Platform Properties

Thanks for your reply Dr. Jonkman.

My platform design comprises of three concrete-ballasted steel cylinders (height: 16m, diameter: 10m), which are arranged in triangular geometry such that the distance between axis of any two cylinders is 40m. They are connected to each other by appropriate truces. The bottom of the wind turbine tower is situated at the centroid of this triangular geometry. The center of mass of this structure is located 8.38m below the centroid.

For your reference, “Cylinder-Model.pdf” is the entire structure and “Cylinder-Model 1.pdf” and “Cylinder-Model 2.pdf” give the dimensions of the structure. Kindly guide me how to calculate platform roll, pitch, and yaw inertias; and platform draft, effective diameter for this geometry. Please let me know if you need further information regarding the same.

Thanks a lot for your help.

-Nikunj.
Cylinder-Model 2.pdf (92.2 KB)
Cylinder-Model 1.pdf (53.1 KB)
Cylinder-Model.pdf (48.3 KB)

Dear Nikunj,

You should be able to calculate the mass moment of inertias through a CAD program; although the geometry of the 3-column semisubmersible is simple enough that you may be able to calculate it by hand.

The platform draft, effective diameter, and drag coefficient inputs available in HydroDyn are not useful for modeling the viscous drag on a 3-column semisubmersible. We are working on an upgrade to HydroDyn that will permit one to model the viscous drag on all elements of the floater (including the main columns and truss structure of a semisubmersible). Until that feature is available, I would disable the viscous drag calculation by setting PtfmNodes, PtfmDraft, PtfmDiam, and PtfmCD all to zero.

Best regards,

Thanks for your inputs Dr. Jonkman!

Dr. Jonkman, how would disabling of the viscous drag calculation (by setting PtfmNodes, PtfmDraft, PtfmDiam, and PtfmCD all to zero) affect the result in fast output file? Will the deflection and other physical properties’ values obtained in fast output file still be reasonable?

Thanks,
Nikunj.

Dear Nikunj,

Of course, if viscous effects are important to the response of a given floating platform, simulating without viscous effects would not lead to accurate solutions. In a semisubmersible for example, the viscous drag on the heave plates is likely very important to the overall dynamic response. Until the full implementation of viscous drag elements becomes available in HydroDyn, including an “additional linear (or quadratic) damping matrix” for the platform DOFs would be a suitable option for enabling FAST/HydroDyn to model the proper response. Such an “additional damping matrix” can be derived from known free-decay response data of the given platform.

However, the standard version of HydroDyn is not currently set up to allow a user to specify additional damping matrix coefficients directly through an input file. This feature is not too difficult to add yourself, although it will require a recompile of FAST. Please see the topic on our design codes forum found here for guidance on how the “additional linear damping” was added to HydroDyn for the OC3-Hywind system: http://forums.nrel.gov/t/additional-linear-hydrodynamic-damping-in-oc3-hywind/383/1.

I hope that helps.

Best regards,

Dear Dr. Jonkman!

Thanks for your inputs on my previous query.

I went through the link you posted in your reply, and am now planning to use OC3 Hywind files after modifying them according to the design i showed you in my previous messages. Will the new codes in OC3 Hywind files work for my design after changing appropriate design parameters such as inertias, et. al.?

Also, I am planning to incorporate few changes in the OC3Hywind.fst file to suite my analysis and would post a query if i get an error.

Thanks a lot!

Regards,
Nikunj.

Dear Nikunj,

I’m sorry, but I don’t understand your question. What do you mean when you say, “will the new codes in OC3 Hywind files work for my design after changing appropriate design parameters such as inertias, et. al.?”

Best regards,

I am sorry for that.

I downloaded the files from the link: wind.nrel.gov/public/jjonkman/NR … Hywind.zip, as posted by you in the topic http://forums.nrel.gov/t/additional-linear-hydrodynamic-damping-in-oc3-hywind/383/1. My platform design is a floating design but the one used in above analysis is a spar buoy design. I am planning to modify the platform file, “NRELOffshrBsline5MW_Platform_OC3Hywind” by changing the parameters such as TwrDraft, PtfmCM, PtfmRef, PtfmMass, PtfmRiner etc. to suite my design.

I was just wondering if i can get reasonable outputs for my design my making aforementioned modifications.

Hope this explanation better conveys my question.

Thanks for your time and swift replies.

-Nikunj.

Dear Nikunj,

The FAST model is parameterized, so, changing input parameters allows you to model different turbine configurations.

That said, as has been discussed in this forum topic, the current HydroDyn module of FAST is limited in its ability to model the hydrodynamic loads on 3-column semisubmersibles. That is, HydroDyn cannot currently model the viscous drag on the various elements (braces, columns, heave plates) of a semi. (HydroDyn can currently model the 1st-order hydrostatics, radiation, and diffraction loads of a semi.) This is why I sent you the link to the OC3-Hyind topic in the first place–to show how you can modify FAST so that you can augment the existing hydrodynamic calculations with an “additional” damping matrix if the viscous effects are important for your design. Please note that the FAST executable included in the OC3-Hywind topic has a costumization that is only appropriate for OC3-Hywind system. If you are not modeling that system, you should not use that executable.

