RISK-BASED MULTI-HAZARD OPTIMIZATION OF PASSIVELY DAMPED STRUCTURES USING EVOLUTIONARY ALGORITHMS
A new computational methodology is developed for conducting Risk-based Optimal Multi-hazard Design (ROMD) of seismic- and wind-excited structures retrofitted with passive energy dissipation (PED) devices. In addition to designing in compliance with the relevant codes of practice, it is important to consider that the performance of PED devices for reducing the structural responses depends on the type, size, and distribution of dampers. The proposed framework provides a genetic algorithm (GA) based methodology to address these optimization issues of multi-hazard design within the context of nonlinear steel frame structures. Steel buckling restrained braces, viscous fluid dampers and solid viscoelastic dampers are all considered as possible design alternatives within this framework. In the proposed algorithm, passively damped structural designs evolve toward configurations that limit damage associated with inter-story drift and absolute floor acceleration, while considering essential conflicts in dynamic response demands of the structures under multi-hazard environments, involving earthquakes and strong windstorms. Unlike previous work in PED optimization, the ROMD approach compares the life cycle costs and benefits of the alternative design or retrofit strategies using an optimization criterion that reflects physical and economical uncertainties, as well as the attitudes of decision makers through the introduction of a risk measure.This measure characterizes the risk tolerance of the decision makers and allows the evolution of rational design or retrofit strategies that depend upon the level of risk aversion to low frequency, high consequence events. Besides providing an outline of the evolutionary risk-based algorithm, this paper includes an example to emphasize the potential benefits of the proposed computational multi-hazard design approach.
Optimal structural design multi-hazard mitigation life-cycle cost risk aversion genetic algorithms
S. Dogruel G.F. Dargush
Civil, Structural and Environmental Engineering, University at Buffalo, Buffalo, NY, USA Mechanical and Aerospace Engineering, University at Buffalo, Buffalo, NY, USA
国际会议
14th World Conference on Earthquake Engineering(第十四届国际地震工程会议)
北京
英文
2008-10-12（万方平台首次上网日期，不代表论文的发表时间）