The paper describes a series of model tests carried out with a Tension Leg Platform (TLP) in both long- crested and short-crested seas. Results from some of the tests are compared with predictions from computer simulations.
The experiments point at a considerable reduction in the total energy for the main response modes in short-crested seas compared to long-crested seas. A similar reduction can also be observed for the mean drift force. The extreme values of the responses (maxima and minima) do not, on the other hand, seem to be significantly affected by the directionality of the waves.
An interesting problem in marine technology is towhat extent the motion response of an offshore structure is influenced by the short-crestedness of ocean waves. So far a great deal of the theory developed for evaluating both first order forces, and in particular mean drift and slow drift forces, have been based on the assumption of long-crested waves.
The objective of this paper has been to identify and investigate certain problem areas by carrying out a comparative study between the response in short-crested and long-crested seas, for a deep water TLP.
The paper is mainly experimental. However, to render some insight into the possible inadequacy of present day theory, results are, to a certain degree, compared with predictions from the TLP computer program TERICA, developed by The OTTER Group in Trondheim, Norway, on behalf of CONOCO Norway Inc.
The experimental findings indicate a notable percentage wise reduction in the mean drift and slow drift motion in the principal direction of the waves. No similar trends are found for peak values, although, it might be that, in this case, the simulation periods (approx. 50 minutes, full scale time) are a little too short to give reliable results.
To give an exhaustive presentation of the computer program TERlCA is far beyond the scope of this paper. A fair description is given in [1], and a more detailed account can be found in [2]. However, for the sake of completeness, the relevant features are listed below:
2-dimensional potential theory (strip theory) is employed in order to calculate first order transfer functions for horizontal pontoons and vertical columns. Interaction effects between different members are neglected.
Maruo's formula [3] is used for calculating mean drift forces, Newmans approximation [4] is then applied to obtain slow drift.
For directional seas an angular averaging is first carried out for first order transfer functions and drift coefficients, the resulting response is then calculated as for long crested seas.
First and second order motions are added in the time domain, and the statistical analysis (giving extreme values etc.) is then carried out for the combined time series.
That 2-dimensional theory (versus 3-dimensional) is in general sufficient is established in [1]. The most critical and questionable point here is the averaging method used for obtaining modified drift force coefficients in directional seas.
