Physical modeling of supported excavations

As a result of the complexities associated with analyzing and evaluating supported excavations, numerical modeling techniques are often employed. Often, there is a great deal of uncertainty associated with the validity of the assumptions and approaches made regarding numerical modeling. Physical models can be used to evaluate complex geotechnical systems whose performances are highly dependent on construction techniques, nonlinear soil-structure interactions, and variable geometries. Physical model tests can be used as alternatives to numerical models or as supplements. Data from physical model tests is often used to verify the validity of the various assumptions and approaches of the numerical models. The data can also be used to calibrate numerical models to the anticipated loading conditions. This paper presents an evaluation of several efforts that utilized physical modeling to investigate the behavior of excavation support systems and the associated ground deformations. Model test results are compared to the state-of-the-practice ground deformation prediction methodologies and to field observations. The information presented herein shows that scale model test data can be reliably extrapolated to equivalent prototype data. Thus, small scale laboratory tests at normal gravity, centrifuge tests, and small scale field tests can all used to evaluate excavation support system behavior and soil response associated with deep excavations. These findings are particularly important in that often full-scale field test cannot be performed. Copyright ASCE 2007.