Nonlinear macroelement for steel shear walls under seismic loading

This paper proposes an efficient macroscale representation for unstiffened thin steel shear walls, enabling accurate and efficient predictions under earthquake loading. The 8-noded macroelement incorporates six nonlinear springs with asymmetric constitutive relationships to represent the cyclic response of steel panels. The macroelement formulation is introduced first, followed by the constitutive model for the nonlinear springs, which allows for strain-hardening in tension, strength degradation in compression, residual strength at the strain reversal point, and stiffness reduction. The material parameters for the constitutive model are calibrated by multi-objective optimisation with Genetic Algorithms based on the numerical results provided by accurate nonlinear FE models with shell elements. Linear regression is utilised to establish the material parameters for infill plates with different geometric properties. The residual of dissipated energy for the calibrated macroelement models lies between 3 % overestimation and 12 % underestimation compared to detailed shell element models, whereas the computational demand is reduced with wall-clock time reductions of more than 97 %. Finally, the proposed macroelement is verified in numerical examples, where a substandard RC frame enhanced with different types of steel shear walls is analysed under cyclic loading using detailed shell element models and the proposed macroelement for the steel wall components. The excellent comparisons confirm that the proposed macroelement model provides an efficient and accurate description of unstiffened steel wall components, and it can be used for realistic nonlinear dynamic simulations of framed buildings equipped with steel walls under earthquake loading. • Efficient macroscale model for steel shear walls. • Material relationship representing the main response characteristics under cyclic loading. • Model calibration by multi-objective optimisation with Genetic Algorithms. • Model validation via comparison against results from physical experiments. • Accurate simulation of substandard RC frames retrofitted with steel shear walls.