Session: SE-07-01 Seismic Evaluation of Systems, Structures and Components

Paper Number: 125200

125200 - Exploration of the Effects of Response Spectrum Matching on Power Spectral Density Functions 

Seismic acceleration time histories have been widely used in seismic soil-structure-interaction analysis, nonlinear analysis of complex structural systems, shake table tests, advanced fragility analysis, among other engineering tasks. Except for situations where actual recorded seismic ground motions are deemed appropriate, these time histories are often developed based on input response spectra (RS) or design RS (DRS) that cover many potential earthquake scenarios. One of the popular approaches is to perform RS matching, in which a recorded seed time history or a random seed time history is modified so that the resultant acceleration time history will have a computed RS close to the DRS by meeting certain acceptance/convergence criteria.

Most response spectrum (RS) matching algorithms do not explicitly consider the power spectral density (PSD) functions of the generated acceleration time histories, and the PSD checks are usually conducted separately. U.S. Nuclear Regulatory Commission Standard Review Plan (SRP), Section 3.7.1, Revision 4, “Seismic Design Parameters,” indicates that the RS-matched acceleration time histories produce PSD functions that are quite different in appearance from one individual function to another, and that fluctuate significantly and randomly as a function of frequency. The SRP also recognizes that as one tries to closely envelop the DRS, the PSD functions tend to fluctuate more significantly and randomly. The fluctuations may lead to unconservative results for the response of structures, systems, and components. This paper provides examples on how RS-matched acceleration time histories behave differently regarding their PSD functions, based on a frequency domain method and a newly developed time domain wavelet-based method. These examples will serve as visual cues to reinforce that RS matching criteria alone are not sufficient and that checking the PSD functions of the resultant acceleration time histories is a necessary part of the ground motion synthesis. SRP Section 3.7.1 Revision 4 provides detailed guidance and the associated procedures for PSD checks. 

Presenting Author: Jinsuo R. Nie U.S. Nuclear Regulatory Commission

Presenting Author Biography: Dr. Jinsuo Nie is a structural engineer at U.S. Nuclear Regulatory Commission. He performs research and technical reviews mainly in the seismic and structural areas and his technical interests also include geotechnical engineering, computational mechanics, fragility, and probabilistic risk assessment. Before joining the NRC, he was a scientist at Brookhaven National Laboratory. He also had multiple years of consulting experience in structural analysis and design.

Dr. Nie serves on the ASCE Nuclear Standards Committee and ASCE DANS Committee. He is also a member of the working group for ANS 2.36.

Authors:

Jinsuo R. Nie U.S. Nuclear Regulatory Commission