Modelling, Analysis and Response-Based Operability Assessment of Offshore Wind Turbine Blade Installation with Emphasis on Impact Damages
MetadataVis full innførsel
- Institutt for marin teknikk 
Installing wind turbine blades is a highly demanding task due to their large lifting height and the requirements of high precision. The single blade installation method – where individual blades are lifted using a jack-up crane vessel – is currently the most popular method, owing to its efficient deck space area utilisation during transportation and cost-effectiveness. However, one of the issues related to this method is the dynamic motion responses developed in the blade while being lifted, which can cause impact of the lifted blade with the surrounding structures. The impact load on fibre composite wind turbine (WT) blades is critical and can induce complex, simultaneously interacting and visually undetectable damage modes and has a high potential to reduce the ultimate and fatigue strength of the blade. An assessment of such impact-induced damages is, therefore, instrumental when planning blade installation tasks since blades are not ideally designed against impact loads during transportation and installation. The present thesis is concerned with impact loads and damages to the blade during installation and develops methods where the damage assessment results can be combined with the global responses of the installation systems to estimate response-based limiting sea states for blade installation. Given that the rated capacity of OWTs is continuously increasing, with 8 MW turbines currently in operation, the case studies presented in this thesis are based on a future class of OWTs featuring a power rating of 10 MW; in this thesis, investigations are performed on a DTU 10 MW reference wind turbine. Standard industrial procedures are considered, and the blade installation task is divided into three distinct sequential sub-operations: sub-operation 1 describes the lift-off phase, sub-operation 2 corresponds with the lifted phase, and sub-operation 3 represents the mating phase. Critical impact scenarios for the blade are identified for each of the sub-operations, and a risk assessment chart is prepared to rank their criticality. The blade root impact with hub during the mating process (sub- operation 3) is ranked the highest, followed by the leading-edge impact with turbine tower during the lifted phase (sub-operation 2). Coupled multibody simulation codes are developed to analyse global motion responses in the installation system corresponding to each sub-operation, and impact velocities are analysed for the blade damage assessment. The numerical model accounts for blade aerodynamics, monopile hydrodynamics, pile-soil interaction, structural dynamics, and mechanical couplings to include lifting arrangements. The results indicate that while being lifted, the blade can impact the surrounding structures with low velocity; nevertheless, impact energy during the collision of the blade is high since a blade of a typical MW-sized WT weighs tens of tons. In the thesis, systematic numerical methods are developed for damage assessments of WT blade and emphasis is placed on efficient finite element (FE) models. For sub-operation 2, where the blade impacts the turbine tower with its leading-edge while being lifted, an energy-based mesoscale approach is adopted and damage assessments are performed on a global scale, local cross-section scale and coupon scale. The results from the global impact analysis of the blade indicate that only a fraction of the impact energy is absorbed in the blade as damage, while the majority is dissipated as recoverable elastic strain energy by means of rigid-body motions. An equivalent impact velocity is derived and damage assessment on the local contact area of the blade cross-section is performed using high fidelity FEM techniques. Three different numerical methods – (1) pure shell, (2) shell-to-solid coupling and (3) submodelling – are compared, and it is observed that the sub- modelling based global-local method is the most efficient technique. Furthermore, a series of experiments are performed on coupon scale adhesive joints with varying bondline thickness; in these, the highest impact resistances under impact loads is indicated by the joints having the least bondline thickness. For cases where the WT blade root impacts the hub during blade mating phase (sub-operation 3), a homogenisation-based macroscale approach is adopted. Two distinct impact scenarios – head-on impact and sideways impact – of the blade root guide pin with the hub are analysed, and the effects of wind-wave misalignment conditions on the impact velocities and subsequent damages are studied. It is observed that the collinear wind-wave conditions are the most critical load cases and induce sideways impact of the blade root guide pin with the hub. This causes permanent deformation in the steel guide pin and damage in the adjoining root laminate. Probabilistic response-based methodologies are developed for estimating operational limits for the blade installation task for impact under normal as well as accidental loads. Under the normal load based-assessment approach, the impact scenarios are always assumed to occur and two distinct approaches are proposed. These are the short-term sea state approach – in which the sea state parameters are assumed stationary – and the long-term sea state approach – where the average failure probability of the installation task is calculated by considering the distribution of the sea states for a given offshore site. It is found that the operational limit obtained using the short-term sea state approach is conservative in general, with operational limits that can further be extended using the long-term sea state approach. Also, the safe domain for the mating task is found to increase with increasing misalignment between wind and wave conditions. On the other hand, the accidental-load based assessment approach is proposed for impact scenarios that can occur at any time of the lifting operations due to factors such as human error, system failure etc. Lastly, a mitigation measure in the form of passive tuned mass damper is proposed for use during blade root mating operations and its feasibility studies are conducted. The device is found to increase the damping ratio of the monopile system in the fore-aft bending mode from 1% critical to 5.6%; thereby reducing impact velocities and increasing the domain of allowable sea states for the mating task. However, the effectiveness of the TMD increases with increasing Hs but reduces with increasing misalignment between wind and wave conditions, and with Tp shifting away from the tuned frequency of the monopile system.