NREL 5MW OC3-Hywind (Test 24) Questions

Discuss theory and modeling of wind-turbine structures.

Moderator: Bonnie.Jonkman

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Tue Nov 17, 2020 9:51 am

Dear Sir,
Thank you for the information. Here are some queries regarding FAST theory.
1. How does FAST compute the loads in case of spar platform.
2. Can Morison theory be used for calculating heave forces. If not which method is used. (My understanding is Morison equation can be used to calculate horizontal loads only)

Jason.Jonkman
Posts: 5771
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Jason.Jonkman » Tue Nov 17, 2020 11:49 am

Dear Akheel,

Regarding (1), what loads are referring to? Applied aerodynamic loads; applied hyrodrodynamic loads, and/or reaction loads within the structure? There are many topics on this forum and related publications in the literature on this topics.

Regarding (2), the strip-theory solution in HydroDyn includes terms beyond just the distributed transverse inertia, added mass, and viscous drag loads that are part of Morison's equation. In addition to these terms, the strip-theory solution of HydroDyn also considers distributed axial loads on tapered members, distributed static buoyancy (about the undisplaced position), and concentrated loads at member ends (joints). And recent work has also added functionality to recalculate the distributed hydrostatic (buoyancy loads) based on the displaced position of the structure every time step (currently under review before being merged into the dev branch of OpenFAST in the following pull request: https://github.com/OpenFAST/openfast/pull/537). More information is available in the HydroDyn documentation, publications, and presentations.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Wed Nov 18, 2020 3:12 am

Dear Sir,
I tried to simulate load case 5.1 in Offshore Code Comparison Collaboration (OC3) for IEA Task 23 Offshore Wind Technology and Deployment, J. Jonkman, 2010. The response I obtained in Surge was not matching. The input conditions were as follows:

1. NRELOffshrBsline5MW_InflowWind_12mps:
WindType = 1
HWindSpeed = 8
ScaleMethod = 1
2. NRELOffshrBsline5MW_OC3Hywind_ElastoDyn
Enabled all dof
Rotor Speed = 6.9
3. NRELOffshrBsline5MW_OC3Hywind_HydroDyn
WaveMod = 1
WaveHs = 6
WaveTp = 10
PotMod = 1
4. NRELOffshrBsline5MW_OC3Hywind_ServoDyn
DLL file name

All other parameters were kept unaltered. Where am I making the mistake?
I have attached the response plot.
Attachments
Doc1.docx
(3.53 MiB) Downloaded 39 times

Jason.Jonkman
Posts: 5771
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Jason.Jonkman » Wed Nov 18, 2020 9:39 am

Dear Akheel,

The results you are plotting appear to be the first 20 s of simulation, with initial platform displacements set to zero. This is still the start-up transient phase of the simulation, which you should generally neglect in post-processing. The results published in the report you reference are shown after all start-up transients have passed. You should run a much longer, which with steady wind and period waves will eventually result in a periodic response. It is this periodic response that you should be comparing.

Please note to minimize the start-up transient in a floating wind turbine analysis, in addition to setting the initial rotor speed and initial blade-pitch angle properly, it is also useful to set the initial platform surge and initial platform pitch angles to be their expected mean values for the given mean hub-height wind speed. Otherwise, due to the low natural frequencies of the floating platform surge and pitch modes, it could take hundreds of seconds of simulation for the start-up transient to pass.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 8:23 am

Dear Sir,
The steady state response was matching well. I need a clarification regarding the second statement of your previous message, that if I am unaware of the expected initial pitch or surge and input some random value, Will the final result be wrong?
Secondly, In the load case 5.1, I tried to enable NTMD with following features, But the responses were not found to be controlled. Can you help me in this regard?

1 TMD_DOF_MODE
true TMD_X_DOF
false TMD_Y_DOF
---------------------- TMD INITIAL CONDITIONS ---------------------------------
0 TMD_X_DSP
0 TMD_Y_DSP
---------------------- TMD CONFIGURATION --------------------------------------
0 TMD_P_X
0 TMD_P_Y
3 TMD_P_Z
8 TMD_X_DWSP
-8 TMD_X_UWSP
0 TMD_Y_PLSP
-0 TMD_Y_NLSP
---------------------- TMD MASS, STIFFNESS, & DAMPING -------------------------
10473 TMD_X_M
0 TMD_Y_M
0 TMD_XY_M
86660 TMD_X_K
0 TMD_Y_K
4203 TMD_X_C
0 TMD_Y_C
500000 TMD_X_KS
0 TMD_Y_KS
500000 TMD_X_CS
0 TMD_Y_CS

