Experimental Validation and Design Review of Wave Loads on Large-Diameter Monopiles
Master thesis
Permanent lenke
http://hdl.handle.net/11250/2350948Utgivelsesdato
2015Metadata
Vis full innførselSamlinger
- Institutt for marin teknikk [3469]
Sammendrag
The objective of the thesis work was to explore challenges related to ULS wave loading on offshore wind turbines, with special emphasis on large-diameter inertia-dominated monopile foundations.
Experimental studies testing ULS loads on a 6.9-m diameter pile in eight different three-hour sea states were performed in two water depths. The wave conditions were set to represent those at location 2 of Creyke Beck B at Dogger Bank, with specified sea state return periods up to 1000 years. The tested structure was a rigid smooth-surfaced bottom-fixed pile with a flexible rotational base stiffness. Wave elevation, pile-top acceleration, shear force and moment time realizations were measured. In addition, the runs were video-taped for deterministic studies.
A numerical model was developed in MATLAB, considering an idealized single degree-of-freedom rotational system with a rigid pile and flexible base stiffness. The program includes two general force models generating the Morison or the FNV excitation forces. The Morison model includes the option of running either shallow, intermediate or deep water, with or without Wheeler stretching. In addition, an impulsive force term is included, aimed at evaluating the impact of an individual spilling breaker. The FNV model is expanded to include the finite-depth vertical kinematic distribution and dispersion relation. This is inconsistent with the assumption of deep water in the derivation of the FNV formulae.
The model tests were shown to give less than a 1 % deviation in Hs for repeated sea states. The agreement in Hs between the calibration tests and tests with the model was within a -3 % to 3 % range for all sea states, whilst the discrepancy in nominal and effective Hs varied between a 12 % reduction to a 3 % increase for the calibrated waves. The greatest decrease in Hs occurred for the roughest sea state, indicating a large amount of breaking waves.
Throughout the tests an excessive first-mode motion was observed. The pile was almost constantly oscillating at its eigenfrequency, even in between large wave groups. The cause is thought to have been a combination of self-excitation from radiated waves reflecting off the wavemaker, the conservative mode shape and the low (but realistic) damping level.
Generally, the longer moment arm, i.e. the greatest water depth, generated the largest loads. The relative difference was smaller for Tp = 15 s than for Tp = 11.25 s, indicating the effect of increased wave non-linearity causing higher loads.
The largest response moments for almost every sea state were a result of a breaking wave impacting the structure. However, uncertainty remains regarding the excitation mechanism, as to whether it is caused by the impulse load or ringing induced by non-linear wave components.
The feasibility of the metocean conditions are questioned, due to the large amount of breaking for the roughest sea states. There is reason to believe that shallow-water effects at Dogger Bank are not properly considered. The relative propagation distance for an unstable energy-dissipating breaking wave before impacting a turbine could be significantly larger in real life than in the wave flume.
The FNV force gave the most conservative response moment values, the finite-depth version more so than the one for deep water. The Morison force gave unconservative responses for the roughest sea states. The FNV formulae implementation was validated, yet uncertainty regarding the linearity of the measured input waves still remains.