The seismic design of precast structures hinges on unique characteristics intrinsic to precast technology. Emphasis is placed on lightweight structural elements for efficient on-site assembly and cost reduction. This leads to increased slenderness in beams and columns compared to traditional cast-in-situ constructions, accentuating the role of second-order effects. Dry pinned joints, favoured for connecting beams and columns, contribute to the overall efficiency in assemblage. Cast-in-situ concrete is typically reserved for connection to foundations and topping precast floor elements. Pinned joints transform the structure into an ideal isostatic system, with cantilevered columns anchored securely at the base, in designers' mind. However, this transformation reduces the energy dissipation capacity, preventing plastic hinge formation in beams and amplifying P-Delta effects in columns. The simplified approach proposed herein assesses dynamic instability in single and multi-storey precast hinged frames, providing a tool for expeditious numerical capacity assessment, useful at the initial design stage. The goal is to predict dynamic collapse and/or the attainment of specific seismic-oriented limit states based on fundamental structural parameters. Incremental nonlinear dynamic analysis, utilising far-field ground motion records, is employed to evaluate performance of a variety of precast structures modelled as equivalent single-degree-of-freedom systems. The outcomes yield inelastic spectra that provide insights into the structural capacity in terms of response modification factor and could help analysts/designers towards seismic performance-based design as well as the assessment problem. These spectra, which can themselves be taken as metrics for structural performance evaluation (in addition to as a reliable tool/means for design), are generated based on various structural parameters, including building height, column aspect ratios, and floor mass configurations, in relation to different limit states typically deemed crucial in the design of these structures for earthquake-induced actions. Regression analyses of median spectra show that polynomial expressions could fit them with good accuracy, as testified by a coefficient of determination in the 0.76–0.98 range for most of the cases.
Read full abstract