Integrated nonlinear modelling strategies for the seismic analysis of masonry structures

In the last decades, significant interest has raised in modelling and analysing the structural response of unreinforced masonry (URM) buildings. This aims at conceiving and designing effective interventions to reduce the vulnerability towards seismic actions. Studies based on costly structural testing are often limited to few benchmark cases, making numerical modelling an excellent option to extend experimental results and a valid solution for understanding URM structural behaviour. Advanced discrete models are widely employed among the available numerical strategies to predict the URM dynamic response, thanks to their ability to account for the heterogeneous nature of masonry and to simulate its behaviour up to the complete collapse. If, on the one hand, the low degree of idealisation of discrete models allows their employment for the extension of experimental tests, on the other hand, they require expert users, the definition of a large number of mechanical parameters and a high computational effort. This last drawback often limits the use of advanced discontinuum models in the engineering practice or for seismic risk studies, which require the execution of multiple analyses. In this work, a modelling approach, based on the Applied Element Method (AEM), was combined with more simplified models to exploit the discrete model potential and overcome its limits. To this aim, the AEM was employed as a benchmark to calibrate/validate simplified modelling strategies, improving their reliability when compared to advanced model outcomes. In this context, AEM models were used as a reference to enhance the Equivalent Frame Model (e.g. the presence of irregular distribution of openings) and to validate a new strength criterion associated with the failure mechanism encountered in a new masonry typology. In the absence of a large suite of experimental tests exploring all the possible setup or configurations, the AEM can provide precious information. On the other hand, the AEM can help to investigate situations requiring a higher level of detail, such as the design of the timber retrofitting system analysed in this work. The ability of the AEM to simulate the structural behaviour up to the complete collapse was also used to investigate the effect of different percentages of ground floor opening on the dynamic response of Dutch terraced houses, performing benchmark analyses to calibrate SDOF models employed for the development of fragility functions associated with the different layouts. Finally, AEM models were employed for substructuring façade models of masonry buildings whose global response was effectively studied by equivalent frame models. The aim of the study was to predict the debris extent involved in the collapse of URM façades in case of earthquake loadings. Such an integrated numerical procedure allowed considering a large suite of seismic inputs, overcoming the time-consuming issue.