Grid4DField_GetVel:Outside the grid bounds

[RoshanThomas1699]

Hello everyone!

I am trying to run a simulation FAST.Farm but end up receiving the following error:

i was trying to simulate the exact case sir @jjonkman validated "V8_TI10_3WT" below i will mention my .fstf file and .inp

@jjonkman sir can you please help me out. I understand that there must be something wrong with how I set up the simulation grid but I have a hard time understanding what exactly causes this error, i.e., which of my input files needs to be updated. I also took a look at the Modeling Guidance section in the documentation but have a hard time to understand where I made a mistake regarding my input files. Thank you

[andrew-platt]

At the very start of the simulation, is there any information from InflowWind about the extents of the wind file?

[RoshanThomas1699]

@andrew-platt sir i couldn't find anything related to that you mentioned. i also tried simulating the same case that sir @jjonkman has validated against sowfa that is the case V8_TI10_3WT. I also tried simulating same case on newer version today. below is the .fstf file I used

And below is my .inp file i sued for turbisim

iam also mentioning IW.dat file which is same i used for .fst file for each turbine (5MW BASELINE)

this is the error iam getting now

I also took a look at the Modeling Guidance section in the documentation but have a hard time to understand where I made a mistake regarding my input files. please guide me through it. Thank you

[jjonkman]

Dear @RoshanThomas1699,

You mention that you are running case V8_T10_3WT. Can you clarify where you obtained your FAST.Farm files for this case?

I have not reviewed all of your files, which are difficult to review when they are pictures. Can you share the actual files, e.g. through Google Drive? Also, your error changed between your first and second post; what did you change between these runs?

It is very difficult to set up FAST.Farm models and generate wind data files by hand that follow of the FAST.Farm modeling guidance. We recommend using the Python-based scripts from the OpenFAST Toolbox, which automates the process for you: https://github.com/OpenFAST/openfast_toolbox/tree/main/openfast_toolbox/fastfarm. FYI: When using TurbSim to generate wind inflow for FAST.Farm, we generally recommend using Mod_AmbWind = 3 rather than Mod_AmbWind = 2. The Python scripts support Mod_AmbWind = 3 well.

Best regards,

[RoshanThomas1699]

sir @jjonkman as iam learning this through process only, my programming background is also weak. My main doubt is whether we can generate wind field by editing existing .inp files and generating the binary files using turbisim for our simulation? Or should we start from using python coding?

[jjonkman]

Dear @RoshanThomas1699,

If you have a functional FAST.Farm input file, you can make small edits by hand, but changes that involve changing the domain size and discretization, such as changing the turbine locations, would be difficult to do by hand. Changing the wind direction would also be difficult to do by hand, because this is often done by changing turbine locations rather than the wind direction. Changing the mean wind speed could also be difficult because the modeling guidance recommends changing the discretization with mean wind speed.

I would generally recommend using Python scripts. This is not so much about programming (unless you want to change the functionality), rather, setting inputs in the Python scripts and running the script to automate the generation of wind inflow and FAST.Farm input files.

Best regards,

[RoshanThomas1699]

@jjonkman so for validating the case V8_T10_3WT i need to generate input files .fstf as well as inflow files using python, right sir? Also where can i find the parameters that used for this case in fast.farm for validating? note: This validation serves as the foundation for my further studies using the software. Thank you

[jjonkman]

Dear @RoshanThomas1699,

FAST.Farm input files for a shortened version (fewer time steps) of case V8_T10_3WT from our original FAST.Farm validation paper presented at TORQUE 2018 (https://docs.nrel.gov/docs/fy18osti/71376.pdf) are available on my Google drive: https://drive.google.com/file/d/1vWlDzIcR6pnuJgb4B3410cxufwn3tFnK/view?usp=drive_link. This model is formatted for an old version of FAST.Farm, so, to run it, you'll have to upgrade the model to the newest version.

Best regards,

[RoshanThomas1699]

Sir, if iam using mod amb 3 for the above case how could i generate user defined time series for the wind inflow?

[rthedin]

For a TurbSim-driven FAST.Farm case (Mod_AmbWind = 3), there is a complete example using the OpenFAST toolbox available here: https://github.com/OpenFAST/openfast_toolbox/blob/main/openfast_toolbox/fastfarm/examples/Ex2a_FASTFarm_TurbSim_driven.py. Specifically, the inflow is generated in two steps: first the low-resolution box, then time-series are extracted from low-res to drive the high-resolution boxes. Those are indicated in the example linked.

As mentioned, however, you will need to update the turbine model to the latest version of OpenFAST and FAST.Farm.

[RoshanThomas1699]

how can i update? or where i could find the latest ones? Also i couldnt find the templates to run those scripts, can you help me out?

[jjonkman]

Dear @RoshanThomas1699,

I'm not sure what you mean by "templates", but documentation of input files with each release of OpenFAST and FAST.Farm are provided on OpenFAST readthedocs: https://openfast.readthedocs.io/en/main/source/user/api_change.html. Examples of FAST.Farm input files compatible with each release of FAST.Farm are available in the OpenFAST r-test: https://github.com/OpenFAST/r-test.

