View description

Session: Challenges for Model Extrapolations Outside of the Data Range [Poster]

Type: Poster

Date: 10/12/2023

Time: 07:00 AM

Room: Stanley Park Ballroom

Numerical simulations of the non-linear response of thick sedimentary deposits give extremely variable results depending on the maximum depth considered for a non-linear behavior of soils. For (non-exceptional) cases where the deep bedrock is hard enough to generate to significant (≥ 2) velocity contrast with the deepest sediments, the application of widely accepted degradation curves to the entire soil column generally leads to a significant reduction of high frequency ground motion even for relatively moderate shaking levels, because deformation levels are high enough to result in significant damping increase over the whole sediment thickness. Nevertheless, as most of these degradation curves have been developed from lab measurements performed under limited confinement pressure, their extrapolation to large depths (exceeding a few hundred meters) can be questioned.

This contribution will present a statistical analysis of strong motion recordings obtained on thick Japanese sedimentary sites from the KiK-net, K-NET and JMA (Shindokei) networks. From a qualitative viewpoint first, the distribution of velocity profiles for sites with very large pga (above 0.4 g) is similar to the distribution for all strong motion sites in Japan, without any bias towards rock or shallow, stiff ;sites. Second, the non-linear effects are quantified by comparing various kinds of site-response characteristics (surface-to-borehole spectral ratios, H/V spectral ratios or site-specific terms resulting from generalized inversion techniques) for moderate to large motion to their weak motion counterparts. While a non-linear behavior can be detected on thick and shallow sites as well, even for relatively moderate loading levels, thick sites do not exhibit any peculiar high-frequency reduction. The inconsistency of such results with the high frequency decrease predicted by most non-linear simulations warns on the caution to extrapolate non-linear degradation curves at large depths, or to use frequency-independent damping in linar equivalent computations.

Presenting Author: Pierre-Yves Bard

Additional Authors

Pierre-Yves Bard

Presenting Author Corresponding Author

pierre-yves.bard@univ-grenoble-alpes.fr

Université Grenoble Alpes, Grenoble, , France

Ziqian Wang

wzq.jddx@gmail.com

Disaster Prevention Research Institute, Kyoto University, Uji / Kyoto, , Japan

Jikai Sun

sunjikai00@gmail.com

Disaster Prevention Research Institute, Kyoto University, Uji / Kyoto, , Japan

Eri Ito

ito.eri.4x@kyoto-u.ac.jp

Disaster Prevention Research Institute, Kyoto University, Uji / Kyoto, , Japan

Hiroshi Kawase

kawasehir0sh1@gmail.com

Disaster Prevention Research Institute, Kyoto University, Uji / Kyoto, , Japan

Boumédiène Derras

boumediene.derras@univ-tlemcen.dz

University of Tlemcen, Tlemcen, , Algeria

Julie Régnier

julie.regnier@cerema.fr

Cerema, Sophia-Antipolis, , France

Items 1 - 30 of 139

Refine your search