The growing awareness of the importance of the dynamic behavior of gravity type deep water platforms is illustrated with a chronological description of the dynamic analyses of the Dunlin "A" platform.
The following different stages can be recognized:
calculation of the dynamic amplification for a one degree of freedom system. This simplification appeared to be too rough especially for the upper part of the structure.
transient response analysis for a multi degree of freedom system. The results of the fatigue calculation were unsatisfactory because the uncertainties of the input parameters were amplified by the dominant influence of the resonance waves.
spectral dynamic fatigue analysis. In this approach the uncertainties are less and their influence is reduced.
The designs of most of the offshore platforms that by now have been placed in the North-sea are based on a static concept although external loadings like waves and wind obviously have a dynamic character.
In most cases the use of a static design concept for a dynamic loading is the result of the, in civil engineering, well established method of simulating the dynamic loading by a static equivalent or even, turning a blind eye to the dynamic character and pretending that the loading is static.
This very simple way of avoiding the still difficult problems of dynamics appeared to be an adequate method for the design of the jacket type offshore platforms as they were designed for shallow, moderate and even for the deep waters of the northern North-sea. A rule of thumb that is often used is that for offshore structures with natural periods below 2.5 to 3 sec. there is no danger of resonance.
For the gravity type deep water platforms the situation is different and the simple approach as mentioned above can no longer be applied the static stresses multiplied by the appropriate dynamic amplification factors showed that fatigue problems for some areas were critical. In fact if the fatigue damage was calculated in the purely deterministic way of dividing the total number of expected waves with a certain height and number of occurences, it governed the design at some positions in the deck structure. This is caused almost entirely by the waves in the resonance area.
The basis for a fatigue calculation in steel constructions is a stress range and a number of cycles; if these parameters cannot be trusted with reasonable certainty, the results of the fatigue calculation have to be very critically reviewed. It will be shown that for a dynamicly sensitive structure and a deterministic approach this is the case because the degree of accuracy of the parameters is not known even by approximation.
It will be clear to the reader that all these analyses were not carried out before the construction had been started, this being impossible with such short delivery times.