Reliability based multi-objective robust design optimization of steel moment resisting frame considering spatial variability of connection parameters

• A multi-objective reliability based robust design optimization method is provided.
• The provided method can improve collapse prevention reliability and reduce cost.
• Seismic sensitivities of connection parameters are studied.
• Effect of spatial variability of connection parameters is studied.
• Perfect spatial correlation of connection parameters is a conservative assumption.

The focus of this paper is on achieving robustness in the seismic design against uncertainties in the design process. In this regard, we present a reliability based robust design optimization approach to achieve designs that are not only safe and cost efficient, but also robust against uncertainties. The aim here is to adjust easy-to-control design parameters to reduce the variability of the seismic response to hard-to-control noise parameters while also considering design safety and budget. The proposed approach is demonstrated with a steel moment resisting frame design, considering uncertainties and spatial variability in connection parameters. Here, the sizes of the steel sections constitute the design parameters, while the uncertain and spatially variable parameters of the Ibarra–Krawinkler connection model are treated as noise parameters. The design objectives include the collapse prevention reliability conditional on a maximum earthquake intensity level, as well as the initial cost of the structure. To reduce computation demands in calculating the seismic response variation, design parameters that have a negligible effect on seismic response are first eliminated through sensitivity analysis. Furthermore, the seismic response variation is evaluated parametrically to investigate the effect of connection parameter spatial variability. Finally, the authors demonstrate the use of proposed robust design optimization approach to obtain a Pareto Front, a collection of optimal designs that are optimized for both reliability and cost.