Use of Cyclic Element Tests to Assess Scale Models

Large scale laboratory model tests have been performed to aid the design of offshore gravity platforms. The paper shows that undrained conditions, necessary for correct model simulation, cannot be achieved if the full field loading program and remolded site material are used. It is theoretically possible to shorten the time of the model event by increasing cyclic load levels and reducing the number of cycles or increasing the frequency or using a dynamically weaker leaner model clay, or by a combination. Cyclic triaxial data are presented for three remolded clays of differing plasticities tested at various frequencies. The results show that for all practical purposes, even if gross model modifications are considered, similarity between model and field events is impossible. It is concluded that laboratory model tests are only indirectly helpful. They may be used to investigate a foundation's response in its weakest fully softened state or to assess a design technique by applying it to a model subjected to a brief arbitrary storm event.