Adjustment of relative amplification factors using kappa2 correction vs. using an empirical lower bound amplification: the case of IFMIF-DONES site (Granada, Spain)

Abstract

The IFMIF-DONES project will carry out the design and construction of a scientific facility near Granada (Spain), whose purpose is irradiation of materials with a neutronic spectrum similar to what is obtained within a nuclear fusion reactor. In the framework of the seismic hazard assessment for the IFMIF-DONES site, this paper compares two approaches to correct the overdamped high-frequency response of soil columns computed via regular equivalent-linear analyses. The site has a soft soil profile, with a V_s30 around 375 m/s. For soft soils, when introducing site effects in seismic hazard assessments, equivalent-linear analyses are known to overdamp high-frequency responses. This may result in unrealistically small relative amplification factors (RAFs) with respect to the host profile response, which is the reference in Ground Motion Prediction Equations. In this paper the overall methodology for derivation of RAFs is presented, based on equivalent-linear analyses, and two approaches to RAF correction are described: an empirical lower bound on the RAFs, taken from accepted practice, and the so-called kappa2 correction of the Fourier Amplitude Spectra. For small ground motions, differences in RAFs computed by the two methods are minimal, since soil degradation is limited. For higher-severity events, significant differences appear beyond 8 Hz. Two empirical RAF lower bounds, 0.5 and 0.6, were tested. The results for the IFMIF-DONES site suggest that the 0.6 lower bound provides a good average fit to the results obtained using kappa2 correction. For the stronger motions, the 0.5 lower bound provides a better fit in the 2.5–10.0 Hz band.