How to transfer steady-state wake distribution to FF inflowwind for FAST?

Wind & Water Inflow Turbulence
1

Dear all,

I have developed a 3D analytical wake model (like some adopted in FLORIS but more sophisticated) and therefore am able to obtain steady-state wake distribution at any given location with a certain value of turbulence intensity. My question is how to generate time-series wind data based on the above steady-state wake distribution as an inflow file for OpenFAST?

Thanks in advance
Ruiyang He

2

Dear Ruiyang,

Well, I could see a couple approaches, e.g.:

  • Writing your inflow wind data into one of the standard file formats supported by OpenFAST’s InflowWind module (e.g., Bladed/WindType = 4 or 7, TurbSim/WindType = 3, or Mann/WindType = 5 formats), or
  • Implementing your inflow in InflowWind’s user-defined wind source file, IfW_UserWind.f90 (WindType = 6).

See the InflowWind User’s Guide for more information: nrel.gov/wind/nwtc/assets/d … Manual.pdf.

Best regards,

3

Dear Dr.Jason,

Thanks for your reply. But it seems that my problems are still unsolved.

Let me make my problem a little clearer: I found that the turbsim can only generate time-series turbulent wind data for shear or uniform wind (based on RefHt, URef, PLExp). But my inflow wind profile is actually the wake distribution and even the multi-wake superposition (provided by my steady-state wake model) since I am going to investigate the structural loads of downstream wind turbines. My problem is how to convert the steady-state wake distribution in YZ plane to time-series turbulent wind data

For the first method. Yes, I can write the inflow wind data into a standard file format, for example, a 13X13 matrix, but there is only one matrix since I can only provide steady-state wind distribution. Also, I can only write the wind distribution in mainstream direction (u), but no lateral or vertical velocities (v and w). Therefore, I haven’t found any of them suitable for me. What do you think?

For the second method. Since I am not familiar with FORTRAN and the code is too long, I don’t know how to start modifying it. Is there any sample for using IfW_UserWind.f90 or more specific guidance?

Best regards,
Ruiyang He

4

Dear Ruiyang,

I’m not really understanding what you want to do. Are you trying to add turbulence to your steady state wake data? Do you already have a method to do this? You could, of course, superimpose the two (sum your wake deficits with the background ambient turbulence from TurbSim), but in reality, the background turbulence will effect the wake evolution and meander the wake. This is, in effect, what NREL’s FAST.Farm extension to OpenFAST already does. Have you considered using FAST.Farm instead?

I don’t have more specific guidance on programming IfW_UserWind.f90. It is hoped that the template is straightforward if you know a bit about Fortran (obviously, you’ll need to learn a bit about Fortran). The NWTC Programmer’s Handbook provides guidance on the modularization framework of FAST/OpenFAST if that that helps: drive.google.com/file/d/1bDV1fB … sp=sharing.

Best regards,