Earthquake Engineering & Structural Dynamics
Ioannis Gidaris, 1; Alexandros A. Taflanidis1,*,†; Diego Lopez-Garcia, 2,3; and George P. Mavroeidis, 1
1 Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Notre Dame, IN 46556, U.S.A.
2 Department of Structural and Geotechnical Engineering, Pontificia Universidad Catolica de Chile, Santiago, Chile
3 National Research Center for Integrated Natural Disaster Management CONICYT/FONDAP/15110017, Santiago, Chile
The design of floor isolation systems (FISs) for the protection of acceleration sensitive contents is examined considering multiple objectives, all quantified in terms of the probabilistic system performance. The competing objectives considered correspond to (i) maximization of the level of protection offered to the sensitive content (acceleration reduction) and (ii) minimization of the demand for the isolator displacement capacity and, more importantly, for the appropriate clearance to avoid collisions with surrounding objects (floor displacement reduction). Both of these objectives are probabilistically characterized utilizing a versatile, simulation-based framework for quantifying seismic risk, addressing all important uncertainties related to the seismic hazard and the structural model. FIS performance is assessed through time-history analysis, allowing for all important sources of nonlinearity to be directly addressed in the design framework. The seismic hazard is described through a stochastic ground motion model. For efficiently performing the multi-objective optimization, an augmented surrogate modeling methodology is established, considering development of a single metamodel with respect to both the uncertain model parameters and the design variables for the FIS system. This surrogate model is then utilized to simultaneously support the probabilistic risk assessment and the design optimization to provide the Pareto front of dominant designs. Each of these designs establishes a different compromise between the considered risk-related objectives offering a variety of potential options to the designer. Within the illustrative example, the efficiency of the established framework is exploited to compare three different FIS implementations, whereas the impact of structural uncertainties on the optimal design is also evaluated.