Floating offshore wind turbines present a great potential for harnessing the power of offshore wind and meeting future energy demands. Though some floating offshore wind farms have already been commissioned, research on floating offshore wind turbine platforms needs to be pursued, in particular for the purpose of cost reduction. In this context, the ability to conduct scaled hydrodynamic testing of floating offshore wind turbine platforms is an important advantage for a wave tank. The COAST Laboratory of the University of Plymouth aims to provide this possibility. The wave tank being built before the decision to add a wind generation system had been made, the space available for the installation of a wind generation system in only limited, in particular without the use of a costly gantry to install axial fans at the location of where the wind is needed. This created the need for the investigation presented in this thesis on producing wind in the laboratory using centrifugal fans and a ducting system as well as an expanding diffuser. Flow distribution at the outlet of the wind generation system was determined both thanks to CFD in OpenFOAM and experimentally for various prototypes. The comparison of the results of 7 turbulence models to the experimental data provided by measurements done with the first prototype allowed the determination of 4 appropriate turbulence models in the context of indoor air flow, namely the standard k-ε, the k-ω SST, the realizable k-ε, and the RNG k-ε model. CFD simulations for larger prototypes using the 2 best turbulence model, that is, the k-ω SST and the realizable k-ε model, were done to predict the flow distribution of air coming out of a 33° angle conic diffuser with various structures inside and choose the best one to be built and studied experimentally. Structures inside the diffuser cone as well as honeycombs and meshed screens were able to help even flow distribution to a certain degree. However, the asymmetry of the flow caused by the 35° angle with which the air flow arrived at the beginning of the conic diffuser was too significant to be evened out. The laboratory wind generation system was used with a model version of the NREL 5 MW wind turbine at scale 1:50, using blades with low Reynolds number aerodynamic profiles. In spite of the asymmetry of the flow distribution, it showed that the model could be used in Froude scaled environment with a wind speed only slightly higher than Froude scaled wind to produce correct thrust. The wind generation system in combination with the low Reynolds number wind turbine allowed to study in laboratory conditions wind speeds corresponding to 13 m/s at full size on a model wind turbine.

Document Type


Publication Date