Today, many offshore structures in the North Sea already reached their predicted lifetime. Since it is still required a huge demand of oil, it results in an important need to keep those structures in operation. The great attention concerning the lifetime of offshore platforms has trigged a need for monitoring these structures in order to gain information about their actual state and hence reduce the uncertainty and allow for more optimal decision planning regarding maintenance, repair and future inspection actions.
Throughout the lifetime, the performance of the structure can be evaluated by analyzing the deterioration process of the structure. In the offshore environment, one of the most common deterioration mechanisms is the fatigue of structural steel induced by wave loading. The deterioration formulation of a structural system subjected to fatigue is nowadays well known. However, many uncertainties may affect the accuracy of the performance evaluation. It can be mentioned mainly the uncertainties related to the materials, the uncertainty on Miner’s rule and the uncertainty on the SN curve but most importantly is the uncertainty on the stress ranges induced by the wave loading.
In this paper, the mainly focus is on the uncertainty observed on the different stresses used to predict the damage. This uncertainty can be reduced by Modal Based Fatigue Monitoring which is a technique based on continuously measuring of the accelerations in few points of the structure with the use of accelerometers known as reliable for long time measurements. An Operational Modal Analysis (OMA) is performed and then a modal filtering of the operating response is considered, so that the modal coordinates of all significant modes are known. Next, the experimental mode shapes are expanded using a Finite Element (FE) model together with the Local Correspondence (LC) principle and the displacements can be estimated in all degrees of freedom of the FE model, allowing the stresses and strains to be obtained from the element equations. It is important to emphasize that even though the accelerations are measured in only a few points of the structure, the stress history can be calculated in any arbitrary point of the structure.
The accuracy of the estimated actual stress is analyzed by experimental tests on a scale model where the obtained stresses are compared to strain gauges measurements. After evaluating the fatigue stresses directly from the operational response of the structure, a reliability analysis is performed in order to estimate the reliability of using Modal Based Fatigue Monitoring for long term fatigue studies.