The development of self‐centering steel moment frames incorporating extended endplate connections with shape memory alloy (SMA) bolts offers a solution to mitigate the structural repair costs following earthquakes. Optimization methodologies have been effectively applied to design self‐centering frames with reduced structural repair costs and minimized use of SMA. However, their seismic performance, considering structural and nonstructural components, has not been evaluated. Using the FEMA P‐58 methodology, this study performs a probabilistic evaluation of the seismic performance of optimal conventional and self‐centering frames. Losses induced in optimal structures due to earthquakes are assessed, considering structural and nonstructural components. The seismic performance of optimal three‐ and nine‐story conventional and self‐centering moment frames is evaluated and compared based on repair costs, repair times, carbon emissions, and the probability of demolition. The findings reveal that optimal self‐centering frames equipped with SMA connections result in up to a 37% reduction in median structural repair costs for three‐story frames and a 57% reduction for nine‐story frames when compared to their counterparts with conventional moment connections across different seismic intensity levels. Based on intensity‐based assessments, self‐centering frames exhibit superior structural performance, with up to 83% and 19% lower demolition probabilities for three‐ and nine‐story structures, respectively. However, because of greater damage to acceleration‐sensitive nonstructural components, the optimal self‐centering structures experience higher downtime and carbon emissions impact after severe earthquakes. Median repair costs associated with acceleration‐sensitive nonstructural components are up to 14% higher in three‐story self‐centering frames and 19% higher in nine‐story self‐centering frames due to higher peak floor accelerations, considering different seismic intensity levels.
Hassanzadeh et al. (2026) studied this question.