Help running FAST v8 for the first time

Dear Dr. Jonkman,

Thank you for your reply. The FAST 8 readme file says if WrVTK > 0, then it’ll generate some files which I can use them in ParaView (am I understanding right?). I’m a bit confused, how I can set WrVTK > 0 since I am just typing sth like “FAST_x64 Test01.fst” at cmd? In other words, should I change sth in the “Test01.fst” file (if yes how)?!! if no, can you please guide me? Thank you.

Dear all,

I am trying to use FAST_v8.16 for the first time. I have used FAST_v8.15 on a previous computer (64 bits Windows) to model a floating wind turbine. Since I have a new machine (again 64 bits Windows), I have now installed FAST_v8.16. I have several questions about this subject.

When I launch the CertTest, it runs correctly. In the CertTest.out file, some little differences appear between the previous and new tests:

  • Test 2: Test02.ED.sum: Rotor Inertia (41755.992 VS 41755.984 kg.m²)
  • Test 3: Test03.ED.sum: Rotor Inertia (41755.992 VS 41755.984 kg.m²)
  • Test 5: Test05.ED.sum: Rotor Inertia (41755.992 VS 41755.984 kg.m²)
  • Test 6: Test06.AD.sum: description of the twist, chord, etc. + tip-loss constant and hub-loss constant: hub-loss constant different in the elements 4 and 7 ([12.55715 VS 12.55714] and [24.58929 VS 24.58928])
    The rest is OK. Is it normal or is there a problem?

Then, I have tried to go on using these previous files (used with FAST_v8.15) with this new version. I have added the section “Linearization” in the .fst file. Is there another place where I should modify something in my previous files?

When I use a previous file with the modification of Linearization, an error message appears: “The dll “DISCON_rect1_gwin64.dll” could not be loaded. Check that the file exists in the specified location and that it is compiled for 64-bit applications”
I have checked that this file exists at the correct location. I had compiled the .dll file on the former computer using “make -f makefile_DISCON_DLL”. In makefile_DISCON_DLL, we had adapted the code to be adapted to our 64 bits machine, as presented below:

32-bit or 64-bit?

#BITS = 32
BITS = 64
Do you have any idea of the reason of this error message?

I have tried to recompile a .f90 file that had already been compiled on the former machine (using make -f makefile) but I get the following error message:
“‘make’ isn’t recognized as an internal or external command, an executable program or a command file” (I hope you’ll understand it because I have to translate the message in english). Do you know where this might come from? I might have misunderstood something in the Compiling instructions (Compiling FAST v8.15.00a-bjj). I have tried to install a Fortran compiler but I’m not sure that it runs correctly.

That’s a lot to ask, I hope you’ll be able to help me with these different subjects.
Best regards,
Léna

Dear Mahmoud,

Yes, the WrVTK input parameter is specified in the Test01.fst file. As with any FAST input file, open up the file in any text editor, make the desired changes, save, and run FAST with the updated input file(s).

Best regards,

Dear Lena,

Here are my answers to your questions:

  1. The differences you are reporting sound like small numerical round-off due e.g. to differences contributed to the optimizer when the source code has changed. I would not expect these differences to impact the results to engineering accuracy.

  2. The FAST v8 ReadMe file explains how the input file(s) have changed between FAST v8.15 and FAST v8.16 (see the section titled “Changes in FAST v8.16.00a-bjj”): wind.nrel.gov/nwtc/docs/README_FAST8.pdf. You can also use the provided MATLAB conversion scripts to update your input files automatically.

3/4) Unfortunately, I have no experience using make files; hopefully some else on this forum can comment.

Best regards,

Dear Jason,

Thank you for your reply. I understand that the issue isn’t related to the differences when I run CertTest.
I will go on looking for the solution.

