Testing Nonlinear Amplification Factors of Ground-Motion Models

ABSTRACT We explore nonlinear site effects in the new Japanese ground-motion dataset compiled by Bahrampouri et al. (2020). Following approach of Seyhan and Stewart (2014), we evaluate decrease soil amplification according to increasing corresponding ground motion on surface rock (VS30=760 m/s). To better predict associated with each record, take into account between-event variability motion, impact nonlinearity, correct observed site-specific linear amplification. Instead grouping stations site-response proxy, focus individual several strong-motion records. develop a framework test recently published models against model compare results recent building codes that include nonlinearity. The show response varies greatly from site, indicating conventional proxies, such as VS30, are not sufficient characterize response. Out all Kiban–Kyoshin network stations, 20 selected having recorded data be used test. these five signs is, performed than linear-amplification for periods T. For most sites, however, get best score. This suggest that, range predicted considered this study (peak acceleration <0.2g), nonlinearity may have sufficiently large justify use terms functional forms seismic moderate-level shaking.