Effect of Fines Type and Content on Liquefaction Resistance and Ground Failure Potential

Clemson University, Clemson SC

Investigators

Abstract

0218365, Charng H. Juang, Clemson University "Effect of Fines Type and Content on Liquefaction Resistance and Ground Failure Potential" The ability to accurately evaluate the liquefaction potential and the liquefaction-induced ground failure potential is important in the engineering planning and design of structures in seismic regions. This research project addresses two important issues in the mitigation of earthquake-induced liquefaction hazards, namely the effect of fines type and content on liquefaction resistance, and the liquefaction-induced ground failure potential. The first topic concerns the triggering of liquefaction, and the second addresses the consequence of liquefaction. With respect to the first topic, recent studies by numerous investigators have concluded that both fines content and fines type affect the liquefaction resistance of soils, but these affects are not fully accounted for in the existing simplified design methods. Data from recent earthquakes offers a rare opportunity to investigate these effects and to improve on the current state of the art on liquefaction triggering analysis. The second topic is more challenging, as reflected in the fact that no readily applicable site-specific model is currently available for evaluating liquefaction-induced ground failure potential. However, the number of case histories from recent earthquakes is now sufficient to establish empirical models for estimating site-specific ground failure potential. This project is collecting this additional case history data on liquefaction triggering and ground failure. This case history data, together with the existing databases, is used to improve the current state of the art on liquefaction triggering analysis, and to develop new empirical models for evaluating liquefaction-induced ground failure potential. Advanced data analysis techniques, including logistic regression analysis, mapping function approach based on artificial neural network, and Bayesian updating technique, are examined and utilized for developing the empirical models. Both deterministic and probabilistic models are being developed. Since the geotechnical engineering profession prefers the deterministic approach, the deterministic models will be developed, but equipped with suggested guidelines including risk-calibrated factor of safety. The probabilistic models are being developed for use in performance-based earthquake engineering, as well as for mapping liquefaction-induced ground failure.