Tower (and blades) mode shapes BMODES vs. FAST

Dear Jason,

Thanks for your reply.

So no need to consider the mass and stiffness matrix from subdyn to BModes right (that means zero in hydro_K and hydro_M), if we are considering both tower and monopile in BModes as above.

Thanks,
Satish J

Dear Satish,

Yes, for a rigid foundation, that is correct. For a rigid foundation, hub_conn should also be set to 1.

In the CS example, hub_conn is set to 3 and a stiffness matrix is provided to represent the coupled springs foundation.

Best regards,

Dear Jason,

Thanks for your reply.

  1. Is it correct input file(attached file - BModes_ip.rtf) if I want to get the modesahpes from the mudline and considering rigid foundation. I have considered hub_conn to 1 and hydro_M, hydro_K and mooring_K to zero matrix.

  2. If I want to consider the soil model in the BModes, Is it just specifying spring stiffness to stiffness (hydro_K) which is at the mudline and hub_conn to 3 in BModes, right?

Thanks,
Satish J
BModes_ip.rtf (8.46 KB)

Dear Satish,

Yes to both.

Best regards,

Dear Jason,

Thanks for your reply.

Could I know where might be wrong as I am getting the different direction mode shapes (1st s-s and f-a) when I considered from the mudline. I have attached the plots of both modes for both cases (modeshapes from mudline and interface level).

Thanks,
Satish J


Dear Satish,

The different signs is related to how the mode shapes outputs are normalized in BModes (mass normalization). You can always normalize a different way, like normalizing so that the tip displacement is unity (which is what is used by the tower mode shape inputs to ElastoDyn).

Best regards,

Dear Jason,

Thanks for your reply.

Could I know how I can normalize in terms of sign by getting the output in similar s-s and f-a directions.

Thanks,
Satish J

Dear Satish,

Simply divide all values of the mode shape along the span by the value at the tip. In this way, the value at the tip is always unity and the remaining mode shape is scaled accordingly.

Best regards,

Dear Jason,

Could you please advice on which of the below frequency sets (Both are taken from same output) are correct which are generated from BModes and I have attached BModes output file.

Frequency Set 1:
1st s-s natural frequency (Hz) - 0.2473 (Mode No. 1)
1st f-a natural frequency (Hz) - 0.2479 (Mode No. 2)
2nd s-s natural frequency (Hz) - 1.3551 (Mode No. 3)
2nd f-a natural frequency (Hz) - 1.4005 (Mode No. 4)
3rd s-s natural frequency (Hz) - 2.8182 (Mode No. 6)
3rd f-a natural frequency (Hz) - 1.5337 (Mode No. 5)

Frequency Set 2:
1st s-s natural frequency (Hz) - 0.2473 (Mode No. 1)
1st f-a natural frequency (Hz) - 0.2479 (Mode No. 2)
2nd s-s natural frequency (Hz) - 1.4005 (Mode No. 4)
2nd f-a natural frequency (Hz) - 1.5337 (Mode No. 5)
3rd s-s natural frequency (Hz) - 2.8182 (Mode No. 6)
3rd f-a natural frequency (Hz) - 3.2161 (Mode No. 7)

Thanks,
Satish J
CS_Monopile.out.txt (74.1 KB)

Dear Satish,

I would say Frequency Set 2 is correct.

Best regards,

Hi, I want to model a small wind turbine in FAST software
In the ELASTODYN module, I have to give the information about the tower to the software in the TOWER section

but the point is that we have used telecommunication towers as towers in wind turbines
Now I want to see where I need to get information about it such as TOWER ADJUSTMUNT FACTORS or TOWER FORE-AFT MODE SHAPES or TOWER SIDE-TO-SIDE MODE SHAPES
Or do we need to provide information about this type of tower at all or not

thanks

Dear Ali,

The tower mode shapes must be input into ElastoDyn. They can be derived in a pre-processing step based on the distributed mass and stiffness data, as well as the boundary conditions at the tower base (e.g., fixed) and tower top (lumped RNA mass/inertia). The software referred to as BModes_JJ on this forum can be used if you haven’t derived these mode shapes some other way.

The tower adjustment factors in ElastoDyn need only be set different from unity if you are confident in the tower properties (distributed mass, distributed stiffness, mode shapes), but you see differences between ElastoDyn’s predictions of mass, natural frequencies, or deflections based on known data (e.g., measurements or a higher-fidelity mode). In that case, the tower adjustment factors can be set different from unity to tune the ElastoDyn response to match the known data.

