Scaling and Selection of Ground Motion Records for Nonlinear Response History Analysis of Structures

Accurate estimation of structural response for a given target hazard level requires a suitable set of ground-motion records that represents a pre-defined scenario event. With the aim of assessing the seismic performance of structural systems, this study presents a ground motion selection and scaling methodology. Given a relatively larger ground-motion database, the method uses the differences between individual records and corresponding estimations of a ground-motion prediction equation in order to determine the optimum subset of recordings. This way, the procedure provides a group of modified records whose median elastic spectral ordinate matches with the target intensity level without excessively manipulating the inherent aleatory variability in the selected recordings. Considering the sensitivity of overall structural response to the nonlinear behavior, the procedure approximately estimates the median and dispersion of inelastic structural response. The case study suggests that, the methodology can be a useful tool for building performance assessment studies