Geotechnical centrifuge modelling at the University of Western Australia

The main requirement for model testing in any area of Engineering is to ensure that key dimensionless groups are equivalent in model and prototype. In fluid mechanics, an area where dimensional analysis was pioneered, there are a number of such groups (Reynolds number, Froude number and so forth) which determine the performance of the prototype, and which need to be matched correctly in any small-scale modelling of the prototype. In solid mechanics, for problems where the self-weight of the material is important, the primary dimensionless group is the ratio of self-weight stress to the strength of the material. This principle needs to be taken even further in geotechnical research, where the response of soil and rock depends critically on the past and current stress state. Under these circumstances, it becomes necessary to simulate the full prototype stress field in any model experiment, distinguishing carefully between the ambient stress state prior to any perturbation, and the stress changes that take place subsequently during the experiment.

It is now widely accepted that centrifuge modelling provides the most versatile technique for obtaining stress conditions that are homologous in model and prototype. Other techniques – such as using a downward hydraulic gradient – may also be used, but none of these offers the same scope as centrifuge modelling. For element tests, or where the variation of stress through the event to be modelled is small (such as in modelling of a very deep tunnel, or the stress changes around an advancing cone), it is sufficient to use triaxial cells or larger so-called ‘calibration chambers’ where the boundary stresses are controlled. However, in problems where not just the ambient stress, but the gradient of ambient stress is important (such as an embankment on soft clay, or the performance of foundations on sand), centrifuge modelling becomes the optimum approach.

This paper describes the centrifuge facility that has been developed in the Department of Civil and Environmental Engineering at The University of Western Australia, and also presents brief descriptions of the current research projects being undertaken using the facility. The paper concludes with a summary of additional research areas where centrifuge modelling can play an important role both in fundamental research and in site-specific design studies.