SharePoint
Aide
Faire avancer la sûreté nucléaire

La Recherchev2

Publications

Non linear site response: PGA distribution from observed and simulated borehole data.


Fermer

Authentification

Email :

Mot de passe :

L.F. Bonilla (1), F. Cotton (2) and C. Berge-Thierry (1) (1) Institut de Radioprotection et de Surete Nucléaire (2) LGIT Université Joseph Fourier European Geophysical Society XXVII General Assembly - Nice, 21-26 avril 2002.

Type de document > *Congrès/colloque

Mots clés > séismes, séismes

Unité de recherche > IRSN/DEI/SARG/BERSSIN

Auteurs > BERGE-THIERRY Catherine, BONILLA Fabian

Date de publication > 14/11/2002

Résumé

In the last years, the deployment of the Kiknet borehole network in Japan has allowed to record valuable data for différent magnitudes and distances. The earthquake engineering community, on the other hand, is interested in knowing the PGA distribution at surface for différent ground motion intensity levels and différent site conditions. The possible feedback between the incident wavefield and the soil medium may lead to nonlinear site effects. For this reason in the last thirty years there has been a great effort to compile strong motion data from boreholes worldwide in order to have a statistical relation between the PGA downhole and the PGA at the surface. These relationships may indicate the acceleration level at which nonlinear effects begin. In addition, they may be used to take into account nonlinear effects in probabilistic seismic hazard scenarios. Moreover, borehole data are also useful to develop attenuation laws for the incident ground motion, which may be used for nonlinear site response scenarios as well. We have compiled the last 2 years borehole data from the Kiknet network in Japan, which include the Tottori M7.1 earthquake. These data show that nonlinear effects begin around 0.1 g (decrease of the surface PGA with increasing downhole PGA). Nonetheless, these effects are not so strong as predicted by previous relationships (e.g. Idriss, 1990). Accelerations recorded on soils are still higher than the ones recorded on rock even at high levels of motion. In addition, from strong ground motion scénarios in Southern California (Archuleta et al., 2000) where synthetized downhole motions were propagated through soil columns with nonlinear properties, we obtain synthetic PGA distributions at the surface as well. Finally, strong nonlinear effects due to dilatancy in sandy materials may produce a high PGA at the surface compared to what is expected from the statistical relations mentioned previously.