Dear @Emanuel.Rergis,
Your SubDyn input file makes sense to me, but values in your SSIfile don’t match what I would expect for the coupled springs representation of the OC3-monopile. Where did you obtain these stiffness values?
Regarding the location in the tower, I would expect that you could identify the natural frequencies from the motion or loading of any point in the tower, unless that point is a node of a given mode (a point which does not move in that mode).
Best regards,
Dear @Emanuel.Rergis,
1- there is no threshold value for WvLowCOff. You can take 1.5 × WvLowCOff_1st_mode_frequency.
Let me be more clear. In fact, i suggest to change WvLowCOff to a value a bit greater than tower first bending mode in order to obtain only the tower 2nd bending mode (i.e. You compute the PSD of tower top fore-aft displacement and as i expect you will obtain one peak that is associated to tower 2nd bending mode).
Also, you can set WvLowCOff to zero and set WvHighCOff to 500 rad/s. In this case, the PSD of tower top fore-aft displacement will show two peaks: the first peak is associated to the tower 1st bending mode and the second peak is associated to the tower 2nd bending mode.
I forget to mention taht when yo apply the method of white noise spectrum, you should disable AerDyn and InflowWind as well as servodyn. This guarante that no effect of aerodynamic damping will be present.
2- I think it is sufficient the log-log scale. You dont need to use decibel.
3- I am not quite sure that i understand your question (i am not very good in English). But, from @Jason.Jonkman, i would say to consider tower top fore-aft displacement (In openfast, it is called TTDspFA) since at this location the tower is free and the mode shapes is equal 1.
Hope i was clear 
Best Regards,
Riad
Dear Dr. @Jason.Jonkman ,
Thank you again for your prompt response. The values in the SSI file were taken from a paper on soil mechanics. Your opinion makes me reconsider and doubt the accuracy of these values. What approximate values should I expect, then? Is there a manual or a reliable source where I could find trustworthy information on this?
I apologize for the question, but I am a beginner in this field. I have never had to work with soil mechanics before.
Thank you.
Best regards,
Emanuel M. Rergis
Dear @Emanuel.Rergis,
The stiffness matrix that has been prescribed for the OC3-monopile is included in the sample UserPtfmLd_CS.f90 file shared through the the following forum topic: Turbine-soil interaction. Influence on the mode shapes. - #13 by Jason.Jonkman. Note that stiffness values are only provided for surge, sway, roll, pitch and their couplings. As prescribed, you should disable the heave and yaw DOFs at the seabed.
Best regards,
Dear Dr. @Jason.Jonkman ,
Thank you very much for your response. I quickly reviewed the file you suggested and found something that confused me. I am sending you a screenshot where the section of code defining the soil stiffness matrix appears.
When observing this matrix, I see that two matrices are defined (you can also see them in another screenshot). One is on the left side of the code, and the other is on the right. After a quick analysis, I found that the differences lie only in the signs of some elements in both matrices.
My question, therefore, is: Which matrix should I consider as the correct one? The one on the right or the one on the left?
Thank you again.
Best regards,
Emanuel M. Rergis
Thank you very much again for your responses, @Riad.Elhamoud.
I am going to implement the method you suggested. My next question would be: With this white noise method, will I also be able to see the third mode?
I understand that the value of WvLowCOff should be > WvLowCOff_2nd_mode_frequency**. Is it correct to assume this for this case?
Thanks once again.
Best regards,
Emanuel M. Rergis
Dear @Emanuel.Rergis,
The left matrix (that is not commented out) is the correct one to use.
Best regards,
Dear @Emanuel.Rergis,
1- For me, yo cannot obtain third mode since ElastoDyn_tower does not contain the third mode shape of the tower.
2- I dont agree. WvLowCOff should be < 2nd_mode_frequency.
Best Regards,
Riad
Dear Dr. @Jason.Jonkman,
Thank you very much for your response. I wanted to share that I have started reading more about this topic, and I even found some papers mentioning the use of BModes as a tool for obtaining the mode shapes I am looking for. As I continued reading, I came across this link:
BModes/docs/examples/CS_Monopile.bmi at master · old-NWTC/BModes · GitHub
This link contains an example of an input file for modeling an OC3-type Monopile. However, it is not clear to me which part of the code I need to modify in order to implement the stiffness cross-coupling model. Could you clarify that for me, please?
My second question is whether this is the correct input file that I should modify, or if I should instead use this one:
BModes/docs/examples/OC3Hywind.bmi at master · old-NWTC/BModes · GitHub
Finally, do you think using BModes is also a good way to find what I am looking for (the fist 3 modes shapes of the tower with soil interaction)?
Thank you once again, and I look forward to your response.
Best regards,
Emanuel M. Rergis
Dear @Emanuel.Rergis,
Yes, you can use BModes to calculate the mode shapes and natural frequencies of a wind turbine tower atop a monopile with coupled springs foundation, as long as you are OK assuming a rigid rotor-nacelle assembly (RNA).
