Seismic design with viscous devices is an effective means of dissipating seismic energy and protecting the main structural system from permanent damage. Despite numerous damper placement methods available, there lacks consensus on the best method, thus leaving design engineers without recommended placement strategies.
The purpose of this research is to investigate strategic placement of viscous dampers for seismic design and to offer design recommendations for placing dampers, vertically and horizontally. Five damper placement techniques were selected for investigation, including standard methods, Uniform and Stiffness Proportional damping, and advanced methods, the Simplified Sequential Search Algorithm (SSSA method), the Optimal Damper Placement for Minimum Transfer Functions (Takewaki method), and the Fully-stressed Analysis/Redesign method (Lavan A/R method).
Effectiveness of the techniques for distributing linear fluid viscous dampers for two building examples was evaluated under a suite of twenty ground motions, two seismic hazard levels, and in terms of peak interstorey drifts, absolute accelerations, and residual drifts using nonlinear time history analysis. The advanced methods showed comparable performance based on performance indicators. Therefore, usability is recommended as the selection criteria. The Lavan A/R method was found to be the most effective and usable method. It is recommended that multiple design ground motions be used for the SSSA Mode and Lavan A/R methods as well as caution against removal of upper storey damping, which prompts susceptibility to larger roof drifts due to higher-mode effects.
Various brace-damper arrangements were explored to determine strategic horizontal damper placement. It was found that brace-damper arrangements with diagonals and multiple bracedamper sets per floor pose effective means of distributing the axial damper force and protecting the lower-storey columns from overstressing. Behavioural testing of two nonlinear viscous devices was performed, and results were used to determine analytical models for the nonlinear fluid viscous and fluid spring devices based on fitted parameters. It was found that the stress-relaxation models better captured the nonlinearity of the devices than standard models but yielded only marginally decreased energy dissipation per cycle. Thus, it is recommended that standard models are adequate for analysis of damped structures with these nonlinear viscous devices.
Source: University of Oxford
Author: Jessica Whittle