Imposing Estimated Motion to Bottom of the Tower

Hi, dear NREL team,

I am working on my master thesis focusing on evaluation of hydroelastic response of a wind turbine platform. I am trying to impose the already calculated elastic motion to the interface between turbine and platform. Briefly, I want to apply the 6-DOF motion parameters to the first node of tower. I checked UserPtfmLd() and UserTwrLd() subroutine. However, UserPtfmLd() assumes there is a rigid platform and UserTwrLd() needs impelementation the loads to all nodes of tower. How can I impose the platform motion to the first node of the tower in either time domain or frequency domain (periodic motion)?

Kind Regards.

Dear Enes,

What you are trying to do sounds similar to what is discussed in the following forum topic: http://forums.nrel.gov/t/regarding-seismic-load-by-fast-code/1107/1.

I’d recommend against trying to prescribe the motion of the first tower node, which is related to both the platform and tower-mode degrees-of-freedom; instead, I suggest you prescribe the motion of the tower base, which is the platform in FAST.

Best regards,

Dear Dr.Jonkman,

Firstly, thank you for your response. Seismic loading seems useful, however it would be more helpful if the seismic loading routine was able to prescribe rotational motion. Is there any trick about rotational motions?

I would be thankful if you have recommendation for this problem.
Kind regards.

Dear Enes,

The prescribing of rotational motion was not needed within Seismic, but it would probably not be too difficult to modify the source code to support this additional functionality (although I have not tried this myself).

Best regards,

Dear Jason,

I am greatful for your quick response and help. I am not good at writing or modifying subroutines but I am going to try at least. Thanks again.

Best regards.

Dear Jason,

I am trying to impose a sinusoidal translational platform motion to a wind turbine in the latest version of FAST v8 just as Enes wanted to do. You mentioned the seismic loading module which however seems to only work with FAST v7 if I understand it right. Is there an easy way to achieve what I want to do in FAST v8 without too much coding as I know little about Fortran? Thank you very much!

Best regards,

Yuanchuan Liu

Dear Jason,

Since the Seismic Loading module is not yet available in FAST v8, I was thinking about at least implementing something similar yet simpler functions in v8. From what you suggested in topic: [url]http://forums.nrel.gov/t/foundation-stiffness-and-damping-in-fast-v8/1059/1] and the applied force model in the documentation of the Seismic Loading module, do you think I could utilise AddCLin and AddBLin and modify the fowllowing equation in HydroDyn.f90:

m%F_PtfmAdd = p%AddF0 - matmul(p%AddCLin, q) - matmul(p%AddBLin, qdot) - matmul(p%AddBQuad, qdotsq)

to something like

m%F_PtfmAdd = p%AddF0 - matmul(p%AddCLin, q - qDesired) - matmul(p%AddBLin, qdot - qdotDesired) - matmul(p%AddBQuad, qdotsq)

where qDesired and qdotDesired represent the sinusoidal equations of the desired displacement and velocity?
Thank you very much!

Best regards,

Yuanchuan Liu

Dear Yuanchuan,

Yes, your approach makes sense.

Best regards,

Dear Jason,

Thank you for confirming the approach is reasonable. And I can also confirm that by tuning the oscillator stiffness k and damping coefficient c in the documentation of the Seismic Loading module, the desired displacement can be achieved to a high level of precision (less than 1/1000 in my imposed surge test). Now I just need to find a way to read the parameters related to prescribed motion from the HydroDyn file.

Best regards,

Yuanchuan Liu

Dear Jason,

As a follow-up to my last attempt to impose platform motion by modifying the code, do you think an alternative simpler approach would be changing the incoming wind speed from steady to uniform but varying in time, e.g. U_wind = U_const + A_platformsin(w_platformtime)? In other words, are these two methods supposed to produce same aerodynamic thrust and torque variation?

Best regards,

Yuanchuan Liu

Dear Yuanchuan,

If the wind turbine structure was rigid, than the wind speed relative to the rotor would be identical between the two methods, so, I would expect that you’d get identical aerodynamic loads. However, if the wind turbine structure is flexible, I would expect a different response between the two methods as the inertial loads (and hence the resulting structural motion and relative wind speed) would be quite different between the two.

Best regards,

Dear Jason,

Thank you very much for your confirmation. I previously had different results between the two methods even when blades are rigid. After some careful examination, it turns out I just made some stupid mistakes. So yes. The results should always be the same whatever method is adopted as long as the blades are rigid.

Meanwhile, thank you for warning me of the shortcomings of varying wind speed for flexible simulations.

Best regards,

Yuanchuan Liu