But I would not use files from the old_NWTC GitHub site. Rather the version referred to as BModes_JJ on this forum is what you want to use, which you can find here: BModes - Google Drive. BModes_JJ needs the extra “Properties of tower support subsystem” section of the input file, which is where you would place the 6x6 coupled springs model of the foundation.
Best regards,
**Dear Dr. @Jason.Jonkman **
Thank you again for your comments. I downloaded the Bmodes_JJ.exe file. However, I only found an executable file that, when initialized, does not generate any changes. Additionally, I did not find any .inp or .bim files to define the characteristics of the soil stiffness matrix (please find attached the snapshots)
Could this be because I did not download something else? Is there an error? What would you advise me to do?
Thank you again.
Best regards,
Emanuel M. Rergis
Dear @Emanuel.Rergis,
There are two BModes_JJ examples in the Google drive directory where you found BModes_JJ.exe: OC3Hywind.bmi, which uses OC3Hywind_tower_secs.dat and CS_Monopile.bmi, which uses CS_monopile_tower_secs.dat.
Best regards
Dear @Jason.Jonkman,
I have some inquires regarding ElastoDyn.
1- When i specify 1 meter for TTDspFA as initial condition, the “1 meter” is considered for the tower 1st bending mode or for both modes ?
2- When specifying damping ratio in ElastoDyn Tower, this value will be used later for computing damping coefficient. However, according to what i know, damping coefficient requires the natural frequency such that the damping coefficient in fore-aft direction, c_fa, is expressed:
c_fa=2 * tower_mass * damping_ratio_in_fore-aft*tower_angular_frequency_in_fore-aft.
Where “tower_angular_frequency_in_fore-aft” does come from before obtaining the natural frequency of the system?
Thank you in advance,
Best Regards,
Riad
Dear @Riad.Elhamoud,
Here are my responses:
Best regards,
Hi @Riad.Elhamoud
I have some preliminary results from several free decay experiments I’ve conducted using ElastoDyn. Some of these involve disabling all the DOFs and then gradually reactivating them one by one to observe how the natural frequency of the first mode changes as each DOF is added. I’ve analysed this through quick PSD-type assessments. If you think this information might be useful or relevant, I’d be happy to share the plots with you.
Kind regards.
Emanuel
Hi @Emanuel.Rergis,
Thank you so much. In fact, i dont think that it would be useful for me since i am working on the 5 MW barge-type FOWT and you are working on the 5 MW monopile-supported OWT.
I have asked @Jason.Jonkman because i could not find anywhere an answer for that. And also i am preparing for my Ph.D. defense so i have to expect any qestion from the jury 
Thank you a lot.
Best Regards,
Riad
Thank you very much, Dr. @Jason.Jonkman, for your advice. I will review BModes a little later. For now, I’m interested in obtaining the FRF (Frequency Response Function) of the tower’s fore-aft and side-to-side directions. I need to do this to evaluate and validate the effectiveness of the TMDs I’m implementing.
To achieve this, I’ll need to compare the results from my MATLAB simulations using ode45 with those from OpenFAST. My question is that I’m not entirely sure how to obtain these FRFs in OpenFAST.
I thought that if I disabled all degrees of freedom in ElastoDyn, and turned off all other modules except for ServoDyn, I could use time-prescribed force files to apply sinusoidal excitations at various frequencies. By sweeping across a frequency range and computing the maximum displacement for each frequency, I could construct an FRF similar to the ones I’ve obtained in MATLAB.
However, the challenge with this method is that I need to know what value to use for the equivalent mass of the system (tower + hub + blades + nacelle, etc.) in order to reduce it to a single degree of freedom. Once I have that, I can tune the values for stiffness (K) and damping (C) accordingly.
What value of mass would you recommend using so that my results are comparable? Or, do you know of a better approach to achieve this goal?
Thank you very much for everything.
Dear @Emanuel.Rergis,
Given that OpenFAST is a nonlinear time-domain model, it cannot directly output a frequency response function (FRF).
You could, of course, linearize the OpenFAST model through a linearization analysis and export a linearized state-space representation of the model, from which you could transform to an FRF.
Best regards,
Dear @Emanuel.Rergis,
Hope you are doing well.
I dont know whether what i will say is useful for you. Instead of computing FRF, you could simply compute RAO of the tower top fore-aft displacement for two cases: with and without wind.
I expect you will see that the presence of wind will dampen out tower bending modes but i am not quite sure since i have not worked on monopile-supported OWT.
The procedure for computing RAO is detailed in page 3 section “RESPONSE AMPLITUDE OPERATOR COMPUTATION PROCEDURE: FAST” of this paper which is published by NREL (you can rely on it):
@Jason.Jonkman What do you think ?
Hope that helps.
Best Regards,
Riad
Dear @Riad.Elhamoud,
I agree with your response; thanks for adding another suggestion.
One thing I’m not clear on is what input signal @Emanuel.Rergis wants in the FRF. RAOs are typically used for wave inputs and HydroDyn / SeaState are directly set up to use white-noise wave input for RAO generation. But you’d have to come up with your own white noise if you wanted to compute RAOs for other inputs.
Best regards,
