Dear Gia,
I quickly skimmed your input file and it looks OK to me, although I haven’t run it myself. What is the exact error message you are getting?
Best regards,
Dear Jason,
Thank you for your prompt reply.
The error message is the following:
"Invalid logical input for file ‘’.\5MW_Baseline/NRELOffshrBsline5MW_Monopile_RF_SubDyn.dat’’ occured while trying to read SSSum.‘’
Best regards,
Gia
Dear Gia,
Ah, I now see that you added a line to the MEMBER X-SECTION PROPERTY DATA table without increasing NPropSets from 4 to 5. The remaining rows in the file are not being read properly.
If this happens again, you can debug an input file by reviewing the echo file created with Echo is set to TRUE.
Best regards,
Dear Jason,
Thank you very much for your help. It finally worked!
Best regards,
Gia
Dear Sir
For the same wind and wave conditions, one would expect lower response from the OWT with increasing apparent fixity depth. However, for a HH wind speed of 30mps (idling rotor), I find that the FATBM is steadily increasing with increasing AF depth. My Hs and Tp values are 10m and 14s, respectively.
I have modelled the idling rotor in FASTv8, using the recommendations in the post Extreme events. I have used the default monopile OWT in FASTv8. Please help me diagnose the error.
Sincerely
Abhinav
FAST8_idle.zip (16.1 KB)
Dear Abhinav,
Can you confirm/clarify a few things:
- By FATBM, you are referring to the TwrBsMyt output from ElastoDyn (the tower base is located 10-m above MSL for each model of different apparent fixity depth)?
- You’ve modeled the apparent fixity by introducing another joint and member in SubDyn?
- The cross-sectional properties of the apparent fixity beam are identical with each depth?
- Using BModes, you update the tower mode shapes in FAST with each change to the apparent fixity depth in SubDyn?
When you increase the apparent fixity depth, you are increasing the length of the beam, which increases its flexibility if the cross-sectional properties of the beam are unchanged between depths. If the pile deflection is substantial, I can foresee an increase in FATBM with increasing depth/flexibility.
Best regards,
Dear Jason
Thanks for the answer. And sorry for my late reply. I was so upset (thinking that you probably considered my question to silly to merit an answer), that i didn’t check this page for over a week! My sincere apologies.
My answers to your questions are given below:
- by FATBM, i in fact, referred to the “ReactMYss” - Base reactions: fore-aft moments at the mudline, since i assumed the “tower” as a long cantilever from the top to the mudline
- i have modelled the AF by introducing several small elements (of the same cross-section and young’s modulus). Please see the attached file. Is it necessary, that i model the AF depth as a single element?
- the cross-sections of the AF beam are same for the different AF depths, but their youngs modulus values vary in minor proportions
- yes, i have updated the tower mode shapes using BModes
To clarify point # 2, i subdivided the AF beam into a number of elements to increase the mesh density. But a glance at the Subdyn.out file tells me that FAST in itself, subdivides the AF beam by introducing nodes. So i hope the AF beam can be modelled as a single element rather than dividing it into sub-elements in the SubDyn input file. The KBB and MBB matrices don’t vary much.
What troubles me is the fact that the increase in the moment is of monstrous proportions, when compared to the minor increase in AF depth.
Also, i carried out a similar analysis for YawBrMyp. I found that the same trend was repeating.
Sincerely
Abhinav
NRELOffshrBsline5MW_OC3Monopile_SubDyn_AF1.txt (8.05 KB)
Dear Abhinav,
OK, thanks for the clarifying information. Your SubDyn input file looks OK to me.
To answer your question about an AF model, the AF beam need not be modeled as a single member. (I’m pretty sure you mean “member” when you refer to “element”.) SubDyn divides each member into NDiv finite elements (you’ve set NDiv = 3 in the SubDyn input file).
I agree that the large variation you are showing in -ReactMYss with AF depth looks odd. However, I’m not sure I know what is causing this behavior. You see the same variation in YawBrMyp? What are the platform fore-aft displacements and tower-top fore-aft deflections from ElastoDyn for these cases–do they look reasonable? I suggest you scan your input files for possible errors.
Best regards,
Dear Jason.
Thanks for the reply.
Yes, it is observed that YawBrMyp also shows the same variation.
I have made the following changes for modelling the idling rotor. These are based on your recommendations in the post viewtopic.php?f=4&t=321&hilit=idling and the chapter “Simulating special events/idling turbine” of the old FAST (v7) manual.
