Non-linear analysis of jack-up structures subjected to random waves
There is a steadily increasing demand for the use of jack-up units in deeper water and harsher environments. Confidence in their use in these environments requires jack-up analysis techniques to reflect accurately the physical processes occurring. This thesis is concerned with the models appropriate for the dynamic assessment of jack-ups, an important issue in long-term reliability considerations. The motivation is to achieve a balanced approach in considering the non-linearities in the structure, foundations and wave loading. A work hardening plasticity model is outlined for the combined vertical, moment and horizontal loading of spudcan footings on dense sand. Empirical expressions for the yield surface in combined load space and a flow rule for prediction of footing displacements during yield are given. Theoretical lower bound bearing capacity factors for conical footings in sand have been derived and are used in a strain-hardening law to define the variation in size of the yield surface with the plastic component of vertical penetration. The complete incremental numerical model has been implemented into a plane frame analysis program named JAKUP. The spectral content of wave loading is considered using NewWave theory, and the importance of random wave histories shown by constraining the deterministic NewWave into a completely random surface elevation. Using this technique, a method for determining short-term extreme response statistics for a sea-state is demonstrated. A numerical experiment on an example jack-up and central North Sea location is shown to emphasise the difference in long-term extreme response according to various footing assumptions. The role of sea-state severity in the variation of short-term extreme response statistics is also highlighted. Finally, probabilistic methods are used to develop further understanding of the response behaviour of jack-ups. A sensitivity study of influential variables (with probabilistic formulations as opposed to deterministic values) has been conducted using the response surface methodology.