Advanced Model Calibration of a Historical Masonry Tower in OpenSeesPy via Genetic Algorithm and Particle Swarm Optimization

Masonry bell towers, vital to European cultural heritage, require calibrated finite element models (FEM) to predict structural responses to ambient and seismic excitations reliably. This study extends previous work on the San Giuseppe bell tower in Aci Castello, formerly calibrated manually, by leveraging OpenSeesPy’s capability to integrate with optimization algorithms. Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) were adopted to calibrate the FEM model by minimizing an objective function, which is the error between numerical and experimentally derived modal properties, including natural frequencies and mode shapes obtained from operational modal analysis (OMA). To evaluate the effectiveness and applicability of GA and PSO in calibrating the FEM of the case study, a two-phase approach was adopted. In the first phase, the numerical model’s outputs, natural frequencies, and mode shapes of the first 10 and 5 modes, served as reference data to test the algorithms’ ability to recover known parameters. In the second phase, to calibrate the model, the actual experimental data, including only the natural frequencies of the first three modes derived from the OMA, were used. Results show that both algorithms estimated the model parameters with notable accuracy, with GA demonstrating superior performance to PSO, particularly when both natural frequencies and mode shapes were incorporated into the objective function. The suitability of OpenSeesPy for automated FEM calibration was confirmed by the findings, along with the superior robustness of GA in estimating mechanical properties, particularly elastic moduli, providing a reliable approach for calibrating finite element models of historical masonry towers.