Comparison of empirical seismic lateral spread displacement models based on probabilistic geotechnical hazard curves developed for Canada

Abstract

Case histories have reported significant damage to structures during earthquakes due to permanent ground displacements (PGDs) arising from liquefaction-induced lateral spreading. The empirical model by Youd et al. (2002) is classical and has been widely adopted in practice over two decades owing to its simplicity. However, the suitability of this model for M>8 earthquakes is at stake, as it is largely based on datasets with M<8 earthquakes. Zhang and Zhao (2005) proposed another empirical model to compute PGDs that accounts for the underlying mechanism and different tectonic source type responsible for earthquake of a given magnitude; accordingly, the outcomes from this model differentiates between crustal and subduction earthquakes. The authors in their recent work developed probabilistic lateral spread displacement hazard curves by implementing Youd et al. (2002) and Zhang and Zhao (2005) models using the Openquake opensource platform, where the users are provided with a choice of using hazard curves based on the model of their preference. Based on the initial comparison of the hazard curves, it has been observed that both the models show similar trends for all the geographic locations considered in the study. However, the annual probability of occurrence reported by Zhang and Zhao (2005) model were found to be significantly higher than those predicted by Youd et al. (2002), for a given level of expected lateral spread displacement. This apparent conservatism of the predictions based on Zhang and Zhao (2005) motivated the authors to perform a comparative assessment in order to understand the underlying reasons for the differences in the outcomes and comment on the applicability of the models. In view of this, Openquake analyses were performed considering some selected geographic locations in Canada, and the findings from this work are presented herein.