Modelling of a two-bladed FWT with leaning tower and guy-wire

Dear Dr. Jason,

I would like to ask several questions about modeling with a two-bladed floating wind turbine with a leaning tower and guy wires. I find myself struggling in modeling with it.


As you can see in the figure, the key point of this FWT is that it has six guy-wires attaching to the top of its tower and the floating foundation. The wires have pretensions so that the whole structure is deformed. The guy-wire tension as well as the structural deformation will be varying at each time instant according to the changes of environmental loadings from wind, wave and current. Therefore, it is a structural-aero-hydro fully coupled problem. I would like to ask that if you have any suggestions on modeling such guy-wires using the great tool FAST. Or if not, is it able to incorporate FAST with other NREL or commercial software to resolve this problem?

The second key point is the leaning tower, which is a little bit different from the traditional vertically straight towers. Are there any differences between the two types of towers when using FAST to model them?

The third key point is the turret mooring at a single point, with this design, the FWT probably does not need a yaw control system. Is it able to model the turret mooring system in FEAMooring or MAP or MoorDyn modules in FAST v8?

Since the design is not a normal case that was commonly modeled by the FAST, I’m not sure on how to correctly model it. Is it possible to model them in FAST, or should I incorporate ADAMS or other FEM softwares? I’m very much appreciated that if you would provide some guidance or suggestion, even if it is just a hint or clue.

Many thanks.

Yingyi Liu
Kyushu University

Dear Yingyi Liu,

Modeling of the guy wires will require a change to the FAST source code, as the current version of FAST does not support this feature. In FAST v8, you could use e.g. MAP or MoorDyn as the guy wire model, but you’ve have to modify the source code to allow one end of the cable to connect to the tower-top/nacelle instead of the inertia frame (anchor).

The modeling of tower tilt in FAST has been discussed briefly for bottom-fixed systems in the following forum topic: Simulation with a tilted tower - #5 by Jason.Jonkman. But in your case, you have a floating wind system, which presumably means that you want to enable the six degrees-of-freedom (DOFs) of the floating platform. Again, you’ll need to modify the source code to consider this effect as FAST assumes the undeflected tower attaches vertically to the floating platform. You would have to modify the source code to set the position vectors and orientation of the tower coordinates to tilt relative to the platform.

Neither MAP nor MoorDyn consider rotational degrees, so if you consider all catenary lines to connect to the platform at the same point (fairlead), then the platform will be allowed to freely yaw around that point without inducing loads on the moorings. This approach would permit you to model a turret mooring without stiffness or damping. If you need to consider turret stiffness or damping or wish to space the fairleads apart, you would need to introduce a turret DOF, which would again require a change to the FAST source code.

Sounds like a fun project!

Best regards,

Dear Dr. Jason,

Thank you so much for your perfect reply. I would love to try to modify the FAST source code following your above guidance if it is not too complex for me to be finished within several months.

Could you please kindly tell me the right place in the source code where I can modify it to: (1) allow one end of the cable to connect to the tower-top/nacelle instead of the anchors, and (2) set the position vectors and orientation of the tower coordinates to tilt relative to the platform? Unfortunately, I could not find the two exact places during these days.

Many thanks.

Yingyi Liu
Kyushu University

Dear Yingyi Liu,

I don’t have the resources to offer detailed step-by-step instructions, but I’ll answer at a high level.

For (1), currently the mooring loads at the fairleads as output by MAP++ and MoorDyn are transferred in the FAST glue code as inputs to ElastoDyn to be applied on the platform via point-to-point mesh-mapping. Likewise, the motions of the platform output from ElastoDyn are transferred in the FAST glue code as inputs to MAP++ and MoorDyn to move the fairleads via point-to-point mesh-mapping. You’ll have to change this mesh-mapping so that loads output and motion input at some of the fairleads (i.e. those connected to the tower-top) get transferred to/from the appropriate tower node via the tower mesh of ElastoDyn. I would suggest that you start by reviewing how the mesh-mapping in FAST works (see: nrel.gov/docs/fy16osti/63203.pdf) and how the existing mapping between MAP++/MoorDyn and ElastoDyn are implemented.

For (2), you’re going to have to change the orientations, position vectors, velocity vectors etc. in ElastoDyn. If you are considering the aerodynamic tower influence, tower shadow, and/or tower loads in AeroDyn, you’ll have to make a similar change there. I would start by reviewing coordinate systems used internal to ElastoDyn (as set in SUBROUTINE ElastoDyn.f90/SetCoordSy().

Best regards,

Dear Dr.Jason,

Thanks for your prompt reply with great help.
I will go into details of your recommendations.

Best Regards,
Yingyi Liu

Dear Dr.Jason,

Due to some proprietary reasons, we need to assume a FAST model of a virtual two-bladed wind turbine instead of directly using the parameters from the real commercialized two-bladed wind turbine which has a rated power of 6MW. At this stage, our goal is to perform a preliminary FAST simulation under some designed conditions (not necessary to make an exactly same wind turbine).

Due to the tight schedule and the complexity of this work, I would rather degrade the well-packaged three-bladed NREL 5MW Baseline Turbine to a two-bladed one (with a rated power around 5~7 MW), than designing a completely new wind turbine. For doing so, from your point of view, what modifications should I make on the existing FAST input files of the NREL 5MW Baseline Turbine? I would be very much appreciated if you could list several points in detail.

In addition, do you think such degrading physically makes sense?

Thanks for your point of views in advance.

Best Regards,
Yingyi Liu

Dear Yingyi,

The change to the NREL 5-MW turbine from three to two blades will likely not be as simple changing FAST input NumBl from 3 to 2. Normally, blade design would have to change (e.g. increased chord) to ensure appropriate rotor performance and blade strength to go from a three- to a two-bladed rotor, with associated changes to the structural, aerodynamic, and controller properties. This change is beyond what I have resources for to comment on.

Best regards,

Dear Dr. Jason,

For the pre-mentioned Nezzy wind turbine (see the picture of our first discussion), I’m thinking that if it possible to couple the ElastoDyn, SubDyn, AeroDyn and HydroDyn modules to calculate the aero-hydro-elastic structural response of the entire structure (contains both tower and floater). But a problem remains that whether SubDyn is applicable to floating wind turbines? since I check its manual that says SubDyn is a time-domain structural-dynamics module for multi-member fixed-bottom substructures.

Best Regards,
Yingyi Liu

Dear Yingyi,

The statement in the manual is made because SubDyn currently requires that at least one of the joints of the substructure is clamped to the seabed (inertia frame) i.e. six degree-of-freedom rigid-body motion of the substructure is not permitted. Without modification to the source code, this would limit the applicability of applying SubDyn to the Nezzy floating offshore wind system.

Best regards,

Hi Jason,

Just re-opening this thread to see if you’ve heard of anyone modifying the source code to enable MoorDyn to model guy wires attached to arbitrary locations on the turbine tower since the original post.

Thanks!

Hi Sam,

We’re wrapping up a project where we’ve added the ability to model substructure flexibility and member-level loads to OpenFAST for floating offshore wind turbines. Included in these changes is the ability to enable the SubDyn module to model the structural flexibility of either the floating substructure (with the SubDyn-ElastoDyn coupling at the tower base), or the floating substructure and tower (with the SubDyn-ElastoDyn coupling at the tower top). And with these changes, SubDyn has also been upgraded to include pretensioned cable elements (in addition to beam elements previously available). These pretensioned cable elements can be used to model guyed towers, as long as both ends of the guy wire (pretensioned cable) attach to distinct parts of the SubDyn model. Please see the papers linked in the following forum topic for more information: Extending the dimensions of matrices related to equation of motion - #3 by Jason.Jonkman. We are still verifying and validating this new functionality, upon which these features will be released.

Best regards,

Very cool!

Any chance you have any guyed-tower models you could post? Even if they’re for fixed turbines, I think it would still be quite useful.

Hi Sam,

No, we don’t have such a model developed, but I could see using this new OpenFAST to be released soon to develop one.

Best regards,

Hi Jason,

I’ve been looking for a way to implement a guyed tower model within OpenFAST, but, unfortunately, I can’t really find anything helpful. Is there any current development regarding a guyed tower implementation?

Thank you! =)

Igor.

Dear Igor,

The functionality I mentioned in my post dated Oct 01, 2020 has been released publicly in OpenFAST v2.6 and newer. So, you can now model the wind turbine support structure (fixed or floating) entirely within SubDyn, including pretensioned cable elements that you can use to model the guy-wires.

Best regards,

Dear Jason,

awesome, I’ll take a look on that! Thank you for your answer!

Take care,

Igor.