OC3-Hywind RAOs

Dear Lorenzo,

I haven’t checked all of the off-diagonal inertia terms, but at least the diagonal entries of your inertia matrix, as well as the mass and center of mass, make sense to me.

Best regards,

Dear Jason,
The diagonal terms return to me too, while I have doubts about the off-diagonal terms of inertias. In fact, in my model I get accelerations in x, y, z right while rx, ry, rz do not coincide very well.
Can you please check them?
Thanks for your help,
Best regards,
Lorenzo.

Hi Lorenzo,

Yes, I agree with your numbers. Here are my calculations:

[code]>> M_swl = [ [ 8.0660481545E+006 0 0 0 -78.0013014648.0660481545E+006 -1.1132995542E-0028.0660481545E+006 ];
[ 0 8.0660481545E+006 0 78.0013014648.0660481545E+006 0 -1.3918125391E-0028.0660481545E+006 ];
[ 0 0 8.0660481545E+006 1.1132995542E-0028.0660481545E+006 1.3918125391E-0028.0660481545E+006 0 ];
[ 0 78.0013014648.0660481545E+006 1.1132995542E-0028.0660481545E+006 6.8025977471E+010 9.9135213851E+006 -8.6878157798E+006 ];
[ -78.0013014648.0660481545E+006 0 1.3918125391E-0028.0660481545E+006 9.9135213851E+006 6.8022535284E+010 6.965876724E+006 ];
[ -1.1132995542E-0028.0660481545E+006 -1.3918125391E-0028.0660481545E+006 0 -8.6878157798E+006 6.965876724E+006 1.9157300092E+008 ] ]

M_swl =

1.0e+10 *

0.000806604815450 0 0 0 -0.062916225372230 -0.000008979927815
0 0.000806604815450 0 0.062916225372230 0 -0.000011226426962
0 0 0.000806604815450 0.000008979927815 0.000011226426962 0
0 0.062916225372230 0.000008979927815 6.802597747100000 0.000991352138510 -0.000868781577980
-0.062916225372230 0 0.000011226426962 0.000991352138510 6.802253528400000 0.000696587672400
-0.000008979927815 -0.000011226426962 0 -0.000868781577980 0.000696587672400 0.019157300092000

TransMat = [ [ 1 0 0 0 78.001301464 1.1132995542E-002 ];
[ 0 1 0 -78.001301464 0 1.3918125391E-002 ];
[ 0 0 1 -1.1132995542E-002 -1.3918125391E-002 0 ];
[ 0 0 0 1 0 0 ];
[ 0 0 0 0 1 0 ];
[ 0 0 0 0 0 1 ] ]

TransMat =

1.000000000000000 0 0 0 78.001301463999994 0.011132995542000
0 1.000000000000000 0 -78.001301463999994 0 0.013918125391000
0 0 1.000000000000000 -0.011132995542000 -0.013918125391000 0
0 0 0 1.000000000000000 0 0
0 0 0 0 1.000000000000000 0
0 0 0 0 0 1.000000000000000

M_cg = TransMat’M_swlTransMat

M_cg =

1.0e+10 *

0.000806604815450 0 0 0 0 0
0 0.000806604815450 0 -0.000000000000000 0 0
0 0 0.000806604815450 0 0 0
0 0 0 1.895050184890263 0.000991227154749 0.000006894335879
0 0 0 0.000991227154749 1.894705909912941 -0.000003858384188
0 0 0 0.000006894335879 -0.000003858384188 0.019157043867686[/code]
Best regards,

Thanks a lot for your prompt reply.
Best regards,
Lorenzo.

Dear Jason,
I report here my problem because despite using the right mass matrix my model does not work.
I noticed that in the case of offshore turbines the two predominant forcings are the loads at the base of the turbine tower and the moorings (which must compensate for the tower loads).
In my model the loads at the base and the moorings add up (with those relating to hydrodynamics, restorings, etc. which are smaller) and, divided by the mass matrix, give the accelerations, from which I get the positions to feedback and obtain the loads of the mooring (via Map ++).
The problem is that Fx and My of the tower start with a step, which if too large leads the moorings to diverge rather than converge and compensate for these loads. To demonstrate this, I tried to manually enter the average value of the forces at the base of the turbine tower, and as can be seen in Figure 1, beyond a certain value of Fx the mooring diverges (at 2 * 10 ^ 5 N converges, while at 3 * 10 ^ 5 N diverges).


I asked you about the off-diagonal moments because I noticed that it is not so much the x position that makes Fx_moor diverge, but rather ry which is much higher than the value obtained with FAST.
By setting an increasing trend instead of a step (Figure 2) it is possible to reach higher values ​​of Fx (up to 10 ^ 6, a value higher than the Fx_tower values ​​of the tests carried out), however this would mean making up the first instants of Fx at the base of the turbine.

How does FAST prevent this problem?
Thanks for the reply, best regards.
Lorenzo.

Dear Lorenzo,

I’m not sure I can really comment on the results from your model, which I’m not familiar with.

What does FAST / OpenFAST do for this same load case?

Please note that FAST / OpenFAST include many nonliearities that may not be present in your model. A related discussion in the following forum topic may provide some insight: http://forums.nrel.gov/t/hydrodynamic-implementation/2284/1.

Best regards.

Hi everyone

I’m a Ph.D. student studying concrete floating offshore wind turbines. (using WAMITv7 and OpenFAST.)

I quite don’t understand the 2nd process.
(2)Redefine the added mass and damping as zero in the file named “spar.1”

As the spar.1 is the output from WAMIT, does this mean the zero values of added mass and damping will be input at WAMIT *.frc files then do the analysis again, or it will be input at PLATFORM ADDITIONAL STIFFNESS AND DAMPING of OpenFAST HyDroDyn?

Thank you in advance for your kindly answer.
Wichuda

Dear Wichuda,

The steps outlined by Minxi where to obtain the rigid-body mass matrix of the system, not including hydrodynamic added mass. In OpenFAST, it is possible to zero out the added mass completely, but this was not possible in FAST v7. So, the workaround was to specify WAMIT output files with all coefficients set to zero. These are not obtained by running WAMIT; instead, the files were created by hand to look like they were WAMIT-generated. I’ve attached the files to my post dated Oct 30, 2012 in the following forum topic: http://forums.nrel.gov/t/questions-about-the-hydrodyn-module/596/4.

Best regards,

Dear Jonkman

Thank you very much. Now I got it.

Regards,
Wichuda