Fragility analysis of masonry structural units by Response Surface method

Structural masonry aggregates include a set of inhomogeneous structural units that can interact under seismic action; therefore the seismic analysis cannot ignore the inevitable interactions resulting from structural contiguity between adjacent buildings [1,2]. The study of a masonry aggregate cannot neglect the unavoidable uncertainties related to geometries and material characteristics of their components. In this work the first results of a research focused on the fragility analysis of masonry structural units by statistical procedures are presented. In order to take the structural and geometrical variabilities and uncertainties involved in the problem into account, the Response Surface (RS) statistical method is used, where the expected value of a response parameter (for instance the acceleration corresponding to the attainment of a prescribed limit state of the building) is approximated through a polynomial function of a set of selected variables [3]. The RS model is calibrated through numerical data obtained by non-linear static analysis, with reference to a masonry building (structural unit) whose geometrical and mechanical properties are varied in prescribed ranges. The number of simulations and the values of the variables to be used to obtain the data for the RS calibration are defined by applying the Design of Experiments Theory [3]. Finally, fragility curves are evaluated through the RS obtained from the numerical simulation data. The results have highlighted the importance of the probabilistic procedures as methods able to consider the variabilities and the uncertainties involved in the problem of the fragility of masonry structures.