AeroDyn power problem

[fbelli997]

Hi everyone,

I'm trying to estimate the freestream wind speed in FAST.Farm using FLORIS. To do so, I take the power production from the FAST.Farm .out files, averaged over a 50-second moving window. In FLORIS, I then sweep the wind speed values and select the one that minimizes the power error.

As a first step, I attempted to validate this approach using WindDiskVelX (and doing sum of square with Y and Z components), and then I tried with RtVAvgxh (and doing the same as before), since, if I got it right, it takes into account for the relative wind speed seen by the rotor which includes platform motions.

• In the first case, I found some mismatch, possibly due to the platform pitch and surge motions (are there any other contributions which can contribute in this mismatch?).

• In the second case, I observed a better trend between my FLORIS predictions and the FAST.Farm results, however, I found a bias around 0.5 m/s. (I will show some plots here.)

this is usign WindDiskVelx

and this represents the error

This is using RtVAvgxh

What is the reason why of the discrepancy between the wind speed values from WindDiskVelX and RtVAvgxh? I’m using a generator efficiency of 95.4%, which could explain a difference of around 0.15 m/s, but not the entire mismatch.

Additionally, I do not fully understand the value of RTcp, which is 0.504 in the outputs. From GenPwr, however, I obtain 0.4536 (for a wind speed of 8 m/s and a power production of 6550 kW). This inconsistency confuses me.

Thank you all in advance for your help!

Federico

[jjonkman]

Dear @fbelli997,

Just a few comments:

• InflowWind output WindDiskVelX and AeroDyn output RtVAvgxh have several noticeable differences, including their orientation (projected along the xi axis of the inertial frame for WindDiskVelX versus along the shaft axis for RtVAvgxh) and underling wind contributions (relative velocity between wind and structure, as well as tower influence effects, for RtVAvgxh versus ambient only (or inclusion of upwind wake effects in FAST.Farm) for WindDiskVelX); there spatial averaging method also differs.
• I'm not sure how your comment about "sum of square with Y and Z components" relates to your question.
• I'm not sure I understand your point about RTcp; are you referring to AeroDyn output RtAeroCp? Regardless, the aerodynamic power should be more than the electrical power. Moreover, I'm not sure if the value of 0.504 you are referring to is instantaneous or time-averaged, but I would only use AeroDyn output RtAeroCp in a time-averaged sense.

Best regards,