Response surface analysis and optimization of controlled rocking steel braced frames

Abstract

As an alternative to conventional seismic force resisting systems, controlled rocking steel braced frames (CRSBFs) can effectively eliminate permanent structural damage after earthquakes. Together with the rocking action in the braced frame, post-tensioning (PT) elements and fuse members are used to provide self-centering and energy dissipation, respectively. This study firstly aims to assess the influence of design parameters related to the fuse and PT materials on the seismic response of CRSBFs. These factors include the yield strength, initial stiffness, and strain hardening ratio of the fuse, the initial force and modulus of elasticity of the PT strands, and the gravity load on the rocking column. Additionally, different analysis cases are considered to include the effects of frame aspect ratio and earthquake intensity level. Nonlinear response history analyses are performed with factor combinations generated using a design of experiment methodology. The second goal of the study is focused on optimizing the seismic response of CRSBFs with respect to the influential factors identified in the sensitivity analyses. Using a response surface methodology and desirability approach, multiple-response optimization is applied to determine the design variable values needed to simultaneously minimize the maximum transient and residual roof drift ratio and peak floor acceleration. Among other results, it is found that the fuse strain hardening ratio and PT strand modulus of elasticity do not significantly influence the seismic response demands in CRSBFs. The results of the multi-response optimization demonstrate that the initial PT force is most useful for minimizing all three seismic response demand parameters.