AeroDyn and FAST

Hello all,
I am new to this forum and a beginner to FAST. Recently, to learn the use of FAST, I downloaded FAST_v7.00.00a-bjj_AeroDyn_v13.00.00a-bjj_BladedDLLInterface.exe and from Jason’s public folder, and carried out some simulations. I encountered two problems:

  1. In the input file of AeroDyn NRELOffshrBsline5MW_AeroDyn.ipt, when the IndModel is set to be SWIRL, the TanInd is always 0.00000E+00, which is inconsistent with AeroDyn User’s Guide, Version 12.50. I downloaded AeroDyn source files (AD_v130000a-bjj.exe) from NREL website, and I found a few lines of code seems wrong, starting from line 424 to 441 in AeroSubs.f90:

[code] SELECT CASE ( TRIM(Line) )
Wake = .FALSE.
Swirl = .FALSE.

     CALL ProgWarn( ' All wake calculations are turned off! This option is recommended only '// &
                    'in high winds or for debugging.' )
     Wake = .TRUE.
     Swirl = .FALSE.
     Wake = .TRUE.
     Swirl = .TRUE.
     CALL ProgWarn( ' Error: Expecting "NONE", "WAKE", or "SWIRL" wake model option.')
     ErrStat = 1

Wake = .TRUE.
Swirl = .TRUE.
Here, I think Wake = .TRUE. should be Wake = .FALSE.
2. In the platform file NRELOffshrBsline5MW_Platform_TLP.dat, there is a line:
“200.0 WtrDpth - Water depth (meters) [USE THE SAME VALUE SPECIFIED IN THE WAMIT .POT FILE!]”
It seems that I can not change the value of water depth with the HydroData unchanged. However, I do not have a licensed WAMIT.
So, I have two questions:
Can I change the WaveHs and WaveTp with the HydroData unchanged?
Is there other ways for me to change the water depth?

Dear Qiang,

When IndModel is set to SWIRL, AeroDyn enables both axial and tangential induction. In the AeroDyn source code, variable Wake is a flag to enable axial induction and variable Swirl is a flag to enable tangential induction. So, the source looks correct to me. I’m not sure why your model shows zero-valued TanInd. What is InfModel set to? Is TanInd zero for all elements for all time, or just some elements some time?

The radiation and diffraction solutions of WAMIT are typically found with a finite-depth boundary condition. The statement in the HydroDyn input file that you refer to is reminding you of this. For the greatest accuracy, it is recommended that you set the water depth in WAMIT and HydroDyn to be identical. Of course, the finite-depth boundary condition may have only minor influence on the WAMIT hydrodynamic coefficients, depending on the geometry of the floater, sea state, etc. When the dependency is minor, it reasonable to alter the HydroDyn water depth without updating the WAMIT hydrodynamic coefficients. I haven’t performed a sensitivity study on water depth for the WAMIT hydrodynamic coefficients of the MIT/NREL TLP, but it is probably reasonable to assume that the coefficients are reasonable for deviations of the water depth around 200 m.

The WAMIT hydrodynamic coefficients are derived independent of the sea state, so, it is possible to set WaveHs and WaveTp in HydroDyn without modifying the WAMIT hydrodynamic coefficients.

I hope that helps.

Best regards,

Dear Jason,
Thank you so much for you detailed response. It is very helpful for me.
I’m sorry I have misunderstood the code.

The InfModel is set to DYNIN, and TanInd is zero for all elements for all time. However, I have only tested three cases when the inflow wind speed is 5 m/s, 8 m/s and 12 m/s.
I have just found that when the InfModel is set to EQUIL, and IndModel is set to SWIRL, TanInd will not always be zero.

Dear Qiang,

It looks like you have found a bug in AeroDyn. After a little checking into the code, we have discovered that while TanInd is calculated, it is not written in the output file when SWIRL is TRUE.

To fix this, you can can add a line in AeroSubs.f90’s vindinf() subroutine. The following code starts at line 7284 in AeroDyn v13.00.00a-bjj, with the additional line surrounded by comments:

IF ( SWRLARG > 0.0 ) THEN A2P = 0.5 * ( -1.0 + SQRT( SWRLARG ) ) VT = VT * ( 1.0 + A2P) !bjj start of proposed change v13.00.01a-bjj AP(iradius,iblade) = A2P !bjj end of proposed change v13.00.01a-bjj ENDIF

We have modified our code and will include it in the next release of AeroDyn. Thanks for finding the error.


Dear Bonnie,

Thank you very much! It’s so exciting that I actually found a bug! :smiley:

Can I just change the water depth to 150m? Thank YOU.
200.0 WtrDpth - Water depth (meters) [USE THE SAME VALUE SPECIFIED IN THE WAMIT .POT FILE!]

Dear Yiqing,

You can, but without updating the hydrodynamic coefficients through a WAMIT calculation you run the risk of reduced accuracy of the solution. The hydrodynamic coefficients computed by WAMIT are only independent of depth for deep water. My guess is that the hydrodynamic coefficients wont change much between 200 m and 150 m, but I wouldn’t go much shallower than that without updating the coefficients.

Best regards,

[size=100][size=100]Dear All,

I am trying to obtain the hydrodynamic coefficients (added mass, radiation damping), first order wave excitation force (magnitude and the phase angle) of MIT/NREL TLP by using NEMOH panel method solver. I am attaching my results where I compared :

  • NEMOH solution (Blue data series)
  • HydroD solution of SESAM (HydroD data series: red)
  • HydroD solution of the same platform except with the draft= 50m. (EB data series: pink)
  • I have digitized the values of Denis Matha , where he published in: NREL/SR-500-45891 Figure 16, 17,18 (Matha 2010 data series: dark dashed)
  • tlpmit.1 & tlpmit.3 WAMIT outputs that are available in the FAST distribution folder (NREL-WAMIT data series: dark-solid)

With the comparisons, except B33 all looks fine (B11 and B55 is acceptable)
All the analyses considers the same platform geometry, draft height. All the geometries are modeled without he pontoons as Matha (2010) also stated.
I am confused with the comparison between Matha (2010) and WAMIT data series which I except them to be the same. What I am missing? Have any idea?
Thanks in advance for your time, suggestions and help.


Dear Mustafa,

I would expect the WAMIT output files provided in the FAST/OpenFAST archives for the MIT/NREL TLP (tlpmit.1, .3, and .hst) would be the same as what is in Matha (2010).

The WAMIT output data is nondimensional, so, I would first check that you’ve redimensionalized the data properly. I’m questioning particularly the B33 plot, where the B33 term does not tend toward zero at low frequency, even though it should be redimensionalized by multiplying by rho*omega. See the following forum post for information on how to dimensionalize the WAMIT output:

Best regards,