Epistemic uncertainties in structural modeling: a blind benchmark for seismic assessment of slender masonry towers

The paper reports the results of a blind benchmark developed as a part of the preliminary activity of the research project RiSEM (Italian acronym for Seismic Risk on Monumental Buildings). The benchmark was aimed at comparing the results obtained with different analytical models and/or numerical analysis techniques (variational approach, finite elements, macroelements, equivalent frame, etc.) for the assessment of the nonlinear structural behavior of two cantilever masonry elements with different slenderness. The analyzed elements were characterized by a deliberately simple geometry, and the comparison between the numerical results had a twofold purpose. On the one hand, it aimed at estimating the effects of the epistemic uncertainties that are related to the different models and numerical techniques. On the other hand, it aimed at reaching a proper evaluation of the influence of parameters describing the postelastic behavior of the structural typology analyzed within the research project (specifically, the masonry towers). Both these objectives were necessary to further proceed with the development of simplified numerical models needed for the subsequent risk analysis. For both slendernesses, the results have highlighted a significant dispersion of both the displacement capacity and postpeak softening branch of the capacity curves. In addition, after some elaborations it has been observed that the dispersion of the results is proportional to both the shear-force and displacement level.