Questions about the HydroDyn Module

Dear Jason,

I’ve read your PHD thesis, it seems that in the HydroDyn Module only the first-order wave forces are considered. But for the calculation of the floaters like a SEMI or a BARGE, the second-order wave forces are also important to the dynamic responses of the system. So I wonder whether there is a way to implement the second-order wave forces in the FAST calculation? Or can I add the second-order wave forces just the same way as adding the first-order wave forces using WAMIT to obtain the results and then adding it as the input parameters?
Thanks.

Best Regards.

Dear Yu Ma,

You are correct that the current version of HydroDyn is based on linear (first-order) wave-body hydrodynamics. You are also right that second-order hydrodynamics could be important for semi-submersibles and barges; it could also be important for spar buoys and tension-leg platforms (TLPs).

We have recently started a project to upgrade HydroDyn to include second-order hydrodynamics for floating platforms. Second-order wave-excitation loads will be implemented by reworking the HydroDyn module to (1) input second-order frequency-domain quadratic-transfer functions (QTF) preprocessed through WAMIT and (2) transform the loads to the time domain based on the specified irregular wave spectrum and directional spreading. The implementation will include bichromatic and bidirectional (short-crested) wave components in finite depth water. The complete second-order solution will be superimposed on the first-order and viscous drag terms currently computed in HydroDyn. But because this project is just getting started, it will be a while before this version is completed and released.

Best regards,

Dear Jason,

Thanks for replying.

I have a few other questions now.

In the output list of FAST calculations, there are “platform loads” including “Platform horizontal surge shear force” et al. I wonder if you can explain what exactly those loads are refering to? Are they internal forces in a certain cross section?

Also, I want to know if the whole horizontal restoring stiffness in x and y directions of the mooring system can be obtained by FAST?

Furthermore, I would like to know if FAST can change the wind direction or it could only be set as coming along the x direction?

Thanks.

Best wishes.

Dear Yu Ma,

I believe your question regarding platform loads was recently answered in the forum topic found here: Clarifications PtfmFxyz/mxyz in FAST. Please let me konw if additional clarification is needed.

It is possible to obtain the complete linear 6x6 stiffness matrix associated with the mooring system from FAST. Let’s call this matrix K_m. When linearizing a FAST model, one has to always consider the influence of “effective stiffness” – see the following forum topic for more information: FAST: Model linearization. To get around this issue, the following process can be used to calculate K_m directly:

  1. Set-up a FAST model of the foating wind system.
  2. Instead of using the correct WAMIT data files for your floating platform, use WAMIT data where all of the coefficients are zero-valued. You can find suitable zero-valued WAMIT data files attached. These were obtained by replacing the values in a normal WAMIT output with zeros at all periods.
  3. Include the 6 platform loads PtfmFxi through PtfmMzi in the FAST output list.
  4. Run a FAST linearization analysis with CalcStdy = False.
  5. Grab the 6x6 matrix relating the 6 platform load outputs to the 6 platform DOFs – that is DspCMat.
  6. Derive K_m = -1000*DspCMat.

The factor of -1000 in step 6 is needed to correct for units (kN → N) and sign.

When using FAST in conjunction with turbulent wind data files from TurbSim, the wind “grids” (“planes”) will propagate along the xi-axis regardless of the wind direction specified in TurbSim, which is why we always recommend setting the wind direction to zero when using TurbSim in conjunction with FAST. See the TurbSim User’s Guide for more information. To model a yaw error with Turbulent wind, set the nacelle-yaw angle to be nonzero in FAST instead of changing the wind direction. When using AeroDyn’s so-called “hub-height wind files,” the wind direction need not align with the xi-axis.

I hope that helps.

Best regards,
Zeros.zip (58.8 KB)

Sir ,
when we use these zeros. files i am getting a increasing response with time.
I have a TLP as a floating structure and in that Pretension of tether is very high , so according to me if the analysis is reading the platform file then due to tether the response will not go higher at every time .
I want to know what are the values will effect the response in Hydrodyn model.
and one more thing i rum two cases
(1)i use only -1 and 0 period for added mass and .3 file are empty.
(2) i use all the values in .1 and .3
.hst is same for both cases
will analyzing i get the same results.why ?? It will only depend on -1 and 0 period added mass .or i have done something wrong ?

Dear Anuj,

I’m not really sure I understand your questions.

Please note that the zero-valued WAMIT coefficients are only useful when following the process defined above for calculating the linear 6x6 stiffness matrix of the mooring system. The zero-valued WAMIT coefficients should not be used for time-domain simulation of floating offshore wind turbines.

Best regards,

What i mean is that in my system the pretension is very high that the structure response is not too much . But when i am doing a coupled analysis using hyrodyn model in the .1 and .3 files when i m using values or either zeros. i am getting the response which is increasing with time , and according to me this should not happen .
Now i want to know that in hyrodyn model which are the important parameters which will affect the responses.
As , i already mention in previous post that i run two types of simulation.
(1) with all the data in .1 and .3 files (i.e added mass for different periods and forces)
(2) only using -1 and 0 period added mass in .1 file and .3 file is empty (i.e no forces the file is empty )
and i am getting the same results .

Dear Anuj,

To get accurate results from HydroDyn, the following WAMIT data must be provided:

*The added mass matrix at infinite frequency (period = zero) in the *.1 file.
*The damping matrix from low frequency to high frequency in the *.1 file (the damping should approach zero at both frequencies).
*The wave excitation force vector over frequencies where there is wave energy in the *.3 file.
*The hydrostatic restoring matrix in the *.hst file.

