Predicting the Pit-to-Crack Transition of Offshore Wind Turbine Foundations

Abstract

Offshore wind turbines are designed to persist harsh environmental conditions, including corrosion. Pitting corrosion is one of the most dangerous forms of corrosion. In combination with a fatigue loading, a corrosion pit could result into the failure of a structure, due to a minimal weight reduction. Of the corrosion fatigue process, the transition of a corrosion pit to a fatigue crack is one of the less understood topics. In this thesis, a combined analytical/numerical model is presented that implements mechanisms of both the fatigue, and the corrosion process to determine the moment in time and location in the corrosion pit of the pit-to-crack transition. The hypothesis used to build the model, is that the pit-to-crack transition occurs at the moment that a fatigue crack initiates. Assumptions are made to be able to predict the level of stress and strain at the surface and below the surface of a corrosion pit using a FEM model. Using the stress and strain levels, damage is accumulated until the failure criterion is reached, i.e. when a fatigue crack will initiate. In this research study, elliptical pit shapes are considered, with and without a certain roughness at the surface of the corrosion pit. From the results it is found that a more sharp shaped corrosion pit is a more favourable location for fatigue crack initiation, than a more blunt shaped corrosion pit. The number of cycles to fatigue crack initiation resulting from the developed model are higher, than the number of cycles to failure found by the DNV GL B1 SN-curve for free corrosion conditions. Also, the DNV GL SN-curve is more steep, than a fitted curve through the data points resulting from the developed model. Through a sensitivity study and by considering the number of assumptions and simplifications made in this research study, it is concluded that the spread and uncertainty of the outcome of the developed model is too large to be able to draw conclusions based on the comparison between the results and the DNV GL SN-curves. More research on the input parameters of the developed model will give more certainty and a better validity to the model. It is found that the effect of surface roughness to the fatigue crack initiation life is not substantial, because it affects the stress and strain levels only up to a relatively small depth. The location of fatigue crack initiation is in the pit bottom for blunt shaped corrosion pits, and in the pit wall for sharp shaped corrosion pits. For circular shaped corrosion pits, the difference between the fatigue crack initiation life at the pit wall and the pit bottom is negligible. It is concluded that the model gives a comprehensive basis for the prediction of the pit-to-crack transition, but more (experimental) research on the input parameters will be necessary to be able to give a substantiated estimate of structural life of offshore wind turbine foundations.