PSI - Issue 78

Paolo Ielpo et al. / Procedia Structural Integrity 78 (2026) 1024–1031

1029

Based on these findings, the numerical model calibration process was initiated, focusing on parameters that most significantly affect simulation accuracy. Particular attention was given to modeling the cracking phenomenon, determining the fracture energy (Gf) according to the Model Code 2010 formulation, resulting in a value of 0.127 N/mm. The bond-slip interaction between concrete and reinforcement was accounted for to improve the accuracy of the model. A splitting-type failure mechanism was adopted, assuming effective bond conditions between reinforcement and concrete. The confining effect of stirrups was also considered. Based on these assumptions, the maximum shear stress was determined as τ max = 7.70 N/mm² (Model Code 2010). The results obtained show good agreement with the experimental data, both in terms of numerical values, with a percentage variation not exceeding 7% for the key parameters, and in the trend of the load-drift curve (Fig. 5). The buckling phenomenon observed at 3% drift was not modelled, as it is beyond the scope of this study.

Fig. 5. (a) as-built FEM model and (b) comparison of numerical the experimental load-drift curve

5. Numerical modeling of the SPEAD device The SPEAD retrofit system was integrated into the calibrated numerical model of the T4 joint to evaluate its effectiveness in enhancing local seismic performance. The device was designed following the procedure proposed by Santarsiero et al. (2020), which allows for customized sizing based on the geometric and mechanical characteristics of the joint. In this study, the device configuration includes three circular holes with a diameter of 60 mm, spaced 10 mm apart, defining the so- called “hourglass zone,” a critical area for energy dissipation. In this zone, the steel component undergoes controlled plastic deformations, ensuring efficient hysteretic energy dissipation in response to cyclic seismic loads. The sizing procedure also precisely defined the internal lever arms of the system: specifically, the internal lever arm B, measuring 540 mm, is key for estimating the device’s yielding moment (My SPEAD ), while the lever arm B ’ , measuring 85 mm, prevents undesirable tensile stresses in the hourglass zone by absorbing the applied shear forces. The device was modeled with a constant thickness of 13 mm, while the upper and lower anchorage zones, located along the vertical axis of the column, each having a length of 500 mm. The material is ordinary steel S355 class, characterized by a yield strength of 355 N/mm². An elastoplastic constitutive law with strain hardening was adopted for the modeling, with a hardening modulus of 1000 N/mm². Finally, the numerical model of the device was discretized using a mesh composed of 9,010 linear triangular finite elements, ensuring adequate numerical resolution in the critical areas subject to plastic deformations.