Performance of Seismically Isolated and Non-Isolated Steel-Framed Buildings: Sensitivity to Amount and Form of Inherent Damping

Abstract

The amount of inherent damping assumed in nonlinear response history analysis of steel buildings is typically set at 2% or less for the fundamental mode of vibration. However, many studies have shown that actual damping levels vary depending on the building characteristics and soil conditions, and methodologies used in measurements of damping in the field. Reported inherent damping values range from 1% to more than 5% for the fundamental mode of vibration. This study investigates the effects on the computed seismic performance of the assumed level and form of inherent damping in nonlinear response history analysis, focusing on seismically isolated and non-isolated buildings with special moment-resisting and concentrically braced frames. The seismic isolation systems considered are sliding friction pendulum type. The findings demonstrate that the assumed value of inherent damping has an impact on the computed floor accelerations, affecting acceleration-sensitive non-structural components, particularly with periods less than 1 s. Collapse probabilities of isolated buildings are minimally affected by the assumption of inherent damping, leading to simplifying modeling for collapse-focused analyses. Comparative studies involving conventional non-isolated buildings reveal significant sensitivity to inherent damping values across various metrics, including floor accelerations, peak story drift ratios, residual drift ratios, and collapse probabilities. It is shown that non-isolated building models exhibit reduced sensitivity of collapse probabilities and floor spectral accelerations when inherent damping is specified using different methods—specifically, capped viscous damping as compared to modal damping. This highlights that while the performance of seismically isolated buildings shows relatively small sensitivity to the model of inherent damping, non-isolated buildings exhibit notable differences.