Best regards,

Dear jason,
thank you for your answer
But most of my emphasis was on the fact that the tower of this small wind turbine is a telecommunication tower(I try to attach a picture of it)

And my question is, is the simulation of this type of tower no different from other patches?
And Is it possible to access data related to this type of tower with software such as BModes in order to provide elastodyn tower inputs?

Dear Ali,

The question I would ask is if you want to model this lattice tower as a single beam, or do want to model each individual element of the lattice tower so as to calculate the loading within each individual element? For the former, the use of ElastoDyn and BModes_JJ to derive the mode shapes would be satisfactory. For the latter, you would want to model the tower in SubDyn rather than ElastoDyn.

Best regards,

Dear Jason,

I’m using BModes to calculate the natural frequencies of the tower for the distributed springs (DS) model in OC3 project, and the first four natural frequencies is 0.2448Hz, 0.2464Hz, 1.3574Hz and 1.5108Hz shown in Fig 1. However, the eigenfrequencies for the flexible foundation model is 0.2456Hz, 0.2476Hz, 1.5327Hz and 1.5459Hz in file of Derivation and Description of the Soil-Pile-Interaction Models by Patrik Passon (Fig 2), and there is a big difference in 2nd side-to-side natural frequency. I’m confused that is there any wrong in my settings? The main input file is attached for your further check, really appreciate!

Best regards,
Kevin


Fig. 1

Fig2.JPG
Fig. 2

Dear Kevin,

From my quick look, I don’t see anything obviously wrong with your BModes input file. And the first pair of modal frequencies are quite close to what you expect.

The second bending modes of the support structure are often heavily influenced by structural flexibilities in the RNA, e.g., generator rotation, drivetrain torsion, and blade bending. While these effects can be captured in the FAST / OpenFAST model, they can’t be captured in BModes, so, the natural frequencies of the second bending modes of the support structure predicted by BModes are typically a bit off. However, from our experience, BModes is still accurate enough to generate the mode shapes needed by the ElastoDyn module of FAST / OpenFAST, for both the first- and second- support structure bending modes. And once you transfer the BModes-generated mode shapes to ElastoDyn (after polynomial fitting), the OpenFAST model can then capture the correct full-system natural frequencies.

Best regards,

hello
I have some questions, thank you for answering me
My questions are about bmodes software

  1. In the tower sections properties file
    I have a problem with the “flp_iner” parameter. Based on its user guidance For a tower, it is mass moment of inertias per unit length about the side-side yT axis.My question is, should this moment of inertias be about the red axis below or the blue one?

2)Another problem I have is that I do not know what the term tower-top mass means.

3)My third question is the same as the first question, because in the “ixx_tip” parameter, according to the user guide For a tower, it is the side-to-side moment of inertia of the tip mass about the tip-section xT reference axis. Here, too, I do not know about which axis it wanted the moment of inertia?blue or red (base of attach file)

Dear Ali,

Here are my answers to your questions:

  1. A similar question regarding the definition of the cross-sectional inertias (flp_iner, edge_iner) was asked and answered in the following forum topic: BModes flp_iner - #2 by Jason.Jonkman.

  2. The-top mass and inertias refer to the entire rotor-nacelle assembly (RNA) represented as a lumped mass and inertias. These mass and inertias can have a strong impact on the tower mode shapes.

  3. The tower-top (RNA) inertias are defined about the x,y,z axes whose origin is the RNA center of mass.

Best regards,

Dear Jason,

I have a question that might be too obvious but I wanted to confirm it with you.
I have access to the coefficients of the mode shape in Elastodyn.tower.dat , but I don’t have access to how these coefficients were calculated (Bmodes or manually)
If I want to get the tower mode shape deflection and its slope at the hub height, should I simply use equation y(x) = a_6x^6 + a_5x^5 + a_4x^4 + a_3x^3 + a_2*x^2 and substitute the normalized coefficients and substitute also x=1 (as I need the information the hub height)?

Thank you in advance,
Kind regards,
Álvaro

Dear @Alvaro.Ponce,

Yes, that’s correct. Of course, the actual deflection is scaled by the tower-bending DOF, that is, y(1) is 1 in you equation above, but the actual deflection is TTDspFA = y(1)*q_TwFA1 if only the first mode is active. If multiple modes are active, you’d sum the contribution of each tower bending-mode DOF.

Best regards,

1 Like