Dear @Kazuya.Nishiyama,
The time series of motion in the PtfmXMotionFn / PtfmYMotionFn / PtfmZMotionFn files is scaled by PtfmMotionFactor. In the El Centro examples, the motion time series are specifed as accelerations scaled by PtfmMotionFactor = 9.80665, so, a value of 0.2 in the time series file is actually 0.2 * 9.80665 = 1.96 m/s^2.
Best regards,
Dear Jason.
Thank you for your response.
Am I correct in understanding that if PtfmMotionFactor= 9.80665, I can input the correct acceleration by dividing the input data by that value or by setting PtfmMotionFactor= 1?
Also, the actual maximum acceleration of the El Centro earthquake is about 3.4 m/s.
I am thinking that even if PtfmMotionFactor= 9.80665, the input maximum of about 0.2 m/s for the El Centro earthquake in FAST is incorrect, why is this?
Dear @Kazuya.Nishiyama
Am I correct in understanding that if PtfmMotionFactor= 9.80665, I can input the correct acceleration by dividing the input data by that value or by setting PtfmMotionFactor= 1?
Yes
Also, the actual maximum acceleration of the El Centro earthquake is about 3.4 m/s.
I am thinking that even if PtfmMotionFactor= 9.80665, the input maximum of about 0.2 m/s for the El Centro earthquake in FAST is incorrect, why is this?
I’m not sure. I was not the one who made this case and I’m not familiar enough with this earthquake to comment.
Best regards,
Dear Jason.
Thank you for your response.
I have an additional question.
・I see on the forum that the highest frequency for 5MW is about 6Hz, am I correct in assuming that I should set ActFreq=60?
(In seismic.dat, it was set to ActFreq=12).
・What is the highest frequency of WindPact 1.5MW?
・I get an error when I analyze under these conditions, what is the cause?
I think the error is caused by a large acceleration, since the error does not occur when the acceleration is reduced.
However, I would like to perform the analysis with this acceleration input.
Analysis conditions↓
FAST-v7,
Turbin model:WP1.5MW
Turbin State:parking(GenDOF=False,PCMode=0,RS=0,BP=90)
Wind:WindSpeed=0
seismic:PtfmMotionFactor=1,Input acceleration as shown in the picture
error↓
Small angle assumption violated in SUBROUTINE SmllRotTrans() due to a large tower
deflection. The solution may be inaccurate. Simulation continuing, but future warnings will be
suppressed.
Dear @Kazuya.Nishiyama,
Regarding the NREL 5-MW turbine, I agree that the highest natural frequency in the FAST model is about 6 Hz, which would suggest ActFreq = 60 Hz if you follow the seismic modeling guidance, but ActFreq = 12 (a factor of 2 instead of 10 higher) was deemed sufficient for this model and increasing to ActFreq = 60 would have required a smaller time step and longer simulation time.
Regarding the WindPACT 1.5-MW turbine, it has been many years since I performed a linearization + eigenanalysis of this turbine in FAST, but the last time I did, I calculated the highest natural frequency of about 20 Hz, which is why the time step (DT) is set to 0.005 s.
Regarding the “large tower deflection” warning, how do the simulation results look otherwise? This warning has been discussed in many other forum posts and indicates that the the accuracy of the results will be diminished when the deflection is that high. I’m surprised the tower would deflect enough to trigger this warning, but perhaps your seismic loading is just that severe.
Best regards,
Dear Jason
Thank you for your response.
I have modified the input file and the error has been resolved.
However, when I input the earthquake in the first picture, the output looks like the second picture.
As the acceleration of the earthquake settles down, the acceleration of the platform increases.
(The phenomenon from about 250s and from about. 450s)
What is the cause of this?
The analysis time is 600s and outputs 100s to 600s.
Seismic is input from 150s to about 450s
※Red line is with seismic input
Blue line is without seismic input
Dear @Kazuya.Nishiyama,
I would expect the platform acceleration output of OpenFAST to match the inputted acceleration time series if the model is set up and running properly. I’m not sure what is going on and would need more information on the model set-up to comment. Are you following the modeling guidance for PtfmMass, ActFreq, ActDamp, and DT?
Best regards,
Dear Jason
Thank you for your response.
The earthquake acceleration and ptfmTAXi do not match and I do not know why ptfmTAxi is larger at 250s and 450s when the earthquake acceleration is smaller.
(Seismic is input from 150s after analysis)
Is it possible that as the seismic acceleration decreases, PtfmTaxi increases?
I have set the following
Seismic Acceleration=Actual earthquake acceleration
PtfmMass= actual weight of Wind Turbine foundation
ActFreq=5
ActDamp=65
DT0.005
Dear @Kazuya.Nishiyama,
It looks like three of your settings are not following the seismic guidance:
PtfmMass should be set 10x the mass of the entire system (for the actuator to work properly), so, about 2000000 kg for the WindPACT 1.5-MW turbine.ActFreq should be set 2-10x the highest frequency of the model, so at least 40 Hz for the WindPACT 1.5-MW turbine.DT should be dropped accordingly with ActFreq, so, with ActFreq = 40 Hz, DT = 0.0025 s.Best regards,
Dear Jason.
Thank you for sharing the details.
I made the change, but ptfmTA and seismic acceleration do not match.
(1st: seismic acceleration, 2nd: ptfmTA)
I have attached the seismic.dat and Ptfm.dat files.
(3rd:seismic.dat, 4th:Ptfm.dat)
Are there any errors in these files?
Dear @Kazuya.Nishiyama,
It looks like your model is unstable for periods of time, but it is unclear to me why. What do you have DT set to?
I’m not sure what other conditions you are setting in the .fst input file, but you I would start the debugging by simplifying the model, e.g., by disabling areodynamics and all DOF except the platform DOFs. Does the platform motion then match what you expect and remain stable? If so, what features, once enabled, make your model go unstable?
Best regards,
Dear Jason
Thank you for your response.
I have set DT=0.001.
I have set the seismic input to start at 400s.
I have disabled aerodynamics and all DOFs except Ptfm DOF, but the output of PtfmTAxi is wrong. (from around 500s).
I have attached the fst and Ptfm files.
Please check it.
(AD.ipt, Blade.dat and Tower.dat files are from WP1.5MW model and have not been changed).
Dear Jason.
We performed additional checks.
Under the same conditions, we found that changing the seismic input time significantly changed the waveform.
I am not sure of the cause of this.
Could you please tell me the possible causes?
Red: seismic input time 500s~.
Blue: seismic input time 550s～
Dear @Kazuya.Nishiyama,
Did you share the correct platform file? The platform file you shared has no platform DOFs enabled and no platform load selected, so, should not result in any platform motion.
When you say you change the start of the seismic input, are you referring to setting EqDelay to 400, 500, 550 s, etc?
For me to take a look, it would be easier if you could pick the simplest model you have been testing (all DOFs disabled except the platform DOF) and share the complete FAST model with me (not just pictures), including inputs and outputs.
Best regards,
