Optimizing Viscous Damper Placement for Stochastic Performance of Buildings with Plan Asymmetry

Abstract

Optimized viscous dampers mitigate the dynamic response of building with plan asymmetry. Previous research efforts limit the optimization method to time domain requiring time history analyses to optimize damper placement for plan eccentricity-induced torsional response combined with translational response. This can consume high computational effort and lead to the variability in the ground motion selection. To expedite the optimization process while taking ground motion variability into account, this study will offer a novel criterion based on stochastic performance. This technique will skip the traditional approach of employing time history analyses and will take frequency domain variability into account. Through given distance between the center of mass (CM) and the target frame, the study establishes the stochastic criterion for the response at the perimeter of the building where the maximum torsional-translational response occurs. This allows the optimization methods to generate high quality solution with less computational efforts compared to traditional approach (i.e. time domain-based optimization). It is noteworthy that the proposed criterion is applicable to extensive kinds of responses (e.g. peak floor acceleration, story shear and etc.) once the geometric relationship is given. The proposed criterion as the objective function is compared against their counterpart in time domain through a case study. The case study is based on a reinforced concrete (RC) moment-resisting-frame (MRF) building with two-way asymmetric plan and adopts Element Exchange Method (EEM) to optimize the damper placement. The effectiveness of optimized design based in frequency domain and time domain is evaluated for their total running time and maximum peak interstory drift against a suite of ground motions. The result showed that the proposed criterion can offer the improvement against the traditional approach in terms of solution quality (i.e. performance in time domain and frequency domain) and computational efforts (i.e. total running time for optimization).