Creating Cp plot using FAST and WT_Pref

Wind & Water Computer-Aided Engineering Software Tools
1

Hi, I am doing a project for the control system of a wind turbine. More specifically, my partner and I are focusing on the control system over region 2.5. At the moment we would like to create a Cp plot using FAST or WT_Pref. Is it possible for you to send me one of the older versions of WT_Pref if the bugs don’t impact too much on the calculations? I tried creating a Cp plot using FAST but I am unsure how to maintain a constant TSR and which VSContrl input would be best. Any guidance on this would be much appreciated. Thanks

2

Dear Micky,

We are not distributing any version of WT_Perf until we have a chance to test it. I would recommend you use FAST with no degrees of freedom and get rid of all controls. Hold the rotor speed constant and apply steady winds with steps of wind speed to sweep the TSR using a hub-height wind file. This is essentially an equivalent of WT_Perf. Make sure to have each step last long enough to for things to settle out (which should not be long for a rigid turbine) and use the last values of each step for the Cp to use in your curve.

Marshall

3

I forgot to mention that I am looking at an offshore wind turbine. Would the settings you described above still be applicable?

4

Micky,

Well, what I described is how to use FAST to mimic WT_Perf and it applies to any kind of HAWT. However, modeling a rigid turbine will give you different results than a flexible turbine. For instance, when the thrust increases, the blades will curve away from the wind and it might essentially reduce the rotor diameter. You will get less power that way. However, if your blades are precurved upwind, they might flatten and sweep a larger diameter.

If you want something closer to a real Cp/lambda curve, you should model a flexible turbine and do the wind-speed stepping described above (but use longer settling times between steps).

I have given little thought to it, but it may be that a flexible turbine will give a different Cp/lambda curve for different rotor speeds, so there would not be a single curve, but a surface of Cp vs. lambda vs. rotor speed. In other words, running the rotor at 10 rpm in a 5 m/s wind might give a different result than running the rotor at 20 rpm in a 10 m/s wind even though the tip-speed ratio is the same. A higher rpm will tend to flatten the rotor and if the blades are precurved upwind, the additional flattening due to the increased thrust at higher wind speeds will make it even worse.

Maybe a Real Aerodynamicist™ would like to chime in on this issue. I have never heard this discussed before, but it seems like it might be important.

Marshall

5

Hi Marshall,

I am having a bit of trouble wrapping my head around this and getting a plot. The way I have tried is to fix the wind speed, varied the rotor speeds, for the various TSR values and varied the blade pitch angles. Would it be best to have the same generator torque for all my test cases or a constant torque for each simulation? I am running the simulations using a matlab interface with FAST and the generator torque is governed by what region the turbine thinks it is running in.

Thanks,
Micky

6

Dear Micky,

To change TSR in FAST in order to derive the Cp versus TSR/pitch surface, its better to force the rotor to spin at a constant speed (by disabling the generator degree of freedom, GenDOF, in FAST) and vary the wind speed (slowyly, or in steps).

Best regards,