About VELACS

Numerous constitutive laws have been proposed for the numerical analysis of dynamic induced liquefaction of soil materials. In order to be useful to the engineering practice, a numerical technique requires verification and validation by comparison of its predictions with observed full-scale in situ field performance. However, much data are unlikely to be obtained soon, because of the scale of the structures involved, the cost of testing and the low probability of having a particular instrumented geotechnical system subjected to design load. The situation is a common problem in geotechnical engineering and is exacerbated in earthquake engineering. Even if a large number of sites were instrumented to monitor liquefaction, there would be only a few cases of liquefaction recorded during actual earthquakes. Consequently, some form of model study is desirable to enable alternative analysis procedures to be checked and validated.

In this context, the VELACS (VErification of Liquefaction Analysis by Centrifuge Studies) project offers a good opportunity to verify the accuracy of various analytical procedures. The numerical predictions were intended to be ``class A'' predictions, and thus were made before the relevant experiments were performed. The verification and validation of the various analysis procedures were to be carried out by comparing their predictions with the measurements recorded in the centrifuge experiments (in terms of excess pore water pressure, acceleration and displacement time histories). However, it should be pointed out that to this date no firm experimental verification and validation of centrifuge test results with the corresponding prototype situations have been made.

A centrifuge is used to simulate gravity-induced stresses in soil deposits at a reduced geometrical scale through centrifugal loading. Conceptually, the technique consists of increasing the confining environment in the model soil, so that the confining stress is identical in both model and prototype at homologous points. The technique allows soil liquefaction tests to be performed at a conveniently reduced scale and provides data applicable to full-scale problems. A program of dynamic centrifuge model tests consisting of nine geotechnical models constructed from or founded on liquefiable soil deposits have been performed within the framework of the VELACS Project as shown in the following figure (from Prevost and Popescu (1996)):

Centrifuge geotechnical models used for the VELACS project
(the dimensions are at the prototype scale)