I would like to further improve my understanding of ‘thrust-matching’: modifying a wind turbine’s (computational) model so the induced loads match a given set of thrust data of a similar turbine, both steady and dynamically and for both the operational and the idling case. An example of thrust matching was the modification of the NREL 5MW WTG for model tests at MARIN by Kimball et al. . They modified the aerofoil sections for the low Reynolds number on the scaled down turbine blades and increased the chord to properly represent the Froude scaled performance of the turbine. This seems quite an elegant but also a drastic way of changing the wind turbine. For the operational case already a lot can be accomplished by modifications of the controller settings: fine pitch and the wind-speed dependent target rotational speed that will be introduced in ROSCO 2.6.0.
However, for the idling case the matching needs to be accomplished by modifying the rotor geometry.
Suppose I would like to modify the NREL 5MW WTG OpenFAST model to decrease only the idling thrust by 10%, without (or as little as possible) affecting the operational thrust.
What would be the geometric properties of the rotor to change?
Thanks in advance and best regards,
Typically I hear the term “thrust matching” to indicate development of subscale physical model (e.g., for wind tunnel or wave basin tests) that has the same thrust characteristics as a full scale computational model. Is that what you are referring to?
During rotor idling, the blades are typically feathered to 90 degrees and the thrust load on the rotor is quite low, at least for cases with aligned flow. Thrust-matched rotors at model scale typically have thin, low Re number airfoils. My guess is the design of the hub may be one of the knobs that could be further tweaked to match a given idling rotor thrust. That said, given the low rotor thrust for the aligned case, the thrust on the nacelle and tower are likely more important than rotor thrust for the idling cases, except for cases with misaligned flow or faults involving blade pitch.
Dear Jason, thank you for your elaborate answer. The goal of my thrust-matching is to have a wind turbine model (e.g. a modification of the NREL 5MW or the IEA-15-240-RWT) that can ‘mimic’ the induced loads of another turbine, for which I do not have the rotor geometry, in coupled simulations.
This involves trying to match a given thrust - wind speed curve in uniform steady wind and/or a set of time traces of thrust in unsteady turbulent wind fields. In the latter case the matching is based on the statistic properties of the thrust.
Can you advise on the best approach to modify the rotor geometry to influence the induced thrust and side force in the idling case with yaw misalignment?
Best regards, Duncan
Am I correct in understanding that you are only dealing with numerical simulations, not a physical model? And you are trying to adjust the geometry of the NREL 5-MW or IEA 15-MW reference wind turbine OpenFAST model to match the thrust of a different turbine in an idling with yaw misalignment condition?
I would first modify the OpenFAST model so that the known properties of the different turbine are set properly in the OpenFAST model, e.g. blade length, blade mass, etc. For the idling with yaw misalignment condition, the induced velocity should be minimal (you should set
WakeMod = 0 to achieve no induction), so, the angle of attack at each aerodynamic node can be computed geometrically. That should provide an equation to relate the known thrust to the combination of chord and drag coefficient you need at that angle of attack.
That is correct, I am interested in performing numerical simulations and the case I am interested in is exactly what you state.
Thank you for the suggestion for the modification. I will look into that further.