For variable speed wind turbines it is possible to perform both a power control and a safety function using centrifugally controlled tips. Present mechanisms (screw-cylinder/spring type) are expensive and maintenance critical. A possible solution to these problems can be achieved with a so-called TenTorTube (Tension-Torsion Tube): a tube with inherent coupling between tension and torsion.
The objective of the project is to obtain reliable predictions on performance and stability and information on costs and applicability of the configuration by developing and testing a full-scale TenTorTube tip mechanism.
The initial proposed design of the TenTorTube was a long slender tube made of aramid fibre reinforced epoxy with an inherent coupling between tension and torsion. During the first phase of the project it became clear that the fatigue properties of aramid/epoxy were dramatically bad due to a creep-like behaviour. Tests on carbon/epoxy tubes showed a considerable increase in design strength, but not in torsional stiffness.
Results of non-linear prediction of aeroelastic stability and performance showed poor stability and power limiting capabilities, mainly caused by the low torsional stiffness. Wind tunnel tests were carried out using a 2 degree of freedom model with an active control loop to simulate the TenTorTube properties like stiffnesses, damping etc. The wind tunnel tests confirmed the results obtained by the aeroelastic stability predictions.
The torsional stiffness requirement cannot be met with the initially specified wind turbine configuration (250 kW - 2 bladed). An alternative (conceptual) design was generated, but was rejected. A parameter study showed that for 3 bladed turbine configurations with an increased ratio between maximum rotor speed and rated rotor speed, a feasible TenTorTube design was possible. Tower fatigue loads will however be larger which will result in more expensive towers for medium-sized wind turbines.
Resulting from the promising test results on carbon/epoxy tubes, flexible resin systems have been evaluated. A more flexible resin in combination with a T300-type of fibre has been selected with improved fatigue strength and acceptable elastic properties.
Based on the promising material properties and the aeroelastic predictions it was decided to continue the project by designing, producing and testing full-scale tubes. The design of the tubes resulted in a realistic tube length (roughly 3 m). Unfortunately the tube manufacturer did not succeed in producing tubes according to spec within the time scheduled, so the tubes could not be tested and evaluated.