دانلود رایگان مقاله انگلیسی ارزیابی روش های طراحی بهینه میراگرهای ویسکوز - الزویر 2017

Abstract

Viscous dampers are often used for seismic protection and performance enhancement of building frames. The optimal design of such devices requires the modelling and propagation of the uncertainties related to the earthquake hazard. Different approaches are available for the seismic input characterisation and for the probabilistic response evaluation. This work analyzes the effect of different characterizations of the seismic input and of the response evaluation on the design of dampers for building frames. The seismic input is represented as a stochastic process and the optimal damper properties are found via a reliability-based design procedure aiming at controlling the frame performance while limiting the damper cost. Two simplified approaches are used to design the viscous damper of a multi-storey steel frame and the design results are compared with those obtained by considering a rigorous design approach resorting to advanced simulations for the response assessment. The first methodology evaluates the response through a prefixed probabilistic demand model, while the second approach considers the average response for a given hazard level only. The comparison allows to evaluate and quantify the effect of the seismic input uncertainty treatment on the system and damper performances.

5. Conclusions

Two approximate methodologies for the optimal design of viscous dampers have been presented in this study. The approaches allow controlling the maximum inter-storey drift by reducing at the same time the total damper forces applied to the viscous dampers that are known to be proportional to the total cost of the dampers. The first approximated approach allows a reduction in the number of simulations during the reliability analysis thanks to the implementation of a probabilistic model that exploits a linear regressions of the structural response vs. the seismic input intensity to calculate the input vector for the optimisation step.

The second methodology presented optimises the system by considering a single level of the seismic hazard during the input selection, and by studying the structural performances always over the same pre-defined set of accelerograms. This methodology is the least computational expensive due to the reduced number of structural analyses required for each optimisation loop.

Although a complete reliability based optimisation allows a more accurate design solution which ensures more strictly the performance objectives, the approximate approaches yield results relatively close in terms of optimal design at a fraction of the total computational cost. The obtain results show that for increasing values of the dampers nonlinearity both the damping viscous constant (i.e., design variable) and the objective function decrease. Thus, dampers with higher nonlinearity levels perform better than linear dampers. Further analyses involving structural systems other than the one investigated in this paper are however needed to generalize the obtained results.