Atmospheric Ice Accretion On The Blades

Hello,Jason

 I want to simulate the atmosphere ice accretion on the blade,I don't know if the FAST can simulate the special case. Can you give me the advice or recommand me some tools ?

Best Regard !

Ruiliang.Wang

Dear Ruiliang,

Although there are currently no specific “ice accretion” inputs in FAST, blade icing can be modeled. Icing is typically modeled with increased blade mass (and perhaps different masses between the blades) and modified airfoil lift/drag data (modified based on the roughness caused by ice accretion).

Best regards,

Hello,Jason

Thank you very much.
I hope I can simulate the “ice accretion” in the FAST .Do your team simulate the specal case ?As you said,I need modify the mass and airfoil lift/drag data ,But I don’t know how to modify .I have to use some other tool to get the modified data .Can you give me some advice about the software tool for getting the modified data what FAST need .For example, Is the CFX useful ?

Best Regard
Ruiliang.Wang

Dear Ruiliang,

We do not have any tools that you can use to determine how much ice is built up on the blades and in which shape. If you knew the shape, you could use an airfoil code like XFoil to predict the lift and drag coefficients and then use them in FAST. And if you knew how much mass it was, that could be easily added to the FAST blade files. You may also want to modify the stiffness.

You will have to look elsewhere for software to help you do all that.

Marshall

Hello,Jason

Thank your reply quickly.
I simulate the ice accretion on the blade,I need to get the airfoil data.Can you give me the 5MW airfoil Coordinates about DU and NACA64 in your 5MW onshore wind turbine. It will helps!
Best Regard!

Ruiliang.Wang

Dear Ruiliang,

Please see my Jun 04, 2012 post in the following forum topic for the airfoil coordinates of the NREL 5-MW turbine / DOWEC blade: NREL 5MW Rotor Geometry.

Best regards,

hello, Jashon

Thanks for your help!
I want to draw the blade using SOLIDWORKS.I want to identify the diameter of the blade root .IN the blade root, is the airfoil “cilinder1” and “cilinder2” the circle? Is the chord same to the diameter?
Hope your answer!

BEST REGARD!
Ruiliang.Wang

Dear Railiung,

For cylinders, “yes,” the chord is equivalent to the diameter.

You can consider the cylinder data and diameter/chord at the root of the NREL 5-MW turbine as approximations chosen to give reasonable aerodynamic loads within aero-elastic simulations, but not necessarily exact specifications for creating a solid model of the blade. For example, it has been pointed out the root diameter is larger than the hub diameter for the NREL 5-MW turbine, which is obviously unrealistic–see the following forum topic: NREL 5 MW Hub Radius.

Best regards,

Dear Jason,

I am currently facing some small problems and I hope we can figure them out.

As I want to model some additional mass in a rotor blade, I changed the mass density in the blade input file.

I increased the mass density of the last three spanwise locations (Radius=62.2 m to 63 m; BlFract= 0.98699 to 1) by 10 kg/m each. However, the FAST summary (.fsm) file still reveals a total blade mass of 17740.082 kg. Even if I increase the mass density by 50 kg/m each or 100 kg/m each nothing changes.

When I change the mass density of the last four spanwise locations (Radius=61.7 m to 63 m; BlFract= 0.97886 to 1) FAST realizes these changes resulting in an increased total blade mass.

However, when I model some additional mass only in the area of Radius=59.7 m to 60.7 m by increasing the mass density, BMassDen, at these locations, the total blade mass is again 17740.082 kg.

I have an idea and hope you can confirm it:

As long as one or more of the 17 blade nodes (see Table 3-1 of “Definition of a 5-MW Reference Wind Turbine for Offshore System Development”) are affected by the change in mass density, these changes will contribute to the total blade mass.

Does this inherently mean that I could model the NREL 5 MW rotor blade by only defining the blade structural properties at 17 spanwise locations at “RNodes” ?

Or back to my example from above:
Would it make any difference if the blade input file contains some random numbers in the last three rows (Radius=62.2 m to 63 m; BlFract= 0.98699 to 1) as these rows seem not to contribute to the total mass, and hence, to the blade inertia.

Best regards,

Sebastian

Dear Sebastian,

Yes, your understanding is correct. The structural model of FAST v7 and the ElastoDyn module of FAST v8 simply use linear interpolation to transfer the input distributed blade structural data to the blade analysis nodes.

Best regards,

Dear Jason,

thanks for your really fast reply!

Do you know if there is a possibility to change the BMassDen during simulation?

For example: At the beginning of the simulation an additional mass, m_add, contributes to the blade root. During simulation this mass linearly decreases at the blade root but simultaneously increases at the blade tip.

Beginning of simulation:

Root: m_add and Tip: 0 kg

End of simulation:

Root: 0 kg and Tip: m_add

Best regards,

Sebastian

Dear Sebastian,

Without changing the source code, you can not change the mass in FAST during the simulation.

Best regards,

Thanks a lot Jason for your help!

Best regards,

Sebastian

I am trying the same thing as Sebastian with FAST V8.16 and the BeamDyn module.

If I manually alter the last mass matrix in the BeamDyn Blade 3 file to add 1000kg at the tip station the BD3.sum file mass results only shows a percentage of the mass increase (~50%).

I am using the NREL 5MW Test26 model.

To add the mass I determine the length of the last section - dr, determine the new mass/length = (1000)/dr + old mass/m . Then all values in the mass matrix are multiplied by the factor = (new mass/length) / (old mass/length).

The goal is to spread a given mass over a length of blade but for a check I started with the tip section.

What is happening here that my mass/m change doesn’t give the expected mass calculation in the BD.sum file?

(putting 2000 kg over the last two stations also only increases the BD3.sum mass by 60%)

Dear Mark,

Assuming you haven’t changed Test26 in any other way, the BeamDyn model in Test26 is set up to use Trapezoidal quadrature for the spatial integration e.g. to obtain the mass matrix. Trapezoidal quadrature will have a bit different effect on the total integrated mass than you are considering with your scaling of (new mass/length) / (old mass/length) at the tip station.

But I’m not sure I understand your statement about 50% mass increase. The total blade mass of the NREL 5-MW is around 17000 kg, so adding 1000 kg should be no were near a 50% mass increase.

Best regards,

Morning Jason,

to clarify:
I expected the mass result in BD3.sum to increase 1000kg (so 17000 + 1000kg)
The difference between the mass in BD1 or BD2.sum and BD3.sum was about (17000 + ~500kg).

Thanks for the information. The quadrature remains Trapezoidal (I think the number of elements prohibits Gaussian…).

I will try (as a lazy wild guess) doubling the increase factor for the first and last elements where I want to increase mass and see where that goes.