Is dual-rotor generator modeling supported in OpenFAST? Can tower and rotor structures be omitted?

[YuYuJou]

Hi
I'm wondering whether the latest version of OpenFAST supports modeling a dual-rotor generator system (see the image below). My generator features two counter-rotating rotors mounted on a single floating structure, without a traditional tower.

❓ My main questions:

1. Has support for dual-rotor systems been added in recent OpenFAST updates?
2. If not, is it feasible to omit the rotor dynamics in ElastoDyn by setting BladeNodes = 1?
3. Additionally, since my generator design lacks a defined tower, can I also set TowerNodes = 1 to ignore tower dynamics?
   To build the model, I modified the IEA 15MW reference wind turbine files and replaced the original WAMIT .1, .3, and .9 hydrodata files with my own WAMIT-generated data.

However, when running the simulation, I encountered the following error related to the .9 file:

📌 Do I need to recompute my .9`file with a finer frequency and wave-direction resolution?

Thanks so much for your support.

[andrew-platt]

Dear @YuYuJou,

We are currently working on implementing multi-rotor support for MHK turbines in OpenFAST version 5.0, which we expect to formally release later this year (August, or potentially sooner). There are still several pieces in development for this that we will be merging into the OpenFAST development branch in the next couple of months if you want to try using it before the formal release.

@luwang00, @hkross, or @jjonkman may be able to answer some of your other questions.

Regards,

[jjonkman]

Dear @YuYuJou,

Multi-rotor (including dual rotor) capability is being added to OpenFAST as recently summarized on our forum: https://forums.nrel.gov/t/multi-rotor-openfast-simulations/8211.

However, your image shows a system with potentially unique needs, such as a support structure with lifting surfaces. Can you clarify: Are you proposing an airborne or underwater system? The system looks like a bi-plane; are both the top and bottom wings lifting surfaces?

Which parts of the system are you attempting to model in WAMIT?

Best regards,

[YuYuJou]

Thank you for your reply.

My system is an underwater one. I’ve simplified the model and already run the hydrodynamic calculations in WAMIT—the input files are as shown in the images. Currently, I’m trying to use OpenFAST to simulate free decay behavior, treating the model purely as a floating platform for now—excluding effects from blades and other components.

May I ask if this approach is feasible? Also, how should I adjust the model to exclude the influence of blades and the tower in OpenFAST?

Thank you again for your guidance.

Best regards,

[jjonkman]

Dear @YuYuJou,

I'm not sure what you are referring to as the "tower", but I would expect that you'd want model the entire support structure (other than the drivetrains and rotors) in SubDyn with HydroDyn. This would allow you to calculate the structural flexibility of the substructure and hydrodynamic loads on all members, including hydro-elastic effects. However, the hydrodynamic loads calculated by HydroDyn would not include lift forces that I expect you'd want to calculate on the airfoil-shaped part of the support structure. Including that would require a source code change. The rest of this system you should be able to model with the multi-rotor capability coming soon in OpenFAST.

Best regards,

[YuYuJou]

Dear OpenFAST team,

Thank you very much for your continued support!

I understand that full multi-rotor modeling is not yet available in the current version of OpenFAST. For now, I would like to simulate the free decay behavior of my floating dual-rotor system.

I have been using the RM1 model as a reference, and these two modules are the primary parts I have modified to suit my simplified platform. Based on this setup, I’d like to ask:

🔹 How should I proceed with configuring my model to simulate free decay, considering that rotor dynamics are omitted?

To clarify my earlier question: I mentioned the tower settings in ElastoDyn because I have been using the RM1 model as a reference, and that model includes tower-related structural parameters by default. However, my system does not include a traditional tower structure.

🔹 In this case, how should I handle the tower-related inputs in ElastoDyn? Should I simplify or remove them, or would it be more appropriate to disable ElastoDyn entirely?

Thanks again for your valuable guidance!

Best regards,

[jjonkman]

Dear @YuYuJou,

In OpenFAST, the 6-DOF rigid-body motion of a floating system is solved by ElastoDyn, and when SubDyn is enabled, the substructure elasticity relative to the rigid-body motion is solved in SubDyn (if you don't want substructure flexibility, you can disable SubDyn). Given that your system does not have a typical "tower", you should define your ElastoDyn model such that the tower is neglected, i.e., by disabling the tower DOFs and setting the tower geometry such that its length is negligible (i.e., TowerHt only slightly above TowerBsHt).

Best regards,

[YuYuJou]

Thank you for your reply.

Since you suggested minimizing the tower height to eliminate tower-related effects, I would like to ask:
Will the tower height setting affect the rotor position?
Additionally, in my case, the rotor is not located at the center of the structure, but rather on one side.
Where should I adjust this offset in the model? Should it be in ElastoDyn, or somewhere else?

Thanks again for your support!

[jjonkman]

Dear @YuYuJou,

This will become more clear once we release the version of OpenFAST with multi-rotor capability. In this version, the origin of each instance of ElastoDyn (each turbine) can be specified independently from the others and differ from (0,0,0) that is assumed now.

Best regards,

[YuYuJou]

Thank you for your response.

If I want to divide my model into the four parts shown in the attached image:
Which part would be most appropriate to include in SubDyn?

Additionally, when the new version is released, if I want to run a complete simulation, should I define two sets of airfoils in AeroDyn—one for the upper wing and one for the blades?

Thanks again for your guidance.

Best regards,

[jjonkman]

Dear @YuYuJou,

In the upcoming version of OpenFAST with multi-rotor capability, I would suggest modeling members 1, 2, and 4 in SubDyn and 3 in (two instances) of ElastoDyn (one for each nacelle). And you'd use HydroDyn to model the hydrodynamic loads on everything but the rotors, which you'd model in AeroDyn). The limitation here would be the absence of a lift force being computed by HydroDyn on member 1, which would require a source-code change to model.

Best regards,