Turbine Aerodynamics

Dear Jason,

I’m looking into the effect of the dynamic stall model and inflow models on the aerodynamic calculations in FAST, and using the NREL 5MW baseline turbine affixed on the MIT/NREL TLP. The wind speed is constant at 6m/s, TSR=7.625, the platform has surge, pitch, and heave DOFs activated, and the waves are regular with height 0.5m and period 4.5s.

Now I’ve considered 4 combinations of the dynamic stall and inflow models as follows:
test 1.1: No dynamic stall + BEM inflow
test 2.1: Beddoes dynamic stall + BEM inflow
test 3.1: No dynamic stall + GDW inflow
test 4.1: Beddoes dynamic stall + GDW inflow

I’ve run these tests in FAST to compare them to the results I get from WInDS for the same platform under the same conditions, to compare the aerodynamic modelling tools. I’m attaching the curves I got for power, torque, and thrust, also superimposing the curve representing the change in pitch with time (degrees per second). As can be expected, the power and torque peak when the change in pitch is minimal (according to the FAST sign convention), and the WInDS results agree for these performance characteristics. However, when it comes to thrust I’m seeing a phase difference (approx. 78 degrees) to the power and torque variations and the change in pitch. The WInDS result for thrust remains in phase with its power and torque counter-parts, but for some reason the FAST thrust is out of phase. Since this happens for all the tests I am assuming it’s not just some numerical error but I cannot understand why this is happening.

My question is, is there some parameter which is involved in the FAST calculations for thrust which could be influencing the phase? Is it not purely aerodynamic?

Any light you could shed on this would be greatly appreciated.

thanks in advance!

best regards

Kurt

Dear Kurt,

Yes, FAST output RotThrust (which is what I assume you’re plotting) not only includes the applied aerodynamic thrust, but also the gravity and inertial loads of the rotor. See the following forum topoic for more information: http://forums.nrel.gov/t/the-effect-of-tilt-angle/769/1.

Best regards,

Dear Jason,

Yes after I posted this I kept on looking and came across a couple of posts explaining how to go about deriving the aerodynamic thrust, so sorry for bringing it up again!

I’ve chosen to add all the ‘ForceNx’ values for each element on a blade together at each time step and then find the total for all the blades at each time step (by adjusting the azimuth angle for blade 1 to 0, 120, and 240 degrees respectively). I did this because I am assuming rigid blades and tower, therefore no flexibility is allowed. I determined the final aerodynamic thrust by transforming the force to compensate for the coning angle: thrust = (ForcN_total)*Cos(cone angle).
My thrust results are now all in phase with power and torque, and even match the WInDS results well.

Thank you for your help

best regards

Kurt