Probabilistic Performance-Based Optimal Design of Steel Moment-Resisting Frames. II: Applications

Abstract

Design codes are migrating from prescriptive procedures intended to preserve life safety to reliability-based design. All stakeholders are given the opportunity to speak the common language of risk and structural designs can be developed to not only reliably preserve life safety after rare ground motions, but minimize damage after more frequent ground motions (minimize life-cycle costs). A companion paper presents a methodology for using an evolutionary (genetic) algorithm with radial fitness and balanced fitness functions to generate solutions to optimal design problems formulated within a probabilistic performance-based design framework. The present paper outlines application of the automated algorithm to design steel frames with fully restrained and a variety of partially restrained connections. Comparison of optimal designs resulting from application of the algorithm proposed with those found in the literature is made. Detailed discussion of algorithm performance and response of the resulting optimized designs during pushover and time-history analysis is provided.