Modification of stochastic ground motion models for matching target intensity measures

Abstract

SummaryStochastic ground motion models produce synthetic time-histories by modulating a white noise sequence through functions that address spectral and temporal properties of the excitation. The resultant ground motions can be then used in simulation-based seismic risk assessment applications. This is established by relating the parameters of the aforementioned functions to earthquake and site characteristics through predictive relationships. An important concern related to the use of these models is the fact that through current approaches in selecting these predictive relationships, compatibility to the seismic hazard is not guaranteed. This work offers a computationally efficient framework for the modification of stochastic ground motion models to match target intensity measures (IMs) for a specific site and structure of interest. This is set as an optimization problem with a dual objective. The first objective minimizes the discrepancy between the target IMs and the predictions established through the stochastic ground motion model for a chosen earthquake scenario. The second objective constraints the deviation from the model characteristics suggested by existing predictive relationships, guaranteeing that the resultant ground motions not only match the target IMs but are also compatible with regional trends. A framework leveraging kriging surrogate modeling is formulated for performing the resultant multi-objective optimization, and different computational aspects related to this optimization are discussed in detail. The illustrative implementation shows that the proposed framework can provide ground motions with high compatibility to target IMs with small only deviation from existing predictive relationships and discusses approaches for selecting a final compromise between these two competing objectives. | INTRODUCTIONThe growing interest in performance-based earthquake engineering 1,2 and in simulation-based, seismic risk assessment approaches 3,4 has increased in the past decades the relevance of ground motion modeling techniques. These techniques describe the entire time-series of seismic excitations, providing a characterization appropriate for dynamic time-history analysis. UndoubtedlyThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Hazard‐compatible modification of stochastic ground motion models

“…This formulation was first introduced in Scherbaum et al, with Vetter et al offering a computationally efficient approach for performing the associated optimization, leveraging surrogate modeling principles. Recent work by the authors extended this framework by addressing a critical shortcoming, the fact that physical descriptors of the resulting acceleration time‐series were incorporated in the optimization merely as constraints, something that required significant experience in ground motion characterization for proper definition of the optimization problem. The shortcoming was addressed by introducing a bi‐objective optimization problem, transforming the aforementioned constraint for the physical characteristics of the resultant ground acceleration time‐series to an explicit objective.…”

“…Recent work by the authors extended this framework by addressing a critical shortcoming, the fact that physical descriptors of the resulting acceleration time‐series were incorporated in the optimization merely as constraints, something that required significant experience in ground motion characterization for proper definition of the optimization problem. The shortcoming was addressed by introducing a bi‐objective optimization problem, transforming the aforementioned constraint for the physical characteristics of the resultant ground acceleration time‐series to an explicit objective. The proposed tuning was performed for specific seismicity scenarios and identified the ground motion model that achieves the minimum modification of the existing predictive relationships that will yield the desired compatibility with the target IM.…”

“…The current study extends approach to (1) match the prescribed conditional hazard (not simply mean IMs) for a specific site and structure (or range of structures) while (2) preserving desired trends and correlations in the physical characteristics of the resultant ground acceleration time‐series, including consideration of the variability of these characteristics. This is again formulated as a bi‐objective optimization problem.…”

Performance‐based seismic design of tuned inerter dampers

“…The advantage of using the model in Equation is due to the fact that it preserves the realistic non‐Gaussian character of earthquakes, a feature usually not considered in other models. () The non‐Gaussianity of real earthquakes is supported by the results in Figure a, which show that the kurtosis coefficient as a function of v s 30 for the ground motions in the NGA‐West data set is greater than 3, the characteristic value for Gaussian processes. Figure b shows the probability distribution function of the kurtosis κ for the type‐C National Earthquake Hazard Reduction Program (NEHRP) soil, similar to the soil assumed for this study.…”

An earthquake-source-based metric for seismic fragility analysis

“…The simplified seismological model used for the purpose of this paper may be replaced by other simple stochastic (Tsioulou et al 2018) or more complex physics-based (Goda et al 2016) models that can produce synthetic processes as functions of (m, r). The general character of the proposed methodology allows for the use of ground-motion records produced by complex seismological models, that account for types of seismic sources, directivity, local-amplification conditions, such as the point-source model SMSIM (Boore 2005), the finite-fault model EXSIM (Motazedian and Atkinson 2005), or hybrid models (Seyhan et al 2013), that combines physics-based and stochastic models for the low and high frequency contents, respectively.…”