Finite Element Modelling of Rocking and Hybrid Precast Walls under Seismic Loading

ABSTRACT In recent years, self‐centring systems have received considerable research interest in the seismic design of precast concrete structures for their ability to limit damage and satisfy resilience requirements. Rocking and hybrid structural walls, characterised by the formation of a single gap at the wall–foundation interface, mitigate damage by permitting controlled rocking motion, while gravity loads and unbonded post‐tensioning tendons provide the restoring force necessary for self‐centring. Supplementary energy dissipation is typically supplied by partially unbonded mild steel bars, producing the characteristic flag‐shaped hysteretic response. This study examines the finite element modelling of rocking and hybrid precast concrete walls, with particular focus on the influence of key modelling assumptions and parameters governing the wall–foundation interface. Both component‐level (single‐wall) and system‐level (full‐structure) models are considered. Three alternative modelling strategies are evaluated to represent the rocking interface: (i) fibre‐based beam elements, (ii) distributed compression‐only translational springs and (iii) lumped nonlinear rotational springs at the wall base. Nonlinear static and dynamic analyses are conducted, and numerical results are validated against shake‐table tests performed on a three‐storey, half‐scale precast concrete building representative of a parking structure, tested on the Large High‐Performance Outdoor Shake Table at the University of California, San Diego. To enable consistent interpretation of the modelling results and to account for the coupled influence of multiple parameters on the nonlinear response, a constrained nonlinear least‐squares optimisation employing a multistart strategy is used to calibrate the numerical models by minimising the discrepancy between experimental and simulated displacement and base shear time histories. The results delineate the relative strengths and limitations of each modelling approach. The proposed wall–foundation rocking interface formulations incorporating either fibre‐based beam elements or distributed compression‐only translational springs demonstrate very good agreement with the experimental measurements.