TOWARDS THE DEFINITION OF REFERENCE MOTIONS (1000 ≤ VS ≤ 3000 M/S): ANALYSIS OF THE KIK-NET DATA AND CORRECTION OF THE LOCAL SITE EFFECTS - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2017

TOWARDS THE DEFINITION OF REFERENCE MOTIONS (1000 ≤ VS ≤ 3000 M/S): ANALYSIS OF THE KIK-NET DATA AND CORRECTION OF THE LOCAL SITE EFFECTS

Résumé

A key scientific component in Seismic Hazards Analysis (SHA) is the assessment of a local hazard for hard rock sites (1000 < V S30 < 3000 m/s), either for applications to installations built on this site category, or as a reference motion for site effect computation. Within the context of SHA, empirical ground motion prediction equations (GMPEs) are the traditional basis for estimating shaking intensities and V S30 , the time-averaged shear-wave velocity in the upper 30 meters from the surface, is the basis to account for site conditions. The current GMPEs, however, are not well constrained for V S30 larger than 1000 m/s (only a few records on high V S30 sites are included in the main accelerometric databases). The presently used approach is based on host-to-target adjustment techniques based on V S30 and κ 0 values. This study is investigating alternative methods to estimate reference motions on site effect free, very hard rocks (1000 < V S30 < 3000 m/s). We explore methodologies to obtain a prediction for reference motions (1000 < V S30 < 3000 m/s) by using the " rock " Japanese KiK-net sites with 500 < V S30 < 1350 m/s. Each site presents the advantages of having sensor pairs (one at the surface, and one installed in a borehole at depth between 100 and 200 m for most sites and up to 2000 m) and geotechnical characterization (P-and S-wave velocity profiles) for surface and down-hole sensors. Firstly, the " rock " transfer functions are estimated in two ways: empirically (spectral ratios between surface and depth records) and theoretically (linear SH1D simulation). These two approaches are compared to validate the input parameters and also to select the stations for which the 1D approximation is verified. Then, two new accelerometric datasets characterizing hard rock sites (1000 < V S30 < 3000 m/s) in free surface condition are developed: 1. Down-hole recordings are modified from within motion to outcropping motion with the depth correction factor developed by Cadet et al. (2012), 2. Surface recordings are deconvolved from site-specific effects with a (surface / outcrop rock) amplification factor derived with the site velocity profile and 1D simulation. GMPEs with simple functional forms are then developed for each dataset, with a site term based on V S30 (assumed to be equal to V S at downhole sensor depth) and the results are compared, for a specific scenario, to the result obtained with the traditional host-to-target adjustment approach: our hard-rock GMPEs lead to significantly lower estimates at short periods.
Fichier principal
Vignette du fichier
4936.pdf (563.67 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01458360 , version 1 (06-02-2017)

Identifiants

  • HAL Id : hal-01458360 , version 1

Citer

Aurore Laurendeau, Pierre-Yves Bard, Fabrice Hollender, Olga-Joan Ktenidou, Laetitia Foundotos, et al.. TOWARDS THE DEFINITION OF REFERENCE MOTIONS (1000 ≤ VS ≤ 3000 M/S): ANALYSIS OF THE KIK-NET DATA AND CORRECTION OF THE LOCAL SITE EFFECTS. Sixteenth World Conference on Earthquake Engineering, Jan 2017, Santiago, Chile. pp.4936. ⟨hal-01458360⟩
369 Consultations
251 Téléchargements

Partager

Gmail Facebook X LinkedIn More