A new procedure to derive typological fragility functions for unreinforced masonry structures: an application to a Chilean case

Abstract The assessment of the seismic fragility of unreinforced Masonry (URM) buildings in cities, using advanced numerical approaches, is hampered by the complex connectivity which develops with the diachronic process of urban growth and regeneration. The building stock forming 43 urban aggregates in the historic neighborhood of Yungay in Santiago, Chile, is the focus of this manuscript. The Failure Mechanism Identification and Vulnerability Evaluation method (FaMIVE), a mechanical approach based on limit-state analysis and failure modes, determines the collapse load factors and derive capacity curves for each of the 423 structures surveyed and analyzed. The objective of the study is to correlate specific sets of architectural features of these buildings to their seismic performance as represented through fragility functions. To this end we have introduced a new selection algorithm to automatically group the buildings using an optimal logic tree analysis (LTA). As a result, we obtain clusters of capacity curves using the observable properties of the façades as the decision variables of the LTA, while minimizing the variability of the parameters which define the capacity curves. The median capacity curve of each cluster is then used to derive Analytical Fragility Functions (AFFs), using a capacity-demand approach, which considers different sets of nonlinear spectra. The structure of the LTA is observed to be adequately preserved for fragility functions, fully justifying the subdivision in clusters. The aim of this work is to provide the data to prioritize mitigation strategies that enables us to preserve this heritage, as well as that of other similar historical urban areas in Chile and Latin American cities, which bear a strong architectural resemblance since their foundation.