Aerodyn input file:
- Stall model - from BEDDOES to STEADY
- IndModel - from SWIRL to NONE
Elastodyn data file:
- BlPitch values to 90
- RotSpeed to 0
Servodyn data file:
- PCMode to 0
- TimGenOn to 9999.9
I have not made any other changes to my input files for modelling the idling rotor. Is it possible that i overlooked something?
The F-A displacement at tower-top and platform level also escalates with increasing AF depth (please note the figure).
Sincerely
Abhinav
Dear Abhinav,
I took the files from the FAST8_idle.zip archive from your Mar 09, 2015 post above and ran one case with the newest release of FAST v8 and I’m not matching your results. I had to create some of my own files because you didn’t provide a complete set (e.g., AeroDyn). However, I obtain a mean tower-top fore-aft displacement of less than 0.1 m, a mean platform fore-aft (surge) displacement of about 0 m, and a mean fore-aft bending moment at the mudline of about 10 MNm. I’m not sure how to interpret your plots – are they not the mean values?
Best regards,
Dear Jason
Thank you for taking time off your work to check my problem. The values in the graph are the absolute maxima from the time series (averaged across 5 TurbSim seeds). I am interested in the serviceability limit state and hence the maximum values. As for the mean values, I believe our results are fairly consistent (< 0.1m for tower-top FA displacement, 0m for platform FA displacement). Sorry for the lack of clarity in the previous posts.
Please allow me to sum up once again.
I have observed that the response (FATBM, FA displacements at tower-top and platform level) of the NREL 5MW OWT increases at an alarming rate (as in the graphs) with even minor increases in the AF depth (I considered values from 19m to 25m, based on a soil-stiffness study). By responses I refer to the maximum value in the time series. The mean values remain constant irrespective of the AF depth. I believe I have made a mistake somewhere, but I am unable to find it out. I am attaching all the files, except for the wind, files and I hope you will be able to identify the error in my simulations.
Sincerely
Abhinav
NREL_5M_OWT.zip (24.4 KB)
Dear Abhinav,
Without your wind files, I couldn’t reproduce exactly your results, but I do see similar maximum values of tower-top and platform displacement when I use my own wind. I also ran the case with a rigid foundation and saw that the tower-top and platform displacements are much lower in that case.
I don’t really see any problems with your input files. To check for consistency between the models with rigid and AF foundation, I ran a case where I increased the Young’s modulus of the AF beam by 3 orders of magnitude. As expected, I obtained quite similar results between this case and the one with a rigid foundation. This tells me that the larger displacements for the case with the true AF foundation are purely result of the properties of the AF beam. I’m not fully sure I understand why the results are so strongly dependent on the length of the AF beam, but I’m not convinced they are incorrect either. Are you confident that the AF beam is providing the effective foundation flexibility that you desire? Looking at the SubDyn summary file, the AF beam is seen to drop the overall stiffness of the substructure quite a lot when compared to the case with the rigid foundation.
Best regards,
Dear Jason
Thanks for checking out my files. I believe that the stiffness values i use are representative of the soil properties. I’ve checked it using the geotechnical code LPILE and the more general USFOS. The soil data I have used, represents soft, offshore conditions in the Arabian Sea. Also, for operating conditions, the AF model is giving good predictions, for similar stiffness formulations. The problem crops up only when I model an idling turbine. I was hoping that the error would be due to some mistake i made with the idling mode.
I have not been able to handle the transition form FAST7 to FAST8, with ease, though i feel that v8 is a great improvement. Are the ReactFXss and ReactMYss values (i.e. mudline shear and moments), in v8 are analogous to the TwrBsFxt, TwrBsMyt values in v7?
Sincerely
Abhinav
Dear Abhinav,
I’m not sure I would say that TwrBsFxt from FAST v7 is analogous to -ReactFXss from FAST v8, etc. in general, but for a monopile model this is effectively true. TwrBsFxt from FAST v7 (and from the ElastoDyn module of FAST v8) is the tower fore-aft shear force at the base of the flexible tower. -ReactFXss from SubDyn is the total shear force at the mudline from all reaction nodes. In FAST v7, a monopile is modeled as a flexible tower, but in FAST v8, a monopile is modeled with the tower in ElastoDyn and the substructure (pile) in SubDyn.
Best regards,