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 8:43 am

Dear Sir,
I tried to simulate load case 5.3 of 'Offshore Code Comparison Collaboration (OC3) for IEA Task 23 Offshore Wind Technology and Deployment' J.Jonkman, 2010. I got an error message regarding Mach Number exceeding 0.3. Initial Conditions such as Rotor Speed = 12.1 and Blade Pitch = 14.92 were set prior to loading. The TurbSim file is as follows:

--------Turbine/Model Specifications-----------------------
31 NumGrid_Z
31 NumGrid_Y
0.05 TimeStep
1200 AnalysisTime
1000 UsableTime
90 HubHt
159.00 GridHeight
140.00 GridWidth
--------Meteorological Boundary Conditions-------------------
"IECKAI" TurbModel
"1-Ed3" IECstandard
"B" IECturbc
"NTM" IEC_WindType
90 RefHt
18 URef

OTHER PARAMETERS WERE KEPT UNALTERED
1. Is my TurbSim input file correct for the load case 5.3?
2. What would be the reason for the warning?

Jason.Jonkman
Posts: 5771
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Jason.Jonkman » Thu Nov 19, 2020 8:49 am

Dear Akheel,

Normally, improper initial conditions just result in a long start-up transient that eventually dies out, after which the results are fine. But if initial conditions are not set properly, sometimes the control system may not recover, which may lead to an instability that is quite obvious to spot (very large, unrealistic deflections and loads).

Regarding your second question, I can't say whether these TMD settings will have a large effect on the system response. I gathered from prior questions that you do at least see some effect from the TMD (e.g. you feel the weight of the TMD in the tower). So, perhaps changes to the TMD parameters are warranted to see a greater influence. I can't really guide you on that.

Your TurbSim settings look fine to me. How quickly do you receive this warning? Could it be triggered because there is still a large start-up transient due to not setting good initial conditions for platform surge and pitch?

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 9:54 am

Dear Sir,
The warning was found to occur at about 600 sec

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 9:56 am

Dear Sir,
Is there any limit for mass ratio, Stiffness, Damping of TMD in nacelle?

Jason.Jonkman
Posts: 5771
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Jason.Jonkman » Thu Nov 19, 2020 9:59 am

Dear Akheel,

There is no hard limit to the mass, stiffness or damping of nacelle-based TMDs in OpenFAST. But I could imagine setting any of these values too large may eventually result in numerical issues with the solution.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 10:39 am

Dear Sir,
How does OpenFAST distinguish between JONSWAP and PM Spectrum? Setting WaveMode = 2 is common for JONSWAP and PM Spectrum. Is there any other parameter which needs to be adjusted depending on spectrum used?

Jason.Jonkman
Posts: 5771
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Jason.Jonkman » Thu Nov 19, 2020 10:46 am

Dear Akheel,

The JONSWAP spectrum is just a modification of the Pierson-Moskowitz spectrum, based on a peak-shape parameter. Setting WavePkShp = 1 results in the Pierson-Moskowitz spectrum.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 10:53 am

Dear Sir,
This is regarding the Jason_PSD.m. I imported the table, coded the time series as 'Data' and entered df. then I used the command

Code: Select all

[ f, S1 ] = Jason_PSD( Data, df );

The results seems to be same as the fft output (A vertical line close to the y-axis). What modifications do I need to make to obtain the one sided PSD with unitm^2/Hz ?

Jason.Jonkman
Posts: 5771
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Jason.Jonkman » Thu Nov 19, 2020 10:58 am

Dear Akheel,

S1 is the one-sided PSD with the units you want. The script is quite simple; open it up and review it for details.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Akheel.Kodambiyakamkizhakkakam
Posts: 51
Joined: Wed Jul 15, 2020 5:28 am
Organization: NIT Calicut
Location: India

Re: NREL 5MW OC3-Hywind (Test 24) Questions

Postby Akheel.Kodambiyakamkizhakkakam » Thu Nov 19, 2020 11:11 am

Dear Sir,
Thank you for the quick response.
I did the simulation of load case 5.3. I used the time series in fore aft response in the script Jason_PSD.m expecting that I would get a plot similar to Figure 49 of the report. That I obtained now was a horizontal line (when x and y axis are arranged as in Figure 49). Correct me if my expectation was wrong. Could you please suggest the exact method to obtain the plot if my theory was wrong?


Return to “Structural Analysis”

Who is online

Users browsing this forum: No registered users and 1 guest