PSI - Issue 78

Livia Fabbretti et al. / Procedia Structural Integrity 78 (2026) 823–830

824

1. Introduction Base isolation currently represents one of the most effective solutions for the protection of strategic buildings, as it allows decoupling the seismic ground motion from that of the superstructure, significantly reducing the transmitted forces. Despite the well-established effectiveness of this strategy, a correct characterization of the mechanical properties of the isolators remains fundamental to ensure the reliability of numerical models and, prospectively, to monitor the efficiency of the isolation system over time. The present work originates from the results of a parametric analysis conducted by Fabbretti et al. (2025) on a base isolated structure equipped with FPS (Friction Pendulum System) devices and a permanent structural monitoring system recently installed on the building. This case constitutes a particularly relevant application example, since permanent SHM systems are rarely installed on base-isolated buildings, despite growing scientific interest. Although FPSs demonstrate excellent performance in the event of strong earthquakes, as shown in Saito (2015) and Celebi (1996), the conducted analysis highlighted a recurring issue in practical applications, as reported in Clemente et al. (2016), Salvatori et al. (2019), Clemente and Martelli (2019), Clemente et al. (2019), and Scafati et al. (2024). Indeed, moderate-intensity earthquakes recorded at the site were not sufficient to activate the sliding of the isolators, although they represent the most relevant local events of the past twenty years. Consequently, future earthquakes of similar intensity will not allow for reliable calibration of the model. To overcome this limitation, a dynamic identification procedure of the mechanical parameters of the FPS isolators has been developed – specifically, the friction coefficients, initial stiffness, and rate parameter – using simulated data. The methodology combines the use of a simplified model, integrated within a hybrid MATLAB-SAP2000 computational environment, with a genetic algorithm for automated optimization. Examples of the use of genetic algorithms for optimization problems related to structural identification can be found in Perera and Torres (2006), Rao et al. (2004), He and Hwang (2006), and Erdogan and Bakir (2013). In the future, as soon as useful monitoring data are available, the methodology will be applicable to real, rather than simulated, data to quickly achieve a reliable calibration of the model concerning the mechanical parameters of the isolators. The building under study is the headquarters of the Guglielmo Marconi Scientific and Artistic High School in Foligno (Perugia, Italy). It is a four-story RC frame structure above ground, with prefabricated predalles-type slabs and deep pile foundations. The integration of the isolation system was achieved through a substructure composed of squat pillars, on which thirty-eight isolators were installed, located between the pilotis level and the first-floor slab. This configuration minimizes the risk of damage in case of flooding from the nearby Topino River. The isolators used are ISOSISM ® PS double pendulum metallic devices, manufactured by Freyssinet. They all share the same basic geometry, with an equivalent curvature radius of 2.5 m, but differ in terms of maximum axial force capacity( N max ) and maximum theoretical horizontal displacement ( δ max ). The structural monitoring system installed in the building includes: five high-sensitivity, low-noise MEMS accelerometers (one triaxial, two biaxial, and two monoaxial), two displacement transducers ±250 mm with a resolution below 0.1 mm, an anemometer for measuring wind direction and intensity, and two thermocouples for temperature monitoring. All sensors are connected to a centralized acquisition system. The sensor layout, visible in the finite element model (FEM) developed in SAP2000, version 22.0.0 (Fig. 1), was calibrated based on the nearly rectangular plan shape of the building and the high diaphragm stiffness of the slabs. 2. Reference building and experimental context 2.1. Structural description and monitoring system

2.2. Mechanical modeling of FPS isolators

As highlighted in Fabbretti et al. (2025), the mechanical behavior of frictional double pendulum isolators can be represented through a bilinear hysteretic model (Fig. 2a), whose dynamic response depends on several fundamental parameters that must be correctly incorporated within the FEM, as they significantly influence the global response of the structural system.