Project engineer, GeoSyntec Consultants, Huntington Beach, CA
University distinguished professor, Virginia Tech, Blacksburg, VA
Professor, University of California, Berkeley, CA
An evaluation of the use of centrifuge modeling as a tool for analyzing the behavior of reinforced soil slopes is presented in this paper. A review of the state-of-the-art indicates that previous centrifuge studies have focused mainly on the performance of reinforced soil vertical walls and that limit equilibrium approaches (used in the design of reinforced soil slopes) have not been fully validated against the failure of models in a centrifuge. As part of an evaluation of the conditions of similarity governing the behavior of reinforced soil structures at failure, scaling laws are specifically derived by assuming the validity of limit equilibrium. It is demonstrated that an Nthscale reinforced slope model should be built using planar reinforcements having 1/N the strength of the prototype reinforcements in order to satisfy similarity requirements. A description of the experimental testing procedures implemented as part of a recent centrifuge testing program is presented, and an example dataset from this investigation is used to illustrate typical results. These include the g-level at failure, visual observation of failure development, and post-failure analysis of reinforcement breakage. The pattern observed in the geotextile reinforcements retrieved after testing indicates that the boundary effects were negligible.
Paper ID: GTJ10413J