Seismic site response analysis

Sponsors: Rutherford Discovery Fellowship (RSNZ), Marsden Fund (RSNZ), Natural Hazards Research Platform (NHRP), NZ Earthquake Commission (EQC)

Collaborators: Jim Kaklamanos, Eric Thompson (Tufts University), Misko Cubrinovski (UC), Chris McGann (UC)

Postgraduate researchers: Saumya Das, Jawad Arefi

Accuracy and precision of 1D equivalent-linear site response analyses based on 100 KiK-net downhole array locations

Site response, or local site effects, are well recognised to have a profound influence on surficial ground motions as a result of the variability in surficial geology, and the complexity in its characterisation, particularly for nonlinear site response.

This work has two main thrusts:

The first is an improved quantification of the efficacy of site response analysis methods via their validation based on downhole array recordings within a formal framework (Bradley 2011). This validation framework can be used for any form of site response analysis method. To date, the framework has been applied using 100 KiK-net downhole arrays in Japan to assess the accuracy and precision of linear and equivalent linear analysis methods, as a function of various parameters (Kaklamanos et al, submitted). The results, summarised in the right panel of the above figure, illustrate the peak shear strains for which linear, equivalent linear, and non-linear analysis methods are preferable (as a function of the spectral period of interest). This methodology is intended to be applied to nonlinear total stress and nonlinear effective stress site response methods in the near future.

The second focus of site response analysis work is improving the characterisation of site response effects in Christchurch. This is principally driven by the objective an improved understanding of the ground motions which have been observed in the region from the 2010-2011 Canterbury earthquakes. This work is utilizing CPT/SPT, Vs, SWS and H/V ratios to document the characteristics of strong motion stations in Christchurch, as well as using equivalent linear, total stress and effective stress nonlinear analyses to illustrate the responses at these sites. The results of this work will lead to an improved understanding of site response effects in Christchurch, and improved characterization of these effects via simplified site classification. The results in Christchurch will eventually be extended to other major urban areas in New Zealand.

Relevant Publications:

1. Bradley BA. A framework for validation of seismic response analyses using seismometer array recordings. Soil Dynamics and Earthquake Engineering 2011 Vol 31 No. 3, pp 512-520.
2. Bradley BA. Ground motions observed in the Darfield and Christchurch earthquakes and the importance of local site response effects. New Zealand Journal of Geology and Geophysics (in press).
3. Kaklamanos J, Bradley BA, Thompson EM, Baise LG. Critical Parameters Affecting Bias and Variability in Site Response Analyses Using KiK-net Downhole Array Data. Seismological Society of America Annual Meeting 17-19 April, 2012. San Diego, US. (abstract).
4. Bradley BA, Araki K, Ishii T, Saitoh K. Effect of ground improvement geometry on seismic response of liquefiable soils via 3-D seismic effective stress analysis. Soil Dynamics and Earthquake Engineering, 2013. Vol. 48, pp 35-47.