Best regards,

ok.

So, if i use WAMIT to get the viscous properties for my platform design (and then input the output of WAMIT to fast), will i still need to add the damping matrix?

Thanks,
Nikunj.

Dear Nikunj,

WAMIT is a tool based on potential flow, so, it cannot model viscous effects. The linear potential flow solution (radiation, diffraction, hydrostatics) can already by imported from WAMIT to FAST.

The “additional damping” would be needed to capture the viscous effects.

Best regards,

Thanks a lot for your help and patience Dr. Jonkman. Inputs provided by you were really helpful.

Regards,
Nikunj.

Hello Dr. Jonkman!

As I mentioned earlier, I am modifying the PtfmRIner, PtfmPIner PtfmYIner of the “NRELOffshrBsline5MW_Platform_ITIBarge4” file to perform an analysis for 3-column semisubmersible design; and as suggested by you, I am also planning to perform viscous calculation for this design by adding the damping matrix for this design.

I wanted know to whether this approach- of modifying the parameters of “NRELOffshrBsline5MW_Platform_ITIBarge4” file- is a good way to go forward and whether it will yield reasonable results.

Thanks a lot for your time!

Regards,
Nikunj.

Dear Nikunj,

Yes, I always recommend users modify an existing input file when developing a new FAST model rather than trying to develop one from scratch. Starting with the ITI Energy barge model is fine for your purposes.

Best regards,

Hello Dr. Jonkman!

WIth the help of “Dynamics Modeling and Loads Analysis of an Offshore Floating Wind Turbine”, I am manually calculating the elements of linear hydrostatic matrix. Would you please tell me how do I calculate C35 (which is integral of x*dA) [on page 26].

Also, I see that equation 2-21 of the document gives the equations for calculating individual elements of the the hydrostatic restoring matrix. Would you please let me know similar equations for calculating the elements of Hydrodynamic added mass matrix (Aij) and Damping matrix (Bij)?

Thanks a lot!

Nikunj.

Dear Nikunj,

Element (3,5) of the hydrostatic restoring matrix is just the first area moment about the x-axis of the body cross section at the still water level. This is equivalent to the area centroid times the area of the body cross section at the still water level.

The frequency-dependent hydrodynamic added mass and damping matrices of general bodies cannot be expressed so simply. Instead, you should use numerical software such as WAMIT for this calculation. (WAMIT can also solve for the hydrostatic restoring matrix, although this is trivial.)

Best regards,

Hello Dr. Jonkman!

Would you please tell me the model specific customization that I will have to make in HydroCalc.f90 in order to specify additional matrices (and recompile FAST)for the design i specified earlier in this topic.

Thanks a lot.

Regards,
Nikunj.

Hello Dr. Jonkman!

As you are aware, I am working on the three-cylinder submersible platform, and in order to generate the damping and added mas matrices, I am using WAMIT. My seniors- who worked on this project earlier- mentioned in their report that they received WAMIT’s *gdf file from you for their platform design. I do not have the aero-hydro program through which i can input the geometry via a CAD program. I would be really grateful to you if you could provide the *gdf file for the platform design I have attached below.

Thanks a lot for your continuous help and assistance.

Regards,
Nikunj.
Cylinder-Model 2.pdf (92.2 KB)
Cylinder-Model 1.pdf (53.1 KB)
Cylinder-Model.pdf (48.3 KB)

Dear Nikunj,

Please see my July 14, 2011 post in the forum topic found here for information on how the “additional linear damping” was added to the OC3-Hywind spar buoy model: http://forums.nrel.gov/t/instructions-for-compiling-fast/514/1.

We can provide you WAMIT input files (e.g., *.gdf) for models in the public domain that we have already generated (such as the for the OC4-DeepCwind semisubmersible), but NREL does not have the resources to develop input files for your own unique design.

Best regards,

Dear Jason,

The thread above relates to a symmetrical floating platform. I am investigating the feasibility of representating an assymmetric platform in FAST.

We are considering building a model of a wind turbine on a floating platform in which the tower interace is offset horizontally from both the platform centre of mass and the centre of buoyancy. I would like to evaluate the changes we would need to make to FAST in order to take account of this offset.

In FAST.F90, Line 5130, the position vector rZY of the platform mass centre (point Y) with respect to the platform reference point (Point Z), is calculated as a scalar. I presume that is because you have assumed that both points lie on the centreline of the tower and sub-structure (the whole support structure is uni-dimensional and axisymmetric). Also on Line 5518, the velocity of the platform mass centre is calculated using the cross product of the angular velocity and rZY. I assume that a2 is defined as the unit vector, [0,0,1].

Can you foresee any problems if rZY is redefined as [rZYxt, rZYyt, rZYzt] so that there is an additional arbitrary horizontal offset of the mass centre [rZYxt, rZYyt, 0]?

Many thanks.

Kind Regards,

Mark

Dear Mark,

I’ve answered a nearly identical question in my May 24, 2013 post in the forum topic found here: http://forums.nrel.gov/t/tower-centerline/684/1.

This change will be included in the upcoming release of FAST.

Best regards,