Simplified Assessment of the in-Plane Seismic Response of Old Brick Masonry Building Aggregates Using DE Macro-Crack Networks

Old buildings were often constructed adjacent to each other, without the minimum gap recommended by modern codes. This further increases their seismic vulnerability by exposure to the risk of pounding, a complex mechanism involving repeated impacts between adjacent buildings. Although post-earthquake surveys worldwide confirmed that seismic pounding can significantly increase the extent of in-plane damage and cause early collapses, this phenomenon still remains largely unexplored, while ad-hoc assessment guidelines are missing. This preliminary study focuses on investigating the mechanical in-plane interaction among low-rise unreinforced masonry (URM) buildings of clay brick, a seismically vulnerable yet common structural typology across Canada and abroad. The main novelties consist in the unprecedented use for this task of experimentally validated numerical models developed in the Distinct Element Method (DEM) framework, enabling us to map accurately crack propagation up to collapse, as well as the quantification of key material and geometrical factors affecting earthquake performance. To reduce the otherwise prohibitive computational expense typically entailed by DEM and consider building-scale models, a new macro-modelling strategy is devised that idealizes masonry as an assembly of solid rigid blocks connected by nonlinear interface springs, forming an equivalent macro-crack network where failure occurs according to linearized softening joint constitutive laws. Using this expedited yet accurate analysis technique, a comprehensive parametric study is conducted to investigate the pounding of adjacent URM façades of varying height, material degradation levels and opening layout, tested under pushover loading schemes. Preliminary results, which also account for the stochastic nature of the mechanical properties of masonry, seem to suggest that the severity of damage due to building interaction is particularly dependent on the material properties, adjacent building numbers and the building height. These results will inform ongoing research on seismic pounding at McGill University, where the effect of dynamic loading will also be considered.