Best regards,

[RoshanThomas1699]

sir @jjonkman my doubt is where i could find those files which relates templates in example file https://github.com/OpenFAST/openfast_toolbox/blob/main/openfast_toolbox/fastfarm/examples/Ex2a_FASTFarm_TurbSim_driven.py Elastodyn.T, ServoDyn.T ,Subdyn.dat and ElastoDyn_Blade.dat?

[rthedin]

The "templates" are simply a set of OpenFAST input files for the turbine of interest. The "template" word comes from the variable name where the user sets the filename of the individual input files. It could be a good idea for you to run standalone OpenFAST with your turbine model before moving to FAST.Farm so you can understand the input files and make sure they are compatible with the software version you are using.

[RoshanThomas1699]

@rthedin i tried generating for openfast and this was i found when i ran it. can you please help me figure this out?

[jjonkman]

Dear @RoshanThomas1699,

You haven't explained your simulation set-up, but I see warnings about large skew ("a large value of chi") and high Mach number and your simulation error is triggered by a Mach number above 1. Are you seeing large blade deflections? Perhaps you model is numerically unstable? You'd have to provide more information to understand what is wrong.

Best regards,

[RoshanThomas1699]

sir @jjonkman iam not sure if iam right? i generated .fst file using the python script and generated .fst file using that. the python script and the openfast files are uploaded in below link. the new simulation result came like this

is this normal? One of my other doubts is, in the python script uploaded in the drive link (got from github) do we really want to refer main .fst file to generate new? Then what will be the difference? https://drive.google.com/drive/folders/1dbfdbpTS8T8O_OknzNznUmLHkErmJf5g?usp=sharing Thank you.

[jjonkman]

Dear @RoshanThomas1699,

I'm not really understanding what you shared with me. You have two fst files in the directory you shared, but neither of them has CompAero set to enable AeroDyn, which was triggering the warnings and errors you shared previously. I doubt this is the OpenFAST model that generated the errors. It would also also help to summarize what you are trying to do.

Best regards,

[RoshanThomas1699]

sir @jjonkman can i update that by hand in .fst file i generated using python? also my main doubt is in the python script uploaded in the drive link (got from github) do we really want to refer main .fst file to generate new?-like i did? thank you

[rthedin]

It's unclear what python code you are using to generate your input files. It seems like you are trying to run standalone OpenFAST-- is that correct? How are you using the code you are using? The toolbox linked above is for FAST.Farm simulations, not OpenFAST.

It's unclear to me what you are trying to achieve, what set of input files you are starting from and what modifications you want to do.

[RoshanThomas1699]

@rthedin Iam not able to understand certain parameters in script? in the fastfarmcasecreation file they are asking for subdyn.dat and elastodynblade.dat files, what does that mean, where i could find that? Also can you give me reference to follow up steps so that i could generate my own wind farm domain? which all python scripts from the link should i use (name). i meant which all python files should i deal with? Sorry that I couldnt figure it out. thank you

[RoshanThomas1699]

@jjonkman @rthedin Can you help me to figure out which pythons scripts only i should deal with? (file name in the link) i was trying to figure out, going through all. Is it only fast farm case creaction or should i g for another script within the link? thank you.

[rthedin]

I've given the link to an example in a prior message: https://github.com/OpenFAST/openfast_toolbox/blob/main/openfast_toolbox/fastfarm/examples/Ex2a_FASTFarm_TurbSim_driven.py. This is a TurbSim-driven FAST.Farm case.

However, we are going to need more information about your setup and your goals to be able to help you. You haven't explained what you are after with these simulations. It seems like you are trying to run FAST.Farm-- I would strongly suggest you start with standalone OpenFAST before moving to FAST.Farm. You need to get an understanding of the different modules and how the inputs are set before moving to FAST.Farm.

I recommend you start by reading the OpenFAST manual.

[RoshanThomas1699]

@rthedin as i mentioned earlier iam trying to simulate V8_T10_3WT sir @jjonkman has validated against sowfa to learn the software (even though he gave me every input files). At first i thought i could edit the input files and reach at the level later only from continous interaction I realised i need to utilise python scripts to generate my own farm and all. I tried the openfast inputs or perquisites got from github. But now i want to know how could i generate case with python scripts available,as I couldn't find any steps to follow. I'll go through manual deep Thank you

[RoshanThomas1699]

Hi @rthedin and @jjonkman , thank you for your guidance.

I am a postgraduate student working on offshore structures, and my project focuses on the wake effect in wind farms.

From what I understand so far, to create a case for study in FAST.Farm, I cannot just edit the values directly in the input files I found on GitHub. Instead, I need to use the Python scripts available in the python_toolbox to generate the necessary input files for whatever cases I need to study.

By using the script called FastFarmCaseCreation, I can create my own .fstf and wind inflow files with the domain parameters I want.

Is this understanding correct? Please correct me if I’m wrong.

Thank you!

[rthedin]

That is mostly correct. A couple of notes:

1. FASTFarmCaseCreation.py is a file that contains the FFCaseCreation class. These are not scripts