Probabilistic Seismic Loss Analysis for Design of Steel Structures - Optimizing for Multiple-Objective Functions

An optimized seismic performance-based design methodology considering structural and non-structural system performance and seismic losses is considered to design steel structures. Multi-objective optimization methodology is implemented considering various sets of optimization objectives which would take into account minimization of the initial construction cost, associated with the weight of the structural system, and the expected annual loss considering direct economic losses, and a social loss parameter defined as expected annual social loss. A non-dominated sorting genetic algorithm method is implemented for the multi-objective optimization. Achieving the desired confidence levels in meeting performance objectives of interest are set as constraints of the optimization problem. Inelastic time history analysis is used to evaluate structural response under different levels of earthquake hazard to obtain engineering demand parameters. Hazus fragility functions are employed for obtaining the damage probabilities for the structural system and non-structural components. The optimized designs and losses are compared for example steel structures, located in two geographic locations: Central United States and Western United States