Best regards,
Léna

Dear Dr. Jonkman,

Thank you very much. With your helpful comment, I could make an animation for the Test01 file by ParaView software. Dr. Jonkman, at this juncture I am able to run FAST at cmd, open Test01 and see the animation of the wind turbine. Now I want to simulate my own 3-blade offshore wind turbine with my specific wave and wind conditions. How can I use FAST to do that? Is it like I should write an input file from beginning by myself (if yes is there any PDF to teach how to do this) or no I can use one of your example input files, like “Test??.fst” and then I can modify the wave, wind, and the structure of the wind turbine there?

Dr. Jonkman, we, in the Civil Eng. Dept. of NTU, are very hopeful to learn FAST and apply it in my Ph.D. research for the first time. Thank you for your valuable comments.

Dear Mahmoud,

I would take the FAST CertTest model (set of FAST input files) that best matches your system and modify those input files accordingly. (The FAST CertTest models are described in the Certification Tests section of the FAST v8 ReadMe file.)

The FAST documentation is a bit scrambled at the moment. But each input file of FAST is described in the various User’s Guide that have been provided e.g.:

I hope that helps.

Best regards,

Dear All,

I was trying to run “Test20.fst” with “FAST_v8.16.00a-bjj”. At cmd I typed “fast_x64 Test20.fst”, the same way I’ve successfully run “Test01.fst”, but I got an error there! My PC-windows is 64-bit, though next time I typed “FAST_win32 Test20.fst” then it terminated normally. BUT, when I’m opening “Test20.outb” or “Test20.SFunc.outb” in a text file, there are all strange letters like “¤%¤þ×%¿ÔÍZÝ‹)Ñ©ó € € € ^éÏ € €ú«:Ò×÷U*ê¾ €”. Any idea how I can fix this problem?

Thank you all.

Dear Mahmoud,

Test20.fst will trigger an error regarding in inability to load the 32-bit DLL controller when running with the 64-bit executable. You can eliminate that error by referencing the 64-bit DLL in the ServoDyn input file. You can run either the 32-bit or 64-bit executable on a 64-bit computer.

The .outb files are output files stored in a binary (non human-readable) format. You can process the data in post-processors adopted to read such data, such as MCrunch, MExtremes, MLife, or the MATLAB toolbox. Or you can change the output to be in text format by changing the setting in the FAST primary (.fst) input file (but text-based output files are about 4-times as large as binary output files for the same data).

Best regards,

I found the FAST v8.12 ReadMe file to be helpful for this. Try the “Normal Simulation: Starting FAST” section.

Did you figure this one out? :stuck_out_tongue:

Hi All,

I am trying to use OpenFast for my graduate study. I have downloaded the source code and successfully installed it under Cygwin. Currently, I have several questions about running and testing the program:

  1. I checked a post at [url]Beginner's guide to fast?] and tried the suggested conversion between FAST7 and FAST8. The Matlab code worked out with some warning and I could not run the converted file. I copied the converted files into ~/OpenFAST/build/glue-codes/fast then I use this syntax ./openfast.exe -h NRELOffshrBsline5MW_Floating_OC3Hywind.fst and the program prompted FAST_InitializeAll:The required input file was not specified on the command line. Anybody know how to resolve this? am I missing something in the syntax?

  2. I saw in the previous posts and README_FAST8.pdf that there are examples provided in FAST CertTest folder, which I could not find in the github distribution. How can I access these example? I assume those examples were compatible with OpenFast and I can directly run it without conversion.

  3. I am trying to set up a simulation of an offshore wind using time series of wind/wave from other models. Is there any similar example from the cert test that I can use as a starting point?

Any suggestion would be very helpful. Thank you.

Regards,
Boma K

Dear Boma,

Here are my answers to your questions:

  1. Remove the “-h” from the command line call i.e. type, “./openfast.exe NRELOffshrBsline5MW_Floating_OC3Hywind.fst”.

  2. Yes, you don’t need to convert yourself any of the NREL-provided models because we update these models ourselves when we change FAST and run the tests. What was called the “CertTest” (certification tests) in FAST v7 and FAST v8 is now called the “r-test” (regression tests) in OpenFAST. The OpenFAST r-test is available here: github.com/OpenFAST/r-test.

