Dear Jason,
Could you check the above quoted my previous post on Fri Feb 14, 2020 3:37 am? I do not understand why the results of ‘FAST-Morison’ and ‘FAST-Potential’ do not agree with each other, because they are FAST built-in functionalities. Do you know why? Are the results reasonable and acceptable?
The calculation was done by FAST_v8.16.
Best regards,
Yingyi
Dear Yingyi.Liu,
Sorry, I did not see your post on Feb 13, 2020 10:37 am, likely because you posted again right afterword.
I would expect that the strip-theory solution and potential-flow solution to give quite similar results for the OC3-Hywind floating wind system if the OpenFAST inputs are set properly.
For the strip-theory only solution, “FAST-Morison”, what else did you change in the HydroDyn input file besides PotMod and PropPot? Did you specify the correct transverse added-mass coefficient (SimplCa = 0.969954), based on the OC3-Hywind specification document: wind.nrel.gov/nwtc/docs/HydroDyn_Manual.pdf?
Best regards,
Dear Jason,
Yes, you are correct, I have not changed other settings in the HydroDyn input file. Regarding the two settings you mentioned above, I have questions:
(1) Should SimplCa = 0.969954 be used only in Morison model? and in potential solution SimplCa should be 0 ?
(2) I notice that the HydroDyn manual suggests the users to calculate AddCLin by themselves. Can we use the hydro restoring matrix data in spar.hst which is calculated by WAMIT for AddCLin ? Since in this case we can avoid calculating it manually by ourselves.
Best regards,
Yingyi
Dear Yingyi.Liu,
Here are my answers to your questions:
(1) For the potential-flow solution (PropPot = True), SimplCa is not used (because the fluid inertia and added mass terms are accounted for in the potential-flow solution rather than in the strip theory solution). So, it doesn’t matter how SimplCa is set when PropPot = True.
(2) Yes, you can use the values from the .hst file to set AddCLin. But please be aware that the values in the .hst file are nondimensional, and so, have to be dimensionalized first (multiplying by WtrDensGravityWAMITULEN^exp, where exp is 2, 3, or 4 depending on the term, but WAMITULEN is often set to 1 m, so, often doesn’t matter).
Best regards,
Dear Jason,
Following your answers, we successfully achieve the agreement between the FAST built-in Potential method (PropPot = True) and Morison method (PropPot = False).
However, we find a new problem which relates with a parameter WaveTMax in the HydroDyn input file.
We compare the results calculated with WaveTMax set as 3630s and 1000s respectively, using the FAST built-in Morison method (PropPot = False).
To our surprise, the two results do match at the beginning, however, there will be an obvious phase difference of pi in the motions when the platform motions get steady (see the graphs below).
(1) Do you know the reason for this problem?
(2) If the value of WaveTMax matters, then How can we set a correct WaveTMax in a simulation? Is 700s ~ 1000s sufficient for the Test 23 case (MIT-NREL TLP)?
Looking forward to your reply.
Best regards,
Yingyi
Dear Yingyi.Liu,
Presumably you see the same phase shift in the wave-elevation time series? My guess is the problem is related to the frequency resolution. The frequency step is dictated by WaveTMax. If your user-specified wave frequency (inverse of the wave period, WaveTp) is not exactly aligned with a frequency step, HydroDyn will round the wave frequency to the closest step. So, the wave periods are likely slightly different between WaveTMax = 3630 s and WaveTMax = 1000s.
In general, I would recommend keeping WaveTMax > 3600 s (even if your simulation length is less) to ensure that the frequency resolution of the wave spectrum is adequate.
Best regards,
