Shake Table Investigation of Structure‐Soil‐Structure Interaction of Idealized Urban Building Clusters

ABSTRACT

The increasing urban density has led to buildings being constructed closer together, which can result in complex interactions between structures during seismic events. This study investigates this interaction, commonly referred to as structure–soil–structure interaction (SSSI), through shaking table tests on scaled structural models founded on a polyurethane foam block. The foam layer is used as a repeatable compliant soft‐soil analogue within a mechanism‐based similitude framework. Isolated SSI configurations and homogeneous and mixed‐height SSSI clusters comprising 9 and 21 buildings were tested, including target buildings located at both central and edge positions. The experimental programme was designed to evaluate how the number of neighbouring buildings, plan layout (symmetry vs. asymmetry), and variations in building typology influence seismic demand and structural response in dense urban environments. Four recorded earthquake ground motions and a white noise excitation were applied. The dynamic acceleration response is quantified using (i) absolute peak roof acceleration (PRA) and (ii) total energy spectral density (ESD), which respectively characterise the maximum instantaneous response and the accumulated acceleration content. The results show that SSSI effects depend on the spatial distribution of buildings and seismic input. For the central low‐rise building, all clustered configurations reduced both PRA and ESD relative to the isolated case. The taller building exhibited a more complex response, depending on the input motion and cluster distribution. Specifically, SSSI can reduce the maximum acceleration while increasing the accumulated response through temporal and spectral redistribution. The experiments also showed that building position is important: edge buildings in mixed‐height clusters exhibited greater response than equivalent edge buildings in homogeneous layouts. These findings demonstrate that SSSI effects cannot be characterised solely by peak response and that cluster heterogeneity and target‐building location are key variables in the experimental assessment of dense urban configurations.