Investigation into the resonance behaviour of the internal water column of an open-ended monopile

More Info
expand_more

Abstract

The offshore wind industry is increasingly constructing wind turbines farther from the coast, in deeper water, and under more extreme conditions. This requires larger (monopile) foundations and necessitates new installation methods. An important factor affecting workability is the dynamic behavior of the monopile during installation.
The objective of this thesis is to develop a method for determining the hydrodynamic loads caused by internal sloshing in an open-ended monopile (MP) as it transitions from a horizontal to a vertical position in the splash zone.
First, the resonance frequencies of the internal water column are predicted with analytical approximations based on linear theory. Distinction is made between piston mode and sloshing. Two numerical methods, linear potential flow (LPF), and computational fluid dynamics (CFD) are used to verify the resonance frequencies. Since the CFD analysis is done in 2D, a 2D representation of the open-ended monopile is considered. Due to the presence of viscous effects in CFD, the resonance observed with CFD consistently occurs at a lower frequency than for the analytical and LPF methods. Also it is found that an decrease in inclination angle of the monopile with respect to the horizontal, while maintaining the same submerged length, results in lower resonance frequency for both piston mode and sloshing in both LPF and CFD.
To assess the accuracy of LPF in describing the motion of the internal water column, it is compared to the CFD model. Input excitation in the CFD model is low enough to avoid non-linear sloshing modes and other non-linear behaviour of the free-surface. The ComFLOW 2D CFD model has been validated against various works from the literature for the accurate representation of gap resonance frequencies.
For both the piston mode and sloshing resonance, discrepancies between the two numerical methods are found, which can be attributed to viscous effects. At resonance viscous effects are nonnegligible, therefore the LPF method over-predicts the severity of the piston mode and sloshing. The influence of both the submergence of the monopile and its inclination angle with respect to the horizontal is considered.
The hydrodynamic coefficients for added mass and damping are found with forced oscillation for both upright and inclined geometries. While good agreement is found between the LPF and CFD results away from resonance, the CFD results in the vicinity of the resonance frequency are used to tune the LPF model, by way of additional linear damping, to achieve more accurate results.
It can be concluded from the present work that the resonance of the internal water column near the first sloshing mode significantly affects the overall hydrodynamic force and must be taken into account. At the peak hydrodynamic force observed during the first sloshing mode, the sloshing induces forces 5.59 times higher (submergence of 5 meters), 3.62 times higher (submergence of 10 meters), and 2.33 times higher (submergence of 15 meters) compared to cases where sloshing is not considered.
Looking forward, it is strongly advised to conduct forced oscillation tests with larger amplitudes, as this explores the effect of non-linear chaotic sloshing. Additionally, expanding the CFD analyses to 3D, where more non-linear sloshing effects are expected, such as swirling, is recommended. Furthermore, given the differences in results between LPF and CFD, it is valuable to validate the findings through model experiments.