Hi, I just wanted a confirmation (or otherwise) of my understanding about the coupling between hydrodynamic loads calculated in hydrodyn and the flexible elements in SubDyn.
“SubDyn also outputs the substructure displacements, velocities, and accelerations for input to HydroDyn to calculate the hydrodynamic loads that become inputs for SubDyn.”
So, through the glue-code Hydrodyn calculates the hydrodynamic loads acting on the elements described in the SubDyn model, and the glue-code imposes these hydrodynamic loads on the SubDyn elements.
My doube is this: does this apply to any element in SubDyn? I.e., beam elements, rigid link, and pre-tensioned cables?
In particular, are pre-tensiioned cables loaded hydrostatically and hydrodynamically (waves), as long as they are discretised in the hydrodyn.dat file as well?
To clarify, HydroDyn receives as input elastic motions and calculates/outputs hydrostatic and hydrodynamic loads at the hydrodynamic analysis nodes known by HydroDyn. Likewise, SubDyn receives as input loads and calculates/outputs elastic motions at the structural analysis nodes known by SubDyn. Given that these hydrodynamic and structural nodes may have different spatial discretizations, the spatial mesh-to-mesh mapping in the OpenFAST glue code handles the data transfer, ensuring that kinematic and load balances are maintained. For more information on the mesh mapping used by OpenFAST, see: https://www.nrel.gov/docs/fy14osti/60742.pdf.
In BeamDyn, members represented by beam elements may have many nodes across the member, depending on the discretization of the beam elements. However, rigid-link elements only have a single node and pre-tensioned cable elements only have two nodes (one at each joint), so, these SubDyn element types can’t support distributed hydrostatic/hydrodynamic loading like beams can.
Well, MoorDyn cannot model pre-tensioned tendons that interconnect two members within a SubDyn model if that is what you have in mind. MoorDyn can be used to connect a SubDyn member to bodies in MoorDyn or the seabed.
Thanks, Jason.
Yes, the members I talk about are members connecting SybDyn elements to the seabed.
Anyway, I fear the issue is that, when SubDyn performs the structural modal reduction, not knowing of the mooring tendon in MoorDyn, will signal some rigid DOF as unconstrainted, and put out an error.
Since the offshore wind turbine that I am modelling has one reaction point at seabed level, it is automatically considered by openFAST as a bottom-fixed wind turbine (see here).
Your link is not working for me (all I see on two different browsers is a blank page). Can you clarify what system you are modeling and what the reaction point on the seabed is?
Withouth infringing any IP, the system is basically the same (has been beefed up a bit) of the previous post. Also an image below.
Long story short, the reaction point at bed level of the main column is a universal joint, so without the 9 tension cables the whole platform will just rotate (rigidly) around that point.
Can you clarify how you are modeling this FOWT now in OpenFAST? Are you modeling all members of the substructure (foundation, universal joint, main column, arms, tendons) in SubDyn and thinking about switching the tendons from SubDyn to MoorDyn?
I would guess you’d then run into the problem discussed in other the other forum post you linked to–switching the tendons from SubDyn to MoorDyn would result in rigid-body mode about the universal joint (unless there is nonzero stifness about the joint) that could not be resolved by SubDyn. Unfortunately, SubDyn may be limited in accuracy for this specific compliant tower concept.
Many thanks for the time taken to discuss/re-discuss this, always good to double check one’s understanding… especially with the “father” of FAST!
I will give it a quick try and see if adding some small stiffness (to avoid a singularity) to the universal joint at the bottom would give me issues or not.