  3. Test_004 and Test_006 from the CertTest of the standalone HydroDyn module are cases with user-defined wave kinematics and wave-elevation time series, respectively. The OpenFAST team is still in the process of migrating all of these module-level CertTests into the OpenFAST r-test. What kind of user wind time history are you referring to e.g. a uniform wind file or a full-field turbulence (TurbSim) file?

Best regards,

Dear Dr. Jonkman,

Thank you very much for the instructions now I can run the examples. I have more simple questions if you don’t mind:

  1. How do I visualize all the structure in the example without running FAST? I am aware that FAST is compatible with ParaView, and able to produce animation through WrVTK options. I am also new to Paraview.

  2. For the simulations that I am trying to set, I work with atmospheric-ocean models (WRF-ROMS) and I want to apply wind field from WRF time series output as an input in OpenFast. Is this possible?

Regards,
Boma K

Dear Boma,

Here are my answers to your questions:

  1. You can use the visualization functionality of FAST to output either surface or stick-figure geometry for model reference and initial configurations and time-series animation through the generation of VTK output files that can be visualized through e.g. ParaView. See the section titled, “Visualization Toolkit Files (Visualization Capability)” in the FAST v8 ReadMe file for more information: wind.nrel.gov/nwtc/docs/README_FAST8.pdf.

  2. I’m not familiar with WRF-ROMS, but I’m assuming this model does not generate wind inflow at the high spatial resolution important for running aero-elastic calculations in OpenFAST (e.g. on the order of meters). So to generate equivalent wind time series for FAST would require that e.g. you use the user time series option in TurbSim to read in the WRF-ROMS time-series output at a point (e.g. at the hub) or a few points and use TurbSim to construct the finer resolution wind data based on the user time series provided. See the TurbSim v2.0 documentation for more information: nwtc.nrel.gov/alphas.

Best regards,

Hi everyone

I’m using FAST for first time. I’ve complied it in Linux (Ubuntu 14.04.5 LTS) and I could run the test files. However, when I try to run it for other models, I get the error below. I guessed it might be the line in which Echo is set, but it’s exactly the same between Test_*.fst and the fst file I’m trying to use. Any hint would be much appreciated.
NB: I’m using version 8 and not OpenFast because I’m using it with WISDEM and have to use that version.

  Running FAST (v8.16.00a-bjj, 27-Jul-2016), compiled as a 64-bit application using single
  precision
  linked with NWTC Subroutine Library (v2.09.00, 23-Jul-2016)


 FAST_InitializeAll:FAST_Init:FAST_ReadPrimaryFile:Invalid logical input for file
 "NRELOffshrBsline5MW_Monopile_RF.fst" occurred while trying to read Echo.

 FAST encountered an error during module initialization.
  Simulation error level: FATAL ERROR

  Aborting FAST.

Dear Hassan,

The only way I can think that would generate that error message is if the file NRELOffshrBsline5MW_Monopile_RF.fst is not formatted correctly. It sounds like you checked this already, so, I’m not sure what else to suggest. You could upload the file so that I could take a look.

Best regards,

Dear Jason

Thank you for taking the time to look into my comment and respond. Yes that was my first guess, but since it’s my first time using it, I’m not sure about the required format. I’ve attached the file here (I got an error when I tried to attach it because of the extension, so I just changed it to txt)

Thanks again
Hassan

Dear Hassan,

I don’t see that you attached the file. Can you try attaching it again?

Best regards,

Dear Jason

Sorry for that, maybe I mistakenly clicked on Submit before file was actually uploaded. However, thanks I could figure out where the problem was.

Thanks again
Hassan

I’m glad you solved the problem!