Validation of Simulated Ground Motions for Bridge Engineering Applications-A Preliminary Study
- Author(s): Habchi, Rachelle George
- Advisor(s): Zareian, Farzin
- et al.
The study presented herein provides validation of five methodologies (EXSIM, GP, Irikura-Recipe, SDSU, and SONG) for ground motion simulation. The 1994 Northridge earthquake is selected alongside a bridge model as the test bed for this validation. This validation comprises a comparison between Engineering Demand Parameters (EDPs) obtained from Nonlinear Response History Analysis (NLRHA) of the bridge model subject to simulated and recorded motions of Northridge. The significance of the differences between the two EDPs are correlated with Intensity Measures (IMs) of the simulated and recorded motions. The IMs considered are normalized Arias intensity (Ia), Significant Duration (Td), time at mid-duration (tmin), rate of energy accumulation (Ia/Td), rate of change in predominant frequency (’), and predominant frequency at mid duration (mid). A regression of logged recorded parameters yielded Ia, Td, and mid as significant IMs. A regression analysis of logged recorded ground motions discovered that ’ was also significant when considering multiple ground motion angles from 0 to 180 at 9˚ increments. A mixed effects regression was performed to establish the influence of simulation realization, ground motion station, and angle of rotation possessed on the model. Statistical distributions of the recorded and simulated IMs and EDP were compared to each other. Regardless of simulation methodology, the mean and variance of the simulated EDP were comparable to that of the recorded. It is recommended that the methods employed in this study be used with various recorded earthquakes in future research to assess the practicality of this validation approach.