PSI - Issue 78

Stefano Bozza et al. / Procedia Structural Integrity 78 (2026) 1213–1220

1214

1. Introduction A large number of existing bridges in Italy were built in the second half of the 20th century, mainly between 1955 and 1975 (Pinto and Franchin 2010), following outdated codes which usually required less severe design actions than current standards (Bozza et al, 2023). In recent years, after the collapse of the Morandi bridge (Calvi et al, 2019) and the issue of the Italian Guidelines of existing bridges (MIMS 2022), a great effort to evaluate existing bridge had been done, following a multilevel approach from rapid risk classification of large portfolios of bridges (Salvatore et al, 2024) to the accurate assessment of structures (Galassi Sconocchia e al, 2024). Since a large part of the Italian territory was not seismically classified until 2003 (OPCM 2003), most of existing bridges were designed without considering the effects of earthquake excitations. Therefore, several in-service bridges can be presently affected by devices (e. g. load bearings or shear keys) whose resistance and displacement capacity may not be adequate to current standards provisions (NTC 2018). In this study, the seismic behavior of a case-study post-tensioned concrete bridge built in 1985 was numerically analyzed. The bearing devices were designed without considering the effects of seismic action, as well as the expansion joints, seized to allow only thermal expansion. So, their displacement capacity could be inadequate, causing pounding between structural parts or even loss of support; moreover, the resistance against horizontal forces of the bearings could be not sufficient to properly connect the deck to the piers. In order to assess the seismic performance of the bridge, a preliminary Finite Element (FE) model of the structure was developed, and nonlinear dynamic time history analyses were performed: displacements and internal forces induced by the seismic action were evaluated, focusing on joints and load bearing devices, and the results were then compared to the nominal capacity of the elements. More in detail, in Section2 the case-study bridge was described, while in Section3 the characteristics of the numerical FE model were described as well as the features of the nonlinear dynamic analyses carried out in the study. The results were reported and discussed in Section 4, and then some conclusions collect the main target of the study. 2. Case study The case study is a post-tensioned concrete bridge 440 m long, composed by nine spans (inner spans of 55 m, and end spans of 27.5 m). The deck is a curved continuous box girder on multiple supports (Fig. 1a, b), built with precast segments using the balance cantilever construction method. Moreover, it is divided into two parts, connected by a hinged joint in the middle of the central span, made with three shear keys (two placed in the box webs and one in the bottom slab). The box girder section has an average height of 2.50m and hosts a 17.9m wide four-lane carriageway; the transversal inclination of the top slab varies along the bridge, due to different curvature radius of the planimetric layout. The piers are in prestressed concrete and have a rectangular hollow section with stiffeners at the edges, with outer dimension of 2.3m x 3.4m, and a slightly bigger section at the base (from 0 to 3m, see Fig. 1c); the upper part of the piers is solid and it widens up to 7 m in order to accommodate two load bearing devices on each pier. According to both design drawings and test certificates made during the construction of the bridge, the piers were built using a concrete with a cubic compressive strength of 40 N/mm 2 , while for the box girder a stronger concrete was chosen, with a cubic compressive strength of 50 N/mm 2 . Reinforcement bars are in FeB44K steel, while prestressing steel with a tensile strength of 1800 N/mm 2 was used for all structural members. Different load bearing devices were used along the structure: the bearings on the abutments allow both longitudinal and transversal displacements (‘multidirectional devices’), the bearing s on piers next to deck joints allow only longitudinal displacements (‘unidirectional devices’) while other bearing s prevent all translations. All devices are aluminum alloy spherical bearing that allow small rotations, bolted on top of the piers and to the bottom slab of the deck.