PSI - Issue 78

Abed Soleymani et al. / Procedia Structural Integrity 78 (2026) 815–822

818

4. Numerical modeling Using SAP2000 software (CSI, 2022), a nonlinear FE model of the RC deck-stiffened arch bridge was built, as shown in Fig. 2. Two-node frame elements with three translational and three rotational degrees of freedom were applied to model the structural components, including pillars, deck beams, and arch elements. Moreover, transverse slender plain concrete walls located between pillars were modeled by shell elements in the FE model. Fixed supports at the end of the elements were assumed for arch elements and pillars in the FE model. Also, the flexibility of abutments was neglected. Therefore, the displacements at both ends of the bridge were constrained in both X- and Y-directions. The deck slab was considered as a diaphragm constraint in the FE model.

Fig. 2. FE model of RC deck-stiffened arch bridge.

In order to define the nonlinear behavior of the two-node frame elements, the plastic hinge model ‘ Fiber P-M2 M3 ’― which considers an integrated solution between the distributed plasticity approach and the concentrated plasticity approach ― was implemented in the FE model. After that, the plastic hinge length was defined according to Eq. (2) (Paulay and Priestley, 1992): = 0.08 + 0.022 (2) where: is the length of the structural element (in mm), is the yielding stress of reinforcement (in MPa), and is the diameter of longitudinal reinforcement (in mm). Shear plastic hinges were assigned to define the shear behavior of all two-node frame elements. All ‘ Fiber-P-M2 M3 ’ elements were assigned to relative lengths of 0.05, 0.5 and 0.95, whilst shear plastic hinges were assigned to the central point of each frame element. It is worth mentioning that the nonlinear behavior of the transverse plain concrete walls located between pillars was considered using the shell-layered/nonlinear approach in SAP2000 (CSI, 2022). 5.Seismic analysis of the examined bridge Seismic resistance of the examined bridge and damage induced by earthquake ground motion were assessed through NTHA. In this respect, a Rayleigh damping model was applied, where 5% damping ratio was assumed for the two global vibration modes corresponding to displacements in directions X and Y. In order to determine the global vibration modes, modal analysis was carried out. According to the results of modal analysis, periods of 0.421 s and 0.276 s were considered as two global vibration modes corresponding to longitudinal (X) and transversal (Y) direction, respectively. In NTHA, selecting appropriate ground motion records is a crucial issue to accurately predict the seismic response of the structure. In order to select ground motion records according to criteria by Eurocode 8 – Part 1 (herein abbreviated as EC8 – 1; see EN1998-1, 2004), the REXELweb software tool (Sgobba et al., 2019) was used. That