Fidelity and robustness of detailed micromodeling, simplified micromodeling, and macromodeling techniques for a masonry dome

Understanding the damage and failure mechanisms of masonry structures can help engineers reduce catastrophic failures and facilitate effective restoration and preservation of historical masonry monuments. This can be achieved through a combination of experimental and numerical studies to gain insights on the macrolevel strength-deformation behavior and microlevel defects and crack growth of masonry structures. Although experiments aid in calibration and validation of the numerical model to reduce errors and uncertainties in predictions, the success of the simulations fundamentally depends on the accuracy of the mechanical principles used to represent the heterogeneous masonry assembly. In this paper, three modeling techniques - detailed micromodeling, simplified micromodeling, and macromodeling - are investigated, considering not only the accuracy but also the robustness of the model predictions. In detailed micromodeling, the brick units and mortar joints are modeled as separate entities. In simplified micromodeling, the bricks and mortar are smeared, homogenized units bonded with zero-thickness interface elements. In macromodeling, the masonry composites are smeared into a homogenous continuum. Linear properties of these three alternative models are first calibrated by exploiting the modal parameters identified through dynamic experiments conducted on a scaled dome specimen in the laboratory. The fidelity of the two micromodeling and the macromodeling techniques are then evaluated by comparing the model predictions against static, load-to-failure tests conducted on the same scaled masonry dome. Finally, the robustness of the three models to uncertainty in the input parameters is evaluated.