Stochastic optimization of uncertain viscous dampers for energy-dissipation structures under random seismic excitations

Abstract

Energy-dissipation devices have been widely used for seismic mitigation of civil structures. Stochastic optimization of energy-dissipation devices has emerged as an appealing means to achieve optimal seismic performance of structures in uncertain environment. This paper proposes an effective optimization framework for nonlinear viscous dampers of large-scale structures in consideration of both random seismic excitations and uncertain damper parameters. A dimension-reduced explicit time-domain method (ETDM) is employed for nonlinear time-history analysis of seismic responses and response sensitivities with respect to the nominal parameters of uncertain viscous dampers. To obtain the statistical moments of critical responses and the corresponding moment sensitivities, the high-efficient ETDM is then used in conjunction with the Monte-Carlo simulation (MCS) with sufficient samples of seismic excitation and damper parameters. The stochastic response and sensitivity results are further utilized for stochastic optimization of the nominal values of damper parameters with the method of moving asymptotes (MMA). The present optimization framework represents a hybrid approach combining the existing ETDM, MCS and MMA. A large-scale suspension bridge installed with nonlinear viscous dampers with uncertain parameters is analyzed to demonstrate the feasibility of the proposed optimization framework, and the influence of the uncertainties of damper parameters on stochastic optimization is also investigated.