Seismic Response of Acceleration-sensitive Nonstructural Components Mounted on Moment-resisting Frame Structures

Abstract

A statistical analysis of the peak acceleration demands for nonstructural components (NSCs) supported on elastic and inelastic regular moment-resisting frame structures is presented. The response of a variety of stiff and flexible frame structures (with 3, 6, 9, 12, 15, and 18 stories) subjected to a set of 40 far-field ground motions are evaluated. The NSCs under consideration are those that can be represented by single-degree-of-freedom systems with masses that are small as compared to the total mass of the supporting structure. The study evaluates and quantifies the dependence of peak component accelerations on the location of the nonstructural component in the structure, the damping ratio of the component, and the properties of the supporting structure such as its modal periods, height, stiffness distribution, and strength. The results show that current seismic code provisions will not always provide an adequate characterization of peak component accelerations especially when the period of the NSCs fall in the higher modal period region of the supporting structure and the provisions do not address the inelastic action of the supporting structure. A parameter called as acceleration response modification factor (<em>R_acc</em>) is proposed to quantify the reduction in component amplification factors and inelastic FRS that is achieved due to the inelastic behavior of the building. A methodology that makes use of the <em>R_acc</em> factor to estimate the acceleration demands on NSCs mounted on inelastic supporting structures from that of elastic buildings is outlined. Separate <em>R_acc</em> factors are proposed for long-period, fundamental-period and short-period regions of the FRS at three different locations in the building namely roof, mid-height, and bottom-third location. A comparison of the proposed <em>R_acc</em> factors to that of results obtained from real multi-bay buildings show that the recommendations fall within 20% error range for both fundamental-period and short-period regions of FRS.