Good afternoon,
I am modeling a platform concept in FAST using a hybrid model. I used WAMIT as preprocessor for the Potential Flow Model (.1, .3 and .hst files), and I now want to add strip-theory members so that the current generates viscous drag forces on the platform. To adjust the viscous drag coefficients of the strip-theory members, I will compare FAST with CFD simulations.
I have already run some simulations in FAST with different current speeds and viscous-drag coefficients for the strip-theory members, and I have removed the mooring lines, so that the platform can be displaced freely. That way, I can compare the acceleration and velocity of the platform from FAST and CFD simulations. I have also disabled Aerodyn so that no aerodynamic forces are generated on the blades, and all DOF except platform surge and sway.
However, I have observed that the platform velocity does not reach the current velocity in the simulations, just a 30-40% of it. I suspect that when the platform moves, forces opposite to the movement are generated on the potential flow model, which compensate viscous-drag forces that current generates on the strip-theory members, so that the platform does not reach current velocity. Still, I’m not sure about that, so I’d like to know if that’s the explanation or if there’s another.
On the other hand, if that is the case, what do you think is the most appropriate simulation to adjust the coefficients?
Many thanks in advance for any help,
Best regards,
Nicolás Deza
Dear Nicolás,
If I understand correctly, you are trying to run a FAST simulation where the platform velocity catches up to the current velocity…is that correct? Because of the use of the relative form of Morison’s equation in HydroDyn, the current will create a drag force as long as the current velocity is different from the structural velocity, so, I would expect that the platform velocity would eventually reach the current velocity unless other external forces are applied. You mentioned that you’ve disabled aerodynamic and mooring loads. You are correct that the potential-flow solution could be generating other forces (these would be oscillatory at the wave frequency); the additional platform linear damping (AddBLin) or quadratic drag (AddBQuad) could be yet other forces.
Regardless, normally viscous drag coefficients are calibrated through current-only simulations with the structure fixed (looking at the applied forces), or equivalently, with the structure towed at a constant velocity and the fluid stationary (as would be done in a towing tank); in both cases, there is a relative velocity between fluid and structure. Viscous drag coefficients are also calibrated through free-decay simulations. However, recent experience has shown that viscous drag coefficients really depend on the sea state, and so, are difficult to calibrate for validity across many conditions.
Best regards,
Dear Dr.Jonkman,
Thank you for the reply.
As you say, I am trying to run a FAST simulation where the platform velocity catches up to the current velocity. If I run a current only simulation with the structure fixed, can I obtain the viscous forces generated on the platform from the outputs of the simulation?
On the other hand, I have noticed that the additional platform linear damping (in surge and sway DOFs) of the potential flow model was the cause for the platform not catching the current velocity. Should I set those coefficients to zero in the model and add strip-theory members for horizontal relative velocity (between water and platform) induced viscous drag forces? (…not for this particular simulation, but for a realistic modelling of the platform)
Dear Nicolás,
Here are my answers to your questions:
Yes, you can obtain the viscous forces generated on the platform from the outputs of the simulation. In a current-only simulation with fixed substructure, the only forces are viscous drag and buoyancy. You can output the viscous drag at the node level or globally through the HydroDyn outputs–see Appendix C in the draft HydroDyn User’s Guide and Theory Manual for more information.
The additional platform linear damping and quadratic drag provide a quick way to calibrate the hydrodynamic damping. The distributed viscous drag across members and joints are likely harder to calibrate, but are likely more realistic and more tuneable.
Best regars,