| The current state of practice for predicting the liquefaction potential of soils relies on Standard Penetrometer Test (SPT) or Cone Penetration Test (CPT) results that are converted into a measure of liquefaction resistance through the use of design charts and empirical formulas. This empirically based methodology, used worldwide is based on statistical analysis of various case studies and provides simple and user friendly tools for practicing engineers. The methodology, however, is limited to relatively simple geometric configurations and boundary conditions.;Analysis of liquefaction and its consequences in more complex geostructures that contain liquefiable soils usually requires the use of more advanced simulation methods that are based on finite element or finite difference analysis techniques along with a constitutive model for the soil stress-strain-strength behavior in monotonic and cyclic loading conditions. Calibration of these advanced constitutive models depends on thorough and reliable experimental data. This work provides a detailed study into the characterization and stress-strain-strength properties of Ottawa F-65 sand. The experimental data includes key characterization tests as well as hydraulic conductivity, monotonic triaxial, and cyclic triaxial tests. A trial calibration is then performed using the experimental data for the constitutive model proposed by Dafalias and Manzari (2004). The results of a centrifuge experiment are compared to an analysis performed using current state of practice predictive techniques. The shortcomings of the state of practice are highlighted. |
