Seismic risk-based optimization of tension-only shape-memory alloy device for steel moment-resisting frames

Abstract

This study introduces the seismic risk-based design of SMA devices for steel moment-resisting frame buildings. The investigation begins with the numerical representation of the device which includes SMA wires exhibiting super-elastic stress–strain response, along with steel components. The model is then compared against an actual test result. Accordingly, several four-story steel-frame buildings are numerically characterized, all designed to withstand similar seismic loads. This includes a control frame building (Bare-SMF), a frame building with SMA devices using a traditional design approach (SMA-D), and a frame with buckling restrained braces (BRB-D) for comparison. As the traditional design method relies on the first modal shape and static loads, the pushover analysis results for all structures possessed similar amount of strength and deformation capacity. However, the nonlinear time history analysis revealed inconsistencies in the responses of the retrofitted buildings (SMA-D and BRB-D) when compared to Bare-SMF. To address this issue, a seismic risk-based design approach that combines nonlinear time history analysis on a steel moment-resisting frame and the particle swarm optimization algorithm is employed. By using this approach, the SMA brace quantities were determined such that the cost of SMA devices can be minimized while the life-cycle probability of collapse and demolition is smaller than that of Bare-SMF. As a result of the risk-based design approach, three brace quantities are determined according to a decision maker’s attitude against the seismic risk. As a result, the cost of SMA devices could be reduced by more than 40% while maintaining a significant reduction in the life-cycle probability of collapse and demolition.