A Bayesian FE-model updating of slender masonry structures for robust seismic assessment

The evaluation of the seismic risk of historic masonry buildings is a significant issue in order to ensure their conservation over the time. Among the different typologies of historic monumental buildings, masonry towers, representing a hallmark of many Italian town centres and widely diffused in the European territory, embody an important heritage to preserve. However, the comprehensive assessment of seismic risk requires a reliable numerical model that captures the nonlinear behaviour of the masonry as well as an explicit consideration of different sources of uncertainty for a probabilistic definition of the seismic capacity and demand. For historic masonry structures, the unavoidable lack of knowledge on the input parameters (e.g., material properties, geometry, construction technique, boundary conditions, etc.), makes still difficult their numerical modeling. In this scenario, the availability of actual experimental data on the real structures can improve the reliability of the finite element (FE) models. This paper proposes a methodology that allows the definition of robust seismic fragility curves for masonry towers through a Bayesian updating of the uncertain parameters, considering different limit states. The methodology is based on nonlinear static analyses of a three-dimensional finite element models, assumed to be representative of historic masonry towers. The nonlinear mechanical parameters of the masonry were considered as modelling uncertainties and different damage states were defined. The towers of San Gimignano have been herein considered as an effective case study to test this methodology thanks to the availability of both geometric data and experimental measurements.