Abstract Summary
Unreinforced masonry (URM) buildings show high vulnerability to seismic loading. The performance of such buildings during past earthquakes has highlighted that wall collapse in the out-of-plane direction represents a major source of concern. Depending on the connection of a wall with diaphragms and transversal walls, different resisting mechanisms to the out-of-plane loading can develop and, hence, different configurations are commonly distinguished. Two-way spanning walls are characterized by effective restraints at least at one lateral side of the wall. The capacity of a two-way spanning unreinforced masonry wall with a single opening under seismic loading has been investigated in several reported studies by means of experiments and further derivation of analytical formulations. However, two-way spanning walls with more than one opening lack both experimental data and analytical formulations proposed to predict the out-of-plane capacity. This study presents an engineering approach to calculate the out-of-plane capacity of four-sided supported walls with two openings. The proposed methodology involves dividing the wall into simple components, whose performance may be assessed separately by means of available formulations, such as those recommended in the Dutch guidelines for the assessment of structural safety under earthquakes, NPR-9998:2020. The division into components is based on simple geometrical considerations. The first two components are defined as the wall portions comprised between side support and one opening. These components are analyzed as equivalent three-sided supported walls. The length of such components is increased fictitiously to account for the presence of the openings which is calibrated based on consolidated expressions defined for walls with a single opening. The second component corresponds to the wall portion between the two openings and is analyzed as a one-way spanning wall. The remaining components corresponding to the portions above and below the openings are analyzed both as cantilever walls and three-sided supported walls. This is dependent on whether the failure of the adjacent components is expected to occur or not and, consequently, whether lateral restraints can be provided. The applied seismic demand used for the analysis of the components is based on the natural frequency of the original wall with two openings and not on that of the components themselves. The adequacy of the proposed approach is evaluated by cross-comparing the outcomes of predictions computed based on both analytical calculations and finite element analyses. By investigating a number of practical configurations, it is shown that the presented engineering approach is conservative and efficient awaiting further validation.
