Floating Offshore Wind Turbine (FOWT) unlock far-offshore wind resources in deep waters that can’t be harvested under economic aspects using bottom-fixed wind turbines. Numerical modelling tools are employed to assess different FOWT designs under various environmental conditions. In order to be competitive and yet guarantee working designs, the numerical tools need to be reliable and computationally efficient. This justifies the need for assumptions simplifying the models.

One such simplification is the rigid floater assumption, where the FOWTs’ substructure is assumed rigid. This reduces the computational effort but at the same time alters the results, like the tower’s first natural frequency and corresponding mode shape.

Multiple approaches are used in literature to match the tower’s first natural frequency of flexible and rigid floaters. This involves adjusting the tower properties, such as length or Young’s modulus, or alternatively, implementing a flexible element between a rigid floater and a flexible tower. Siemens Gamesa currently makes use of the latter method by tuning the flexible elements’ properties to achieve a match in the models’ tower first bending natural frequency.

So far no studies have been conducted on the impact of the flexible element parameters on the tower first mode shape when tuning towards a matching tower first bending frequency. Additionally, the effect of differently correct mode shape variants on the tower dynamics is investigated. This leads to the goal of improving the tower dynamics for a FOWT with a rigid substructure.

The analysis was based on two versions of the U-Maine FOWT model. One fully flexible floater design served as a reference, whilst a fully rigid floater design was used to incorporate the different flexible element designs. Various flexible elements with distinct properties were evaluated to understand the sensitivity of the mode shape to these parameters. Subsequently, selected designs exhibiting varying degrees of accuracy of the mode shape were compared to the flexible floater design in time domain simulation. Furthermore, three separate methods of identifying the tower’s first bending mode are proposed.

In the course of modelling the flexible floater design, a modelling error was made that resulted in double counting of the heave motions. Despite this error, it was concluded that for constant and turbulent wind, all flexible element designs outperform the rigid floater design. Furthermore, a close mode shape match likewise results in an increased match of bending moment and tower top rotation for high and low wind speeds. In the range of rated wind speeds, the shortest flexible element design with the worst mode shape match performs best. Comparing the tower top acceleration also indicates an overall improvement of the results, but less significant. This is expected to result from tuning towards the first tower mode rather than higher-order modes.

Generally, using any flexible element design already results in an improved mode shape match with minor differences. The impact of these discrepancies on the tower dynamics is small. Therefore, it is concluded that any flexible element, even when only tuned to match the tower’s first natural frequency, is an improvement over the rigid floater design.