3.4 Principles and Scaling of Centrifuge Modelling

Owing to high cost, inconvenience and complexity, extensive field instrumentation studies are not suitable for detailed study of the behaviour of deep excavations. Laboratory study is a feasible approach to overcome these difficulties. However, small scale model tests under 1g condition would be at a much lower effective stress level than the actual prototype. Since soil behaviour is highly stress dependent, the test results of a small model would be quite different from the corresponding prototype. To overcome the shortcomings mentioned above, centrifuge modelling technique has been employed to test small scale model (1/Nth-scale) under high gravity (Ng) so that prototype stress levels are reproduced. Thus a centrifuge model is a reliable means to simulate the behaviour of a prototype.

Since 1970, some centrifuge tests for excavation have been conducted by a number of researchers (Lyndon and Schofield, 1970). From these early centrifuge studies, it has been established that the centrifuge technique is a powerful tool to simulate the excavation process and to evaluate the ground deformation due to excavation.

Table 3.2 summarises the scaling relations of various parameters between the model and the prototype (Leung et al., 1991).