Question about tower and nacelle modelling

[AlexGuntner]

Dear everyone,

I am currently looking to include the effects of the nacelle and tower in my simulations, but I haven’t been able to find detailed information on how these components are implemented in the code. I would appreciate some clarification on the following points:

In the input file, there is a parameter nacelle_area, described as “the frontal area of the nacelle used to compute the force.” Since the IEA-15MW wind turbine has a circular frontal area, I assumed this could be calculated using the formula $A = \pi r^2$ based on the nacelle radius. However, when inspecting the domain using this assumption, the nacelle occupies a much larger area than expected. Could you clarify how the actual area is determined?

• How are the tower actuator points modelled? Specifically, what method is used, and what drag coefficient is applied?

Kind regards

[hgopalan]

I am not familiar with the tower actuator points modeling. For Nacelle, there are three options:

1. We have a PR for nacelle modeling: F/superellipsebody #1549 accounting for its deformation when coupled with OpenFAST and NALU-Wind. I am not sure if this is something you are looking for.

2. The other option is to use the IB method to mark the nacelle region as a single shape.

3. If you want a static Nacelle, you can use the code in PR# 1 and move the code from initialize_fields to post_init_actions().

The method for nacelle can also be used for tower if you want to move the tower points from actuator model.

[AlexGuntner]

Dear @hgopalan,

thanks for the answer. I'm not interested in the coupling with Nalu-Wind, I am performing simulations of floating wind turbines using just the moving actuator lines. The IB method sounds reasonable to me - I'm not exactly sure what you mean by "static nacelle".

Is the IB method then the default method used in the code? What drag coefficient would you recommend for the nacelle? Could you also kindly clarify the nacelle_area parameter?

Best regards

[hgopalan]

When you use OpenFAST, you can account for the deflection of the nacelle which was the purpose of the coupling with NALU-Wind. As that's not of interest to you, you just need a method to do a simple bluff body drag model for the Nacelle.

You can take a look at the example here: https://github.com/Exawind/amr-wind/tree/main/test/test_files/ib_cylinder_Re_300. This uses the IB method to represent the drag force effects. You can approximate the nacelle by a cylinder/sphere.

The other option is the use of immersed forcing (https://exawind.github.io/amr-wind/walkthrough/terrain.html, we just use the immersed forcing for the Nacelle). We do not have a setup for nacelle using the method yet. Do let me know if you want to use this method.

I am not familiar with OpenFAST and this is what we have in documentation:

amr-wind/docs/sphinx/user/inputs_Actuator.rst

Line 300 in 72801c0

Actuator.TurbineFastLine.nacelle_area = 0.0

. I assume you just compute the projected frontal area of the nacelle and provide that as an input parameter.

[AlexGuntner]

I'm guessing that if I were to use the IB method to approximate the nacelle, then I could only use it for a fixed wind turbine, but I would like to follow the motion of the floating platform.

Regarding the nacelle frontal area, I wanted to replicate the diameter of the IEA15MW from an experimental study - 18 m. I then calculated the nacelle frontal area using the formula $A = \pi \cdot r^2$ = 254.34. If I use this assumption, and set $C_d = 1$, I get the following:

As you can see, If I generate the actuator contours I get a very large nacelle area, which is unrealistic. Still, the wake does not look like is affected by this apparently large nacelle.

[hgopalan]

If you would like to follow the platform motion the method proposed here will work: #1549. In this case NALU-Wind assumes control of OpenFAST and generates the output file which AMR-Wind reads at every instant and controls the motion of the nacelle. I am not sure how to get those values from OpenFAST directly into AMR-Wind.

For the other model, I am not sure.

@lawrenceccheung @gyalla @gantech any suggestions?