Progress and Challenges in Validation of Simulated Earthquake Ground Motions for Engineering Practice

Abstract:

A fundamental assumption in seismic response simulations of structures is the approximation of possible seismic excitations by a set of selected and modified ground motions recorded from past events world-wide. The assumption that records from past earthquakes can be used for analyses of structures in other locations with different seismic characteristics has several limitations, which can be overcome by the use of simulated ground motions. Recently, there have been significant improvements in the methods and required infrastructure to simulate ground motions. Due to such advancements, much attention has been given to simulation and validation research; hence, the engineering community is gradually moving towards embracing the use of simulated ground motions.

This presentation summarises progress and challenges in validating simulated ground motions for engineering practice. Significant challenges in validation of simulated ground motions include the diversity of engineering applications, validation methods, and the lack of consensus on the acceptable accuracy for the estimation of structural response. We focus on recent validation efforts that can be divided into three main types. Type I Validation uses statistical criteria to show if the ground motion simulation methods can replicate waveforms of past events and affect structures in the same way that the natural recordings of the same events do. Type II Validation methods compare the variation of significant ground motion intensity measures (IM) with respect to event parameters. In Type II Validation methods, scalar or vector IM are used as surrogates of structural response and utilized for validation; the closeness in trends for simulated and recorded motions is used as a measure for validation. Lastly, Type III Validation investigates the use of simulated ground motions as a database of records from which engineers can select and scale records for structural response simulations. In Type III Validation, a ground motion simulation method is judged by the statistical similarity of structural response parameters obtained from sets of simulated and recorded motions anchored to a target spectrum. The talk will end with a discussion on remaining validation challenges.