9503673 Bardet This is an extension to the VELACS (Verification of Liquefaction Analysis by Centrifuge Studies) project, the purpose of which was to evaluate the relevance and effectiveness of the existing numerical and constitutive models used in the analysis of soil liquefaction. The motivation is to resolve the difficulties encountered in the Class A predictions of VELACS, and to isolate the main cause of these difficulties. VELACS revealed that the current procedures used for the analysis of geotechnical structures are not suitable, in general, for predicting the consequences of soil liquefaction. Some of the most promising procedures available for the analysis of soil liquefaction and its consequences will be identified in this project. A series of carefully-performed triaxial and torsional tests are to be conducted and carefully examined, so as to provide a small but reliable data base. These data will be used for the calibration of constitutive models, which in turn will be used for a Class A prediction exercise of two centrifuge model experiments. The predictions will be compared with the experimental results, and reasons for potential discrepancies will be discussed. Multiple checks for reliability will be conducted in an effort to isolate reasons for the lack of reliable predictions of the experiments already considered in the VELACS project. ***
