Failure probability minimization of structures with multiple friction tuned mass dampers under seismic excitations
Giancarlo Zibetti Mantovani1; Leandro Fleck Fadel Miguel1; Rafael Holdorf Lopez1; Letícia Fleck Fadel Miguel2; André Jacomel Torii3
1 Federal University of Santa Catarina; 2 Federal University of Rio Grande do Sul; 3 Federal University of Paraíba
doi:10.20906/CPS/CILAMCE2017-0542
Resumo
Optimum design of tuned mass dampers (TMDs) and multiple tuned mass dampers (MTMDs) in vibration mitigation due to earthquakes is an important research topic. A literature survey reveals that the controlled system is often considered linear and a deterministic approach is applied. Moreover, we could not find applications of global design optimization of multiple friction tuned mass dampers (MFTMDs) which considers a robust reliability based approach. Hence, this paper presents a novel application on robust optimization of nonlinear MFTMDs to control vibrations in structures submitted to seismic excitations, in order to minimize the failure probability of these structures. The nonlinear solution of this stochastic dynamical system under uncertainties is carried out through a statistical linearization technique in the state space, which requires considerably fewer computational effort than a classical time domain analysis. A hybrid global local scheme composed of the algorithms Firefly (FA) and Nelder Mead (NMA) is applied for the optimization. A numerical example on a ten-story planar steel building frame with one, two and five dampers is demonstrated. The results point out that the proposed methodology can reduce significantly the failure probability of civil structures under seismic excitations.
Palavras-chave: Earthquakes; MFTMDs; Robust optimization; Failure probability; Statistical linearization