SPATIAL VARIABILITY OF EARTHQUAKE GROUND MOTION FROM DENSE-ARRAY OBSERVATIONS AND 3D NUMERICAL SIMULATIONS

Abstract eng:
The evaluation of spatial variability of earthquake ground motion plays a crucial role in several applications of earthquake engineering, typically for the seismic analysis of spatially extended structures, such as bridges, viaducts and pipelines. Regarding the design of bridges, seismic norms, namely, Eurocode 8 - EC8 and the 2008 Italian Building Code - NTC08, recommend to estimate the spatially variable nature of the seismic action by accounting for the relative displacement (Dij) between adjacent points along the structure. The latter is determined from empirical relationships as a function of the Peak Ground Displacement (PGD). In this contribution we present an investigation on the spatial variability of earthquake ground motion, based on both observationally- and numerically-based approaches. First, spatial coherence is computed as a function of both separation distance and frequency for different earthquake scenarios, making use of the results from 3D physics-based numerical simulations in different areas worldwide. This numerical dataset is useful to study the behavior of spatially variable ground motion on “virtual” dense arrays with desired separation distance, for a variety of source features (focal mechanism), site-to-source configurations and geologic conditions, with special emphasis on the near-source region of large earthquakes, where recorded data are very scarce. Results suggest that spatial coherence at small separation distances tends to be smaller than that given by empirical models, owing to the complex interaction of the extended source, propagation path and local topography/site effects. Second, the accuracy of the empirical relationships provided by seismic norms is reviewed and tested against the Dij/PGD correlation obtained using worldwide strong-motion observations from dense seismic arrays (see Fig. 1). It turns out that the EC8 and NTC08 formulas tend to envelope the observations on the conservative side for large separation distances, while at short distances, EC8 is not conservative. Finally, we found that the dataset from 3D numerical simulations is in reasonably good agreement with observations, showing a moderate dependence on source-to-site distance, with largest values of normalized relative displacement in near-source conditions.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket