Symptom-Based Prognosis Through Integrated Digital Models and Experimental Data

Damage identification of civil structures presents significant challenges due to their intrinsic complexity and high uncertainty. Among them, masonry bell towers are exposed to several risks due to their distinctive structural morphology, especially when surrounded by adjacent buildings. In this context, achieving a high level of damage identification is of paramount importance, because the collapse of these historic systems can occur even in operational conditions due to degradation and aging. To address this challenge, the authors propose a Bayesian inference method to quantify damage in structures integrating data from observations and simulations. The concept behind the proposed work lies in estimating the boundary structural states (i.e., theoretical intact and imminent collapse states) within which the damage estimation is performed, providing the non-exceedance probability of a certain damage value that evolves based on experimental observations and data from digital models of the structure. This allows for a more accurate approximation of the residual service life of the structure. The developed methodology is applied to the masonry bell tower of the old parish church of S. Antonio Abate in Monta' (CN), Italy. Starting from experimental data of the damaged state, a numerical model was subsequently modified and calibrated to reconstruct the theoretically intact state of the structure and to estimate, within the Bayesian framework, the value of mechanical parameters related to the imminent collapse state. The proposed approach prevents the underestimation of the damage level commonly inferred by comparing two or more structural states from observed structural features.