Time-domain fatigue response and reliability analysis of offshore wind turbines with emphasis on welded tubular joints and gear components
MetadataShow full item record
- Institutt for marin teknikk 
Presently fixed support structures are mainly used in the offshore wind industry, e.g. monopile, tripod, gravity base, et al., where the water depth are usually less than 30 m. Research work is ongoing for deeper water depth. Jacket substructure is very competitive in moderate water depth like 40-100m, due to its light weight, high stiffness and efficient construction. Due to the harsh offshore environments, the fatigue performance of welded connections in jacket support structure is therefore a design driving criterion for successful operation of jacket wind turbines. In addition, as one of the most expensive components of the wind turbine system, gearbox has experienced higher-than-expected failure rates in the wind energy industry. In order to refine the design and make a better balance between the target safety levels and the project costs, and hence increase the long-term reliability of gearbox, more reasonable probabilistic design methods are desired. Compared with the jacket platforms used in the offshore oil and gas industry, the dynamic response analysis of jacket wind turbine subject to stochastic wind and waves loads simultaneously is more complicated. In this thesis, a decoupled procedure is applied for the dynamic response analysis of a fixed jacket offshore wind turbine (NREL 5 MW) designed for a North Sea site in a water depth of 70m in time domain. The time series of each of the nominal stresses in 4 different tubular joints with respect to geometry and location in the jacket support structure for each shortterm environmental condition (wind/sea state) are obtained separately, and then superimposed together. The hot-spot stress ranges at the critical hot-spot locations of each joint are obtained by using rainflow counting method. The long-term probability distribution of the hot-spot stress ranges due to wind and wave loads is investigated. The contributions to fatigue damage from wind loads, wave loads and the interaction effect of wind and wave loads are identified with consideration of three different load cases: wind loads only, wave loads only and a combination of wind and wave loads, from a long-term point of view. Strength degradation due to the effect of harsh offshore environment, e.g. corrosion, has an influence on the fatigue reliability of tubular joints since the service life of offshore wind turbines is usually taken as 20 years. In order to maintain safety of offshore wind turbines in service life with respect to fatigue, wear and other deterioration phenomena especially, inspection, monitoring and repair are important measures. In this thesis, a general simplified probabilistic corrosion model is presented and used to modify the long-term hot-spot stress ranges, which can be used to account for the influences of the increased hot-spot stress range due to changes of nominal stress and stress concentration factors (SCFs) produced by the thickness thinning effects of the reference brace, the planar brace and the chord of tubular joint due to corrosion wastage. The fatigue reliability of tubular joints considering the effects of corrosion and inspection is investigated. The main uncertainties in the whole time domain analysis procedure are identified and analyzed. The sensitivity of the reliability index on important random variables is estimated. Another problem discussed in this thesis is time domain based contact fatigue analysis of gears in wind turbine drivetrain. Three practical problems in time domain modeling and analysis of dynamic gear contact force in wind turbine gearbox with respect to fatigue assessment are discussed: (1) torque reversal problem at low wind speeds conditions, (2) statistical uncertainty due to time domain simulations and (3) simplified postprocessing method for long term gear contact fatigue analysis. A simplified predictive pitting model for estimating gear service lives is presented and validated by comparing with the published experimental evidence. The model is applied to estimate the contact fatigue lives of the sun gear and the planetary gears in the drivetrain system of a 750 kW land-based wind turbine under stochastic dynamic conditions. The long-term probability distribution of the maximum gear contact pressure at a certain contact point is investigated. Furthermore, A frame work for reliability-based probabilistic gear contact fatigue analysis for onshore and offshore wind turbines is presented. A gear contact fatigue limit state function is established. Gear contact fatigue reliability analysis and sensitivity study are performed.