Generate floor response spectra, Part 2: Response spectra for equipment-structure resonance

Abstract

When generating floor response spectra (FRS) using the direct spectra-to-spectra method developed in the companion paper, probability distribution of t-response spectrum (tRS), which deals with equipment-structure resonance or tuning, corresponding to a specified ground response spectrum (GRS) is required.In this paper, simulation results using a large number of horizontal and vertical ground motions are employed to establish statistical relationships between tRS and GRS. It is observed that the influence of site conditions on horizontal statistical relationships is negligible, whereas the effect of site conditions on vertical statistical relationships cannot be ignored. Considering the influence of site conditions, horizontal statistical relationship suitable for all site conditions and vertical statistical relationships suitable for hard sites and soft sites, respectively, are established. The horizontal and vertical statistical relationships are suitable to estimate tRS for design spectra in USNRC R.G. 1.60 and NUREG/CR-0098, Uniform Hazard Spectra (UHS) in Western North America (WNA), or any GRS falling inside the valid coverage of the statistical relationship.For UHS with significant high frequency spectral accelerations, such as UHS in Central and Eastern North America (CENA), an amplification ratio method is proposed to estimate tRS.Numerical examples demonstrate that the statistical relationships and the amplification ratio method are acceptable to estimate tRS for given GRS and to generate FRS using the direct method in different practical situations.