Evaluation of different approaches for the estimation of the seismic vulnerability of masonry towers

A series of simplified approaches are evaluated for their effectiveness to estimate the seismic vulnerability of historical masonry towers. First, collapse loads are evaluated on sixteen ‘‘idealized’’ benchmark cases with different slenderness and shear area. Both analytical and computational approaches are used, namely the analytical procedure proposed by the Italian Guidelines on the Built Heritage and pushover analyses conducted using the commercial codes UDEC and 3Muri. The sixteen towers are representative cases which can be encountered in practice. Their geometry is idealized into parallelepiped blocks with hollow square cross-sections, thus favoring the utilization of 2D approaches, beneficial to drastically reduce the effort required for repeated computations. In addition, a Monte Carlo MC upper bound limit analysis strategy is proposed, in order to have an insight into the possible failure mechanisms for the different cases investigated. Deliberately is avoided the introduction of any form of irregularity and they are supposed isolated from the neighboring buildings, to obtain results exclusively dependent from geometric features. Among all the possible collapse mechanisms, five of them are selected according to the probability of occurrence based on past earthquake experiences. Five million cloud points of collapse accelerations are obtained by carrying the height, slenderness and shear area of the idealized towers. The approach is very fast and allows identifying different regions where single mechanisms are active. The results are confirmed repeating MC simulations with a triangular FE upper bound limit analysis discretization of the idealized towers. A series of equations are provided in order to assist engineers and practitioners to obtain a preliminary estimation of their expected collapse acceleration. For validation purposes, the results obtained previously with refined full 3D FE models of 25 towers located in the Northern Italy are reported. Satisfactory agreement between the predictions provided by simplified methods and sophisticated analyses are obtained.