Water-saturated sandy soils are susceptible to liquefaction when subjected to strong ground shaking, such as occurs during an earthquake. Liquefied soils experience a decrease in bearing strength: this can lead to the partial or total collapse of buildings standing on liquefied soils. The aim of this proposal is to verify the analyses used to predict liquefaction due to earthquake loading. Since the field data needed for this verification is extremely difficult to obtain, data is being used instead from laboratory experiments performed on geotechnical centrifuges. This is a relatively new technique for soil testing which is becoming well accepted in the geotechnical research community. The technique for simulating earthquakes on centrifuges is being developed, and the results obtained have generated confidence in the method: questions that remain to be answered are being addressed in this study. Different geotechnical structures, including sloping ground, dams, coastal dikes, and embankments are being investigated so as to validate the applicability of the liquefaction analysis methods to a variety of geotechnical structures. Class A predictions - which involve predicting the soil response without knowing the measured results - are used for later comparison with measured results. The Earth Technology Corporation is responsible for the management of the project, for providing carefully calibrated soil samples, for overseeing the testing program, as well as being responsible for receiving, distributing and documenting data, in conjunction with the Adjudication Committee. It is also responsible for managing the prediction studies. The Adjudication Committee acts as an independent steering group. The participating institutions are the University of California at Davis and Berkeley, the California Institute of Technology, the University of Colorado, the Massachusetts Institute of Technology, Princeton University, and Cambridge University, England. Collectively, they are examining the mechanisms of liquefaction, the development and verification of analytical and numerical models, soil-structure interaction for conditions of limited and large deformations, and the effect of interfaces between soil layers of differing permeabilities. This study will contribute significantly to improving the understanding of the soil liquefaction phenomenon, and to the reliability of methods used to predict the occurrence of liquefaction and the resulting damage to structures. This action is for support of the program at the University of Colorado at Boulder.
