Many thanks Jason.
Regards,
AOAW
Hi Jason:
I have carried out a simulation by comparing the OC4 DeepCwind model (200 m deep water) against the DeepCwind model (NREL) that I have developed for 100 m deep water (UoP). The sea state used in both models is Hs, 1.76m, Tp, 0.8s (Jamaica Sea conditions). I have used a JONSWAP spectrum.
I have compared the motions for the respective degrees of freedom and I am not certain about the outputs for the heave and the surge degrees of freedom. More so, the surge degree of freedom. I have attached the plots.
Heave motion
From the plots, the general motion in heave looks the same. However, for NREL model it oscillates steadily around zero. For UoP 100m model it oscillates around - 0.25 m. I am not not sure what to make of the shift in the average have value from zero as per the NREL model to 0.25 m for the UoP model. Can you please provide some clarification?
Surge motion
For the NREL model, it oscillates around zero for wave only. This is the same behaviour for UoP model. However, when wind is applied, UoP model shows same oscillation about zero. However, NREL model oscillates about the original perturbation of 5m with maximum value of 7.0 m (after model stabilized) for turbulent wind condition. Whereas maximum value for UoP was about 2.1 m. Can you please say what could be the reason (s) for this?
Thank you.
Dear Andre,
Regarding surge, the primary surge restoring is from the mooring system. It looks like your new moorings for 100-m depth are quite a bit stiffer in surge than in the original model.
Regarding heave, the lower mean is likely the result of higher mooring pretension in the 100-m depth case relative to the original model. In general, it is recommend that a FOWT model in OpenFAST be set up so that the mean heave (or static heave in the absence of external forcing) is zero. This can be achieved by ensuring that the full system weight and vertical mooring pretension balances with the undisplaced buoyancy of the floater.
Best regards,
Jason:
Thanks for the prompt response. You said the following about achieving a mean of zero for a FOWT system in OpenFAST:
“This can be achieved by ensuring that the full system weight and vertical mooring pretension balances with the undisplaced buoyancy of the floater.”
How exactly is this done? Would be very grateful for some ideas.
My thoughts are that I do not want to change the weight of the system and that I should play with the mooring line properties until the zero mean is achieved in heave.
Regarding the vertical mooring pretension:
Is it the tension at the fairlead computed by, say, MoorDyn?
I had wondered if this mooring pretension is pre-determined by the manufacturer. I am guessing it could be considered this way because the mooring line would be manufactured for a maximum pre-tension.
Thank you.
Regards,
AOAW
Dear Andre,
Assuming you don’t intend to change the floater design, then I agree that changing the vertical mooring pretension is the one to focus on (a change to the mooring input file). For catenary moorings, the vertical mooring pretension is dominated by the weight of the mooring lines not resting on the seabed, minus their buoyancy, and so, would be most sensitive to the line mass density, line diameter, and unstretched line length.
Best regards,
Thanks Jason.
Regards,
AOAW
Jason:
I adjusted the mooring line using a combination of the dia., mass density and unstretched length and it now tuned such that mean heave is zero. However, I have a query about the connection properties and unstretched length of the mooring line. I am using the properties for the OC4-DeepCwind model to raise the queries, see attachment.
I notice that the X connection points of the mooring line are such that one is 837.6 m and the other 2 are 418.8 m (1/2 x 837.6). I have applied the same principle for the 100 m deep water mooring line. I understand that the 837.6 m is the distance from the vertical centreline axis to the anchor.
1)Does this mean that one of the mooring line (call in ML1) is assumed to have a perturbation such that the dimension of the other two from the centreline to the anchor is 0.5 x ML1, although the lengths of all three mooring lines are the same?
I had read in another forum where you stated that the unstretched length of the mooring line is the natural length before tension is applied; so, that after tension is applied it is longer. My queries are:
2) Is the unstretched length the portion that is not resting on the seabed? I am trying to visualize what is really happening when the unstretched length is adjusted and how it compares to the overall length of the line. For example, at node 2, length from centreline to anchor is 837.6 m and unstretched length is 835.35 m. That’s a difference of 2.25 m. Does this mean that 2.25 m length is resting on the seabed?
In addition, in a physical sense, does changing unstretched length mean that I am trying a new mooring line each time with a new natural length, although the distance X from the centreline is unaltered? No change was made to connections 1, 2 or 3 in X, for the change in unstretched length. Please clarify.
Thank you.
Regards,
AOAW
Dear Andre,
Regarding (1), in the OC4-DeepCwind semi, the three anchor locations are located on the horizontal seabed such that they form a circle of radius 837.6 m and are spread evenly, each 120-deg apart from the other. The (X,Y) locations for anchor “i” are thus:
X(i) = -837.6COS( 120degi )
Y(i) = -837.6SIN( 120degi )
Regarding (2), no. The unstretched length is the total length of the mooring line in the absence of tension. In the OC4-DeepCwind semi, the distance from the anchor to the fairlead is Horizontal = (837.6-40.868) = 796.732 m and Vertical = ( 200 - 14 ) = 186 m (using data from the MoorDyn input file), or a total distance (hypotenuse) of 818.1552 m. Given the unstretched length of 835.35 m, the line is slack (a catenary).
Best regards,
Jason:
Thank you.
If I understand 1) correctly i = 0,1,2. Using those values I get the coordinates. Also, I am going to assume that based on the convention, a negative sign was placed in front of the radius of the circle.
Regards,
AOAW
Dear Andre,
Yes, your understanding is correct.
Best regards,
Thanks Jason.
Regards,
AOAW
Hi Jason:
I saw something in the results for the OC4 DeepCwind model which I found very peculiar and would be very grateful if you could clarify.
The blade pitch angle, BldPitch1, remained at zero for the entire simulation. I, however, thought it would have changed throughout the simulation. Is it that the blade is not utilized for turbine control mechanism in this case? Would be very grateful for some clarification.
Thank you.
Regards,
AOAW
Dear Andre,
I’m not sure what your simulation set up is, but the blade pitch should change if you are running an operating turbine with hub-height wind speeds between rated and cut-out.
Best regards,
Noted Jason. I will check it again.
Thanks.
Regards,
AOAW
Jason:
I have checked again. The simulation I am running is the original OC4-DeepCwind model. I have made no modifications. However, I realize that the initial pitch angle that was set in ElastoDyn as zero, does not change for the entire simulation. I modified ElastoDyn to use pitch angle of 0.1 and it defaults to zero after the first time step and remains the same.
I am not sure, why this happened as I was expecting the pitch angles to change. I also checked the FASTv8 OC4-DeepCwind model (Test25) and the results were the same. I have attached the outputs for both pitch angle of zero and 0.1 (see cells highlighted in yellow) and the related ElastoDyn file(0.1 pitch angle only attached).
I have also checked the ServoDyn file and the control is set to PCMode 5. So I am not sure.
Please let me know if you can think of any other reason why this would be so.
Thank you.
Regards,
5MW_OC4Semi_WSt_WavesWN_test1.xlsx (433 KB)
5MW_OC4Semi_WSt_WavesWN_test.xlsx (433 KB)
Dear Andre,
The model you running uses a steady, uniform wind speed of 8 m/s (also seen in your results in column B), which is below rated, where blade-pitch control is not active. The blade-pitch angles should remain at 0deg for this simulation.
Best regards,
Jason:
Many thanks for that clarification.
Regards,
AOAW
Hi Jason:
I have a query regarding peak shape parameter of incident waves, WavePkShp. When it is set to 1.0 that is for a Pierson-Moskowitz specturm. When set to “Default” that is for the JONSWAP spectrum. 1) What is the value of WavePkShp in that case? Is it 3.3?
2) Is there a range of values that can be used for WavePkShp for JONSWAP? I came across a DLC in one of the offshore wind turbine manuals where a WavePkShp of 2.5 was used.
Thank you.
Regards,
AOAW
Dear Andre,
Here are my answers to your questions regarding WavePkShp:
-
When WavePkShp = “DEFAULT”, the value recommended in the offshore wind turbine design standard from 2008 (IEC 61400-3 Annex B) is used, derived based on the peak-spectral period and significant wave height, which you can see evaluated in FUNCTION WavePkShpDefault() of Waves.f90.
-
WavePkShp must be set between 1 and 7 (inclusive). The higher the value of WavePkShp, the more peaked the wave spectra is.
Best regards,
1 Like