You should not get the same result between your cases (1) and (2). I suspect there is some problem with your model unrelated to the *.1 and *.3 files, but it is difficult for me to guess what that problem might be.

Best regards,

Dear Jonkman,
I want to know that when we are using the Hydrodyn module in FAST . we have to make a Platform.dat file, in that file we have to give mass and inertia of the platform. so we have to give the inertia’s of only platform or platform+wind turbine .

Dear Anuj,

FAST input parameters PtfmCM, PtfmMass, PtfmRIner, PtfmPIner, and PtfmYIner are all values pertaining only to the floating platform (without the tower, nacelle, rotor, or moorings), which is treated as a rigid body in FAST.

Best regards,

Dear Jonkman,
I would like to know that ,can we do a coupled analysis in FAST with a morison model. Actually my floater (TLP) comes under morison region so the hydrodynamic software what i am using calculates the results by using morison formulation. And according to me what i learn is that FAST works on Diffraction formulation . Then how can we solve this, please suggest some ideas.
Thanking you

Dear Anuj,

We have developed a new version of HydroDyn that allows for an arbitrary specification of Morison elements for multi-member substructures. The basic features of this version include:
*Multiple members and intersecting members at joints (including accurate calculation of overlap of intersecting members)
*Inclined and tapered members
*Flooded and ballasted members
*Marine growth
*Morison-based inertia, added mass, and viscous drag loads
*Static buoyancy and dynamic pressure loads

We are currently performing unit tests and verifying its solution (uncoupled to FAST). We then plan to couple it to FAST and verify the coupled solution over the summer, with a full public release by the end of September.

Best regards,

Dear jonkman,
I would like to know that can we do the analysis without mooring using Hydrodyn module. I tried without mooring but it gives error .
And would also like to know what is this error refers to " Unstreched length of mooring is large " whether i am putting the input correctly.

Dear Anuj,

I wouldn’t suggest simulating a floating offshore wind turbine without moorings – without moorings, the rotor thrust would cause continual drift of the platform! And for some floating platforms (e.g., tension-leg platforms (TLPs)), the mooring system is very important to the overall hydrostatic stability of the platform.

Please see my Apr 20, 2011 post in the forum topic found here for information on how to deal with the error regarding “unstretched mooring line length too large”: http://forums.nrel.gov/t/raos-in-fast/353/1.

Best regards,

Dear jonkman,
i m asking that if the floating structure is stable without mooring . Then , can we do the analysis without mooring.?? Actually i want to see something why my hydrodynamics is not matching with other hydrodynamic software.
When i m deleting the mooring input from the .dat file it is giving errors.
can you please suggest me something.
Actually myself and my guide wants to match the hydrodynamics of FAST and some other hydrodynamic software and which is not matching , so in that case we want to do a problem in which platform is stable without mooring . then may be we can know what is the problem in matching. That’s why i want to know CAN WE REMOVE THE MOORING LINES??

Dear Anuj,

Yes, FAST does allow you to model a floating wind turbine without mooring lines. My prior comment was simply to warn you that I wouldn’t recommend doing if the platform requires moorings for hydrostatic stability or if there is wind excitation enabled.

To remove all mooring lines from a FAST model, set NumLines to 0 and remove all lines from the table (leaving the two rows of column headers) in the “MOORING LINES” section of the platform input file.

Best regards,

Dear jonkman,
Please reply for my above post.
Thanking you.

I am running FAST 7.01.

FAST crashed. I have traced the error back to a call to the function Catenary in which the allocatable array LNodesX was not allocated and therefore could not be indexed. This was because the integer variable LineNodes was set to zero by default in FAST_Mods. There is also a duplicate local definition of the variable LineNodes in FloatingPlatform. LineNodes is not defined anywhere in FltngPtfmLd but is passed to Catenary as the 12th argument. LineNodes is enumerated (equal to LineNodesIn) in InitFltngPtfmLd but I cannot find where it is declared. Is this in a module which has been included by calling a USE statement?

Where does LineNodes get passed from or initialized?

Is it intended as a future user-defined variable?

I can see the sense of offering output at a user-specified number of locations along mooring lines and setting up a default number of zero until implemented in the future. I can also see that it ought to be possible to compute the mooring forces with LineNodes = 0. So why is it not possible to run HydroCalc in debug mode?

Kind Regards,

Mark

Hi, Mark.

I assume the problem you’re having is this one: http://forums.nrel.gov/t/compiling-fast/411/12

Using the array syntax (:slight_smile: to indicate “use the entire array” causes an issue with the array bounds checking (a feature enabled in debug mode) when the array is allocated to length 0 (you say it’s not allocated, but I think it is allocated to size zero). I don’t have the source code in front of me now, but I believe if you remove the “(:)” on the arrays sent to the Catenary routine, you should be able to run the code in debug mode. I fixed that problem in FAST v7.02.00d-bjj, so you might want to see how that differs from the version you’re using.

Dear Jonkman,

I am making the comparison of 1st and 2nd order wave excitation for the Hywind spar-buoy.
I read the article “Assessment of First and Second-Order Wave Excitation Load Models for Cylindrical Substructures” and I would ask you why you compute the wave forces " by setting the drag coefficient to zero and by disabling all structural degrees of freedom."?
Is this the way to compute the wave forces in FAST?

Thank you.

Best Regards,
Gabriele