The project was a feasibility study on the use of flexible blades to increase the durability and survivability of tidal turbines. The highly turbulent flow experienced by tidal turbines leads to continuous load variations limiting the fatigue endurance. It was proposed to use flexible blades in order to control the flow field around the blades and allow constant stresses on the turbine’s blades and shaft. Under the effect of a variation of the onset flow velocity at a certain location along the blade span, the flexible blade will change the local angle of attack and the foil shape. The new flow field around the blade section will be as such to minimise the load change on the blade section. The reduced, if not completely avoided, load variations will enhance the fatigue endurance of the blade and the shaft.
1. The unsteady flow fluctuations experienced by a tidal turbine blade may result in dynamic loads more than 20% higher than for steady conditions, i.e. if the turbine operated in a uniform tidal stream with a constant velocity (Tully and Viola, ISROMAC 2016).
2. The load fluctuations are maximum at the tip of the blade, and when the tidal stream velocity is maximum. Decreasing the tidal stream velocity, or considering sections closer to the root, the amplitude of the angle of attack variations increases but the load fluctuations decrease.
3. Low frequency turbulent eddies and long period waves cause the most significant load fluctuations on a tidal turbine blade (Scarlett and Viola, OTE 2016).
4. If frequencies lower than two per revolution were filtered by, for instance, a pitch control system, then load fluctuations would decrease by one order of magnitude (Scarlett and Viola, OTE 2016). Further research is ongoing to determine the effect of yaw misalignment, which was neglected when the above conclusion was drawn.
5. It is unlikely that the pitch control system will be capable to match frequencies much higher than two per revolution. Therefore, in low current speed and steep waves, the turbine will experience high- frequency large angle of attack variations due to turbulence, wave and tower shadow (Scarlett and Viola, OTE 2016). These will result in small load fluctuations, because of the low tidal speed, but in periodic trailing edge separation - non dynamic stall - associated with loss of efficiency of more than 40%, averaged over a cycle (Tully and Viola, ISROMAC 2016).
6. In these conditions, Theodorsen’s method, which is widely used in industry as a predictive model for unsteady loads, may predict load fluctuations more than double that which was measured (Viola et al., OTE 2016). This highlights the need for advanced predictive tools that predicts the occurrence of non-linear effects and take these into account for the computation of the dynamic loads.
7. A flexible trailing edge can be used to mitigate load fluctuations by more than 30% and to increase efficiency by more than 25%, preventing large trailing edge separation (Viola et al., OTE 2016; Tully and Viola, ISROMAC 2016).