internal axial forces and bending moments acting the support

Wind & Water Structural Analysis
1

Dear,
I’m a biginner on the use of FAST.
So, I would like to do a fatigue analysis on the connexions of the support structure of the floating wind turbine OC4DeepCwindSemi.
For that I want to calculate the internal axial forces and bending moments acting the support structure members.
I would like to know if Fast calculate them. Because I didn’t found them on the output items of the hydrodyn module.

Thank you for your help.
Regard

2

Dear Amirouche,

Because FAST currently models a floating platform as a rigid body, with no DOFs for structural flexibility, it is not currently possible to calculate or output internal (reaction) loads within the floating platform.

HydroDyn is a hydrodynamics module, so while it can calculate and output the hydrodynamic loads applied on a floating platform, it cannot calculate internal (reaction) loads.

Adding DOFs for structural flexibility of the platform, and the corresponding member-level internal (reaction) loads, is on our future “to-do wish list” for FAST, but is not something we are currently funded to work on.

Best regards,

3

Dear Jason,
Thank you for your answer.
I would like to calculate that internal axial forces and bending moments acting the support structure members by using a FEM model subjected to the loads calculated with Fast.

On the “HydroDyn User’s Guide and Theory Manual, Appendix C”, I took a look on the calculated hydrodynamic loads and I didn’t found the total hydrodynamic distributed loads applied on the members of the flaoting platform.
Perhaps I didn’t understand very well the available output loads. If that loads are not available, could you please, give some advice how to use the available output data in FAST to feed my FEM model to calculate the internal axial forces and bending moments.

Thank you for your help.
Best Regards

4

Dear Amirouche,

At this time, HydroDyn can output the total hydrodynamic loads at the platform reference point, as well as the individual force contributions (viscous, fluid-inertia, buoyancy, etc.) from strip theory at various nodes of various members. If you need, the total hydrodynamic load from strip theory at various nodes of various members, you must compute these manually e.g.:

Total strip theory load at node β of member α along xi = MαNβFDxi + MαNβFIxi + MαNβFBxi + MαNβFBFxi + MαNβFMGxi + MαNβFAMxi + MαNβFAGxi + MαNβFAFxi

Adding an output for the total hydrodynamic load from strip theory at various nodes of various members is on our HydroDyn to-do list.

Best regards,

5

Dera Jason,

Thank you for your help.

If I understand , there are no way to have the hydrodynamic loads (hydrostatics, diffraction from incident waves, radiation from waves generated by platform motion) applied to the members ?

So, it’s impossbile to calculate the internal axial forces and bending moments on the members on the floating platform by FEM model using loads calculated by FAST.

Best regards,

6

Dear Amirouche,

I’m assuming you are referring to the potential-flow solution of HydroDyn. In this solution, the loads (forces/moments) calculated by HydroDyn are lumped at the WAMIT reference point; they are not applied as pressure distributed across the platform/members.

That said, the potential-flow solution requires as input to HydroDyn frequency-dependent hydrodynamic coefficients (e.g. the added mass matrix, radiation damping matrix, and wave-excitation load vector), as derived from a frequency-domain wave-body interaction panel code e.g. WAMIT. WAMIT itself computes the frequency-dependent pressure distributed across the platform to calculate these hydrodynamic coefficients. So, it is feasible that the pressure could be back-calculated using WAMIT from the HydroDyn output, however, I not worked out the details of this calculation myself. This development is on our long-term “to-do wish list” for HydroDyn, but we are not currently funded to work on that development.

Best regards,

7

Dear Jason,

I am doing a very similar work of Amirouche’s, also regarding the DeepCWind.

I also want to obtain the hydrodynamic loads distributed along the platform, and for that I understand I must use strip-theory, since potential flow calculations only output the loads on the WAMIT reference point, right?
If what I wrote above is correct, I need the matrix with all the hydrodynamics coefficients of the members of the structure (MEMBER-BASED HYDRODYNAMIC COEFFICIENTS), that is HydroDyn input. Do you have these values or do you know where can I find them? I know some are on the Definition document of the semi-submersible from NREL (60601), but it does not have them all. It lacks, for example, all the pressure coefficients.

All the coefficient have the indexes 1 and 2, for example: MemberCd1 and MemberCd2. What do the indexes mean?

I also have another question regarding the coefficients. In the HydroDyn input (input document line 93), where it is defined the drag coefficient for the base columns in z-direction, it has a value of 9.6, exactly the double of what is indicated on the definition document of the semi-submersible. Can you explain me why that happens?

Thank you in advance for your time,
Best regards,
Diogo Dias

8

Dear Diogo,

As I mentioned in my post dated Mar 24, 2016 above, with the potential-flow solution in HydroDyn, you may be able to get the hydrodynamic pressure using WAMIT, but I have not done this myself.

You can use a pressure coefficient of 1.0 for the all members of the DeepCWind semisubmersible.

In the member-based hydrodynamic coefficients, 1 and 2 identify the starting and ending joint of the member, respectively (to include linear variations of the coefficient along the member, if desired).

The OC4-DeepCwind semisubmersible documentation describes the heave-plate drag in terms of the cross-sectional area of a double-sided flat plate. However, HydroDyn is more sophisticated and models different areas on the top and bottom of the offset base columns. To avoid this complication in our HydroDyn model, we treated the viscous drag on only the bottom side of the offset base columns, and so, the viscous drag coefficient was doubled to accommodate this one-sided treatment.

I hope that helps.

Best regards,

9

Dear Jason,

Thank you very much for your fast reply. It is of much help.

With what you have told me, now I mostly have all the coefficients needed except for the axial added mass. Since the OC4-DeepCwind semisubmersible documentation only refers to the base columns axial added mass coefficient, I do not have them for the remaining members. What values do you recommend for the axial added mass coefficients? Or should I leave them as zeros?

Once again thank you very much for your help.

Best regards,
Diogo Dias

10

Dear Diogo,

See section III.C of the following paper for a description of the Morison-only model that we developed at NREL, including the hydrodynamic coefficients in the axial direction: nrel.gov/docs/fy15osti/63116.pdf.

Best regards,

11

Dear Jason,

Once again thank you for your help.

Regarding the distributed loads (forces and moments) applied on the semisubmersible, I read on hydrodyn manual that it refers to loads applied along the member. I wonder if you could explain it to me better because it is not very clear for me how they are defined. I will try to explain my doubts the best I can:

  1. Imagine I define 9 nodes per member, one at each end and all equally spaced. As the loads are defined per unit length, how exactly they are defined? For example, for the loads on the first node, the value means that there is a load of that magnitude (considering of course it is per unit length) along the member from the first node to the second? This is the way I can imagine they are defined, but on the other hand it does not make sense if we use this logic until the 9th node because it would not define the load from the 9th to the 10th , as the 10th does not exist. Basically, I wonder how they are exactly defined so I can make all the necessary considerations needed for a FEM structural analysis? How do I have to process this loads?
  2. For this FEM analysis I will only use moments or forces, right? Considering them both on the same FEM analysis would somehow “double” the loads the structure is subjected to, am I correct?

Thank you in advance for all your help,
Best regards,
Diogo Dias

12

Dear Diogo,

Regarding loads calculated per unit length along the member, you can assume a linear interpolation of the load along the element between two nodes. For example, if the x-component of the load was 4 N/m at node 1 and 8 N/m at node 2, then the load on the element half-way between these two nodes would be 6 N/m.

If you are referring to applied loads (as opposed to reaction loads), then it is not double booking to take both the applied force and applied moment for use in your FEM analysis. The loads calculated within HydroDyn are hydrodynamic loads applied to the structure, not the reaction loads felt within the structure.

Best regards,