estimated using the empirical formula ( = 22000 ∙ ( /10) 0.3 ) of EC2-1 [CEN, 2004]: (i) = 37680 for the piers; (ii) = 40202 for the arch; (iii) = 34290 for the girders. Based on original drawings , the steel grade “Aq42” was used for both arch and piers ’ rebars while the “GS 4400” for girders ’ reinforcement [CSLP, 1957]. To assess the rebars’ properties, tensile tests were carried out according to EN 10002-1 [CEN, 2001] on three ∅ 8 mm stirrup-specimens extracted from low-stress areas of arch and girders. Due to the limited number of tests, and to the absence of specimens taken from columns (expected to be the most critical under seismic loads), average mechanical parameters reported in literature [Verderame et al., 2011] were considered in this study. Although the measured mass loss was only 2.2%, samples came from less exposed zones. Literature reports that spalling and strength loss can occur at 5 – 10% corrosion [Imperadore et al., 2017; Rodríguez et al., 2004]. Therefore, a 10% level was conservatively adopted. (a) (b)

(c)

Fig. 1. The Morandi’s bridge in Valvestino (Italy): (a) overview; (b) corrosion of edge-girders; (c) corrosion of piers . Note: the piers’ numbers indicate the type of cross-section with details reported in [Gandelli et al., 2025]. 3. Seismic assessment of the “as - built” bridge The seismic behaviour of the “as - built” bridge was evaluated through non-linear time history (NLTH) analyses carried out according to the Italian Building Code [CSLP, 2018]. An overview of the whole set of non-linear modelling assumptions is shown in Fig. 2. In particular , the piers’ flexural behaviour was simulated using fiber -section beam elements while the remaining structural members by means of elastic beam elements [Midas-Support]. In fiber-sections, the rebars’ hysteretic response was modelled by means the “Menegotto - Pinto” model [M enegotto and Pinto, 1973]. To account for the effects of corrosion, the yield stress ( , ), the ultimate strength ( , ), and the relevant strain ( , ) were estimated according to a degradation model [Imperadore at al., 2017]: , = , ∙ (1 − 0.0143453 ∙ [ % ] ) = 276.1 (1) , = , ∙ (1 − 0.0125301 ∙ [ % ] ) = 406.0 (2)