| description abstract | The study of the liquefaction phenomenon and its consequences has been of permanent research interest in geotechnical earthquake engineering due to the wide recognition of its significant damage potential. However, despite the constant emergence of new case histories of liquefaction from recent earthquakes, the incorporation of these data has not led to substantial changes in semiempirical methodologies, generating considerable uncertainties about their performance on underrepresented data. Accordingly, the aim of this study is to evaluate the performance of liquefaction models that are less dependent on available databases. Existing correlations derived from calibration chambers and cyclic tests from the literature were revisited to develop a new liquefaction triggering model based on the cone penetration test with pore water pressure measurement based (CPTu-based) state parameter approach. When testing the models on three liquefaction databases, namely, Global, Canterbury, and a new subduction zones database, the proposed approach demonstrated better predictive performances compared with most state-of-practice models. Additionally, when using a purely state parameter–based approach, the estimated cyclic resistances (CRR7.5) in dense sands were considerably lower than those predicted by the state-of-practice models, which is consistent with recent developments. Finally, given the significant similarities between the proposed approach and the cyclic strain approach, the advantages of the former are discussed to enhance CPTu-based models, including the one proposed herein. | |
