Çelik yapıların çok-amaçlı performans tabanlı tasarım optimizasyonu

Seismic performance of structures designed in accordance with conventional force-based design (FBD) codes can vary significantly since (1) estimation and distribution of earthquake loads are based on initial stiffness of members and (2) force reduction factors are based on the rough assumption that all members will yield simultaneously. In recent decades, attempts to predict the seismic performance of structures resulted in the development of several guidelines for evaluation and rehabilitation of existing buildings (e.g., FEMA 273/356, ASCE/SEI 41). On the contrary, currently there is no performance-based design (PBD) code. As a matter of fact, while seismic performance assessment of existing buildings is a difficult task alone due to wide range of variabilities involved in seismic design parameters; PBD of a new structure by conventional means is practically impossible as size and detailing of members are also unknown at the start of the design process, yielding in numerous alternative design solutions. Although it is possible to automate the PBD process by optimization methods, a large number of nonlinear analyses required during optimization process take excessive computational time. In this study, a novel design-driven optimization technique called Capacity Controlled Search (CCS) is proposed for achieving time-efficient optimum design of steel structures under the FBD and PBD methodologies. The success of the proposed method is first numerically investigated and justified by comparing its performance with those of several metaheuristic techniques on FBD optimization problems featuring various 2-D and 3-D ordinary moment resisting steel frames. Later, the same steel frames are optimally designed by PBD approach using the CCS method. A comparison of optimally designed structures via FBD and PBD methodologies is then carried out in terms of design cost and seismic performance. Finally, a practical multi-objective optimization approach is adopted to form the trade-off relationship between design cost and seismic performance, and alternative performance-based design solutions are presented. The numerical results indicate the computational efficiency of the proposed optimization technique and suggest that more economical designs with predictable seismic performance can be produced for steel frames by the PBD approach than the conventional FBD.