Evaluation of Liquefaction Ejecta Potential from Case Histories and Insights from Nonlinear Dynamic Analyses

High-quality data from 24 benchmark liquefaction case histories from Christchurch (New Zealand) and data from 123 global case histories were used to develop a family of empirical relationships for estimating the severity of liquefaction ejecta manifestation. The proposed empirical model provides means to develop liquefaction manifestation charts and estimate ejecta quantities (i.e., area covered by ejecta and ejecta-related settlement) as a function of the intensity of the earthquake excitation. The model is applicable to all types of liquefiable soil deposits within conventional liquefaction assessment using the cone penetration test. The principal factor discriminating between different ejecta potential is the soil deposit type. Deposits comprising loose liquefiable soils but interbedded with nonliquefiable soils have very limited ejecta potential. Such deposits have never exceeded 15% ejecta area coverage, based on current evidence. Conversely, deposits composed mostly of liquefiable soil show very high potential for extreme ejecta manifestation (up to 100% ejecta area coverage) when comprising loose critical layers (qc1Ncs(CL)≤110), though the severity of ejecta manifestation for such deposits substantially decreases with the increase in density of the critical layer (i.e., qc1Ncs(CL)). Results from evaluation of the model performance using 6,200 liquefaction case histories and recommendations for application of the model to engineering practice are presented and discussed. Results from nonlinear dynamic analyses are presented to provide further insights on the liquefaction response of deposits and their potential for ejecta manifestation.