Free-Vibration Response of OC3HYwind Spar turbine

Dear Zahid,

I don’t see that you have the Echo, NTypes, NConnects, or NLines input parameters in your MoorDyn input file. I suggest that you do a file compare to an existing MoorDyn input file for comparison (e.g. NRELOffshrBsline5MW_OC4DeepCwindSemi_MoorDyn.dat in the CertTest\5MW_Baseline directory of the FAST v8 archive).

Best regards,

Dear Dr. Jonkman
I wanted the response of the system at location near to center of mass (instead of water surface) therefore following your previous suggestions, i specified “PtfmRefzt = -78 (for OC3hywind)” in ElastDyn input file. It is working fine however i am little confused about HdryoDyn outputs which i want to understand.
1). If i specify “PtfmRefzt”, different than water surface level,Then the output e.g. AddMxi, AddMyi, AddMzi, are calculate with reference to mean sea level or with reference to “PtfmRefzt = -78” ?

2). In the HydroDyn input file we specify additional linear stiffness coefficient for hydrostatic stiffness. The K44, and k55 (for roll and pitch) stiffness coefficient depends on the ZCOB, i want to know after “PtfmRefzt” to be different than mean sea level, do i have to input the k44 and k55 with reference to “PtfmRefzt”, or these coefficient must be calculated with reference to mean water level.

I will highly appreciate your help.
Regards,
Zahid

Dear Zahid,

The reference points and coordinate systems in FAST’s ElastoDyn and HydroDyn modules are not tied together. Changing PtfmRefzt in ElastoDyn does not effect the inputs or outputs of HydroDyn. So, when PtfmRefzt = -78 m, AddMxi, AddMyi, AddMzi etc. are still calculated relative to mean sea level and K44 and K55 are still specified relative to mean sea level.

I hope that helps.

Best regards,

It was definitely helpful. thank you Dr. Jonkman for all your help.

Dear Dr. Jason Jonkman

I have another small query about the Free vibration response and your response will be highly appreciated. I want to understand the free vibration response of OC3Hywind, as i have posted many questions. I am using FAST V8.16. All the hydrodynamic forces due to waves, and aerodynamic forces are made zero. My question is regarding the variation of Hydro-static restoring force (AddFzi) with rotational degree of platform (pitch or roll). I applied initial pitch rotation of 0 deg, 2 deg and 4 deg and checked the free vibration response (all other initial displacement initially set to zero). I plot the heave displacement and AddFzi as shown in Figure below. I am a little confused about the AddFzi, because it varies with the pitch rotation which i am unable to understand. I thought it will not vary because the linear hydro-static stiffness is diagonal. I will appreciate if you can look at the plats and help me in understanding the source of variation. I highly appreciate your help and cooperation. (I notice that as i apply initial pitch the magnitude of AddFzi varies at t = 0 as well).

Regards,
Zahid

Dear Zahid,

HydroDyn output AddFzi is the heave forces due to the additional preload, stiffness, and damping. Assuming you haven’t modified the additional preload, stiffness, and damping properties of the OC3-Hywind model that NREL has distributed, then the only nonzero term in the heave direction is AddBLin(3,3) i.e. a an additional linear damping in the heave direction due to heave velocity. Thus, HydroDyn output AddFzi for this case represents the additional linear damping force due to heave velocity. You haven’t plotted the heave velocity, but you do show that the heave displacement varies as a result of different initial pitch offsets.

I hope that helps.

Best regards,

Dear Dr. Jason Jonkman

I wanted to check the effect of hydro-static stiffness, therefore i made the other terms zero (preload and damp). I have attached the heave velcoity and also a snapshot of preload, stiffness and damping terms. I will appreciate if you can look at the figures.


Dear Zahid,

The response looks reasonable to me. Is there something specific you wanted me to comment on?

Best regards,

Dear Dr. Jason Jonkman

First i sincerely apologize for late reply. I wanted your comment on the Addfzi at t = 0 sec. From figures, the heave displacement and velocity both are zeros at t =0, for pitch =2 and 4 deg, but the AddFzi is not zeros at t = 0 sec. As the AddFzi depends on heave velocity and heave displacement (also preload = 0), if these are zeros at t = 0, i thought AddFzi will be zero as well. I want to know the value of AddFzi.

Dear Zahid,

OK, looking more closely at your results I don’t see that the platform heave velocity you are plotting is consistent with the platform heave displacement you are plotting. E.g. the velocity should be zero when the displacement hits a local minimum or maximum, but this is not the case in your plots. Are you sure you are outputting/plotting the correct channels?

By the way, AddFzi is based on the platform motion expressed in the inertial frame e.g. you should be plotting ElastoDyn outputs PtfmTDzi (or equivalently, PtfmHeave) and PtfmTVzi (instead of PtfmTVzt).

Best regards,

Dear Dr. Jason Jonkman

Following your suggestions, i plot the PtfmHeave, PtfmTVzi (instead PtfmTVzt ) and AddFzi. Additionally i wanted the response at PtfmRefz = -78m. So i plotted the the three mentioned quantities at PtfmRefz = 0m, PtfmRefz = -78m for three initial pitch rotation of 0 deg, 2 deg and 4 deg. All the plots are attached below. Basically i am unable to understand that why there is change in heave displacement with change in initial pitch rotation. I was thinking may be it is due to AddFzi where AddFzi is function of AddClin and AddBLin. I have also attached the snap of input file of AddClin and AddBLin. I would really appreciate your help in understanding the variation in Heave displacement and AddFzi with Pitch rotation. If any other input information i will post.
(All the picturte could not be uploaded here so i uploaded few in the next post).
3.PNG
1.PNG
2.PNG

The same quantifies at PtfmRefzi under same initial pitch rotation of 0, 2 and 4 deg, are shown in figures attached below.
6.PNG
5.PNG
4.PNG

The Input of addClin and AddBLin for this case are shown in the snapshot.

Dear Zahid,

Your results look reasonable to me. The platform heave velocity and displacement are now consistent with each other (the velocity being the derivative of displacement) and AddFzi is clearly proportional to PtfmHeave for PtfmRefzt = 0.

There is obviously something in your system that couples pitch and heave. While the coupling is clearly not from AddCLin (because it is diagonal), there is probably coupling from another source e.g. the mooring system.

Best regards,

Dear Dr. jason,

I am doing project on 10MW Spar-Bouy wind turbine. My system is in the hydrostatics equilibrium condition but if I do the simulation without wave and wind for the decay test then I am not getting platform motion frequency similar as compare to other 10MW spar-Buoy decay test results,

You can see the platform motion frequency different as below both figure I have attached as below, I think it’s because of Additional linear damping(AddBLin) matrix and I have taken those values from NREL 5MW Spar-Bouy wind turbine. So can you please guide me how to define Additional linear damping(AddBLin) matrix non-zero terms. I could not find in the HydroDyn manual.is it Important to define for my system(10 MW Spar-bouy wind turbine)?

I can see the equation in this following pdf nrel.gov/docs/fy10osti/47535.pdf , But I could not get the idea how to define the Fi additional Damping and qj.

I have also attached my HydroDyn file as below you can see the values those I have used for the Additional linear damping(AddBLin) matrix.I hope that is the only reason for the frequency of platform motion.

I hope you will guide me and help me out as soon as possible.

Thank you very much

Best regards
Mitesh Ramani
Ptfm motions comparison for the decay test.zip (181 KB)
HydorDyn.rtf (18.3 KB)

Dear Mitesh,

The additional linear damping matrix (AddBLin) in HydroDyn will have little influence on the platform natural frequencies. Instead, AddBLin will influence the level of platform free-decay damping. The values of AddBLin defined for the OC3-Hywind spar where set to ensure that the free-decay response matched between FAST and wave-tank test data provided by Statoil. I would assume that these same values would not apply to another platform.

The platform natural frequencies would be dictated by the body mass/inertia, hydrodynamic added mass, gravitational restoring, hydrostatic restoring, and mooring restoring. If your free-decay derived natural frequencies are not matching between numerical simulation and experimental measurements, I would expect a problem with one more of these terms.

Best regards,

Dear Dr. jason,

I really appreciate your help,

As you told me so I have changed platform inertia as per body mass.

1)I am using OpenFAST and PotMod- Potential-flow model {0: none=no potential flow, so I do not need to add potfile (.hst) that is why I do not need to define hydrodynamic added mass and hydrostatic restoring. Am I right??

2)Gravitational restoring and mooring restoring values are correct in my files because My system already in the hydrostatics equilibrium condition. I think so I don’t need to change any values for that too.

3)I have set the additional linear damping matrix (AddBLin) value all zero in HydroDyn as you can see in below my HydroDyn file.

Only I have changed the platform inertia and the additional linear damping matrix (AddBLin) value set to zero after that I have seen huge different in the platform motion frequency.

Thank you for your explanation, the additional linear damping matrix does not matter much in the platform motion for the decay test but can you please guide me how to define that matrix values for the free decay test?

I have uploaded my platform motion frequency and NTNU report frequency, My Spar-Bouy is little different so do you agree that this frequency is good enough for my 10MW Spar-Buoy system.

I have uploaded my ElastoDyn and HydroDyn file as below can you please overlook if I need to change anything important.

I hope you will help me as soon as possible.

Thank you

Best Regards
Mitesh Ramani
HydroDyn.rtf (18.3 KB)
ElastoDyn.rtf (27 KB)
Platform Motion comparision.zip (503 KB)

Dear Mitesh,

Correct, when PotMod = 0 in HydroDyn, you don’t need to provide WAMIT-calculated added mass, damping, hydrostatics, or wave-excitation loads.

In your comparisons between FAST and the NTNU data, I don’t see that you’ve actually run the equivalent free-decay simulations in FAST. To run a free-decay simulation, set the initial platform displacement(s) to a nonzero value and simulate the motion response in still water. From this simulation, you can identify the natural frequencies and damping of the platform. You can run a separate free-decay simulation for each platform degree of freedom (surge and pitch are likely coupled; likewise for sway and roll). If the damping does not match, you can tune the viscous drag coefficients and/or the AddBLin and AddBQuad matrices to better match the damping from the NTNU data.

I hope that helps.

Best regards,

Dear Dr. Jason,

Thank you so much for your help and support,

I got your point that I did not run the free decay test simulation but It was only no wind and no wave simulation thank you for that information.

  1. As you told me to run free decay test I have to set the initial platform displacement(s) to a nonzero value,Those value directly I have to put from the Windchart data means from the platform motion charts (I should use those displacement values of simulation with No wind and No waves ) AM I right ??

  2. AS you said if I run that simulation for each DOF then I can identify the natural frequency and damping of the platform.Otherwise I can tune the viscous drag coefficients and/or the AddBLin and AddBQuad matrices to better match damping from NTNU data.

could you please tell me how can I get this matrix value I am very confused about this because I think It is also Important for the platform frequency and damping and I had set it to zero for my last free decay simulation.

I hope you will guide me and help me out as soon as possible.

Thank you very much

Best Regards
Mitesh Ramani

Dear Mitesh,

The initial platform displacements are set in the Initial Conditions section of the ElastoDyn input file.

You could guess and check different viscous drag coefficients and/or the AddBLin and AddBQuad entries, but this would be a very crude way of calibrating them. A better approach would be to set up a least-squares optimization problem, so as to choose the correct viscous drag coefficients and/or the AddBLin and AddBQuad entries that minimize the error between the FAST simulation and experimental data. I’m sure you can find papers detailing various calibration approaches in the literature.

Best regards,