| dc.description.abstract | Floating Offshore Wind Turbine (FOWT) is a fairly new concept. There are limited
number of full-scale prototypes to provide real data. Therefore, most of the research today
is reliant on numerical models. This is required, so that adequate amount of confidence
can be gained before venturing into large scale production. The major challenge ahead is
proving their reliability and robustness. There need to be supporting studies that consider
most factors that can go wrong. The computer program FAST was a groundbreaking
contribution from National Renewable Energy Laboratory in this regard. FAST is capable
of doing an integrated loading analysis of FOWTs. However, the numerical model used
for the hydrodynamics can be improved further.
Non-linear hydrostatic and wave forces on floating structures become very important
during large amplitude waves and motions. SIMDYN, a blended time domain computer
program developed by Marine Dynamics Laboratory, Texas A&M University, is capable
of capturing the role of such forces. SIMDYN has previously been used to demonstrate
that nonlinear hydrostatics become very important in the problem of parametric excitation.
In the current work, SIMDYN is coupled with FAST. FAST-SIMDYN is now a tool that is
capable of studying large amplitude motions of FOWTs in extreme seas.
FAST-SIMDYN was then used to study the classic instability of negative damping.
This phenomenon occurs in FOWTs that use conventional land based control. The development
of platform pitch and surge instability are studied in relation to different wave and
wind scenarios. The objective of this project was to do an analysis to see if non-linear
hydrodynamic forces do play a significant role in large amplitude motions induced by
negative damping. Such a study would give an idea on whether the development of more
sophisticated hydrodynamic modules is justified. This study indicated that non-linear hydrostatic
and non-linear Froude Krylov forces resulted in higher platform pitch and heave
motions compared to that obtained using a corresponding linear analysis. A bifurcation
was also identified, which was induced by the controller algorithm. The system motion
response depended on the wind speed history more than what was expected as a result of
the bifurcation. This was used to explain the large motions obtained near the controller
algorithm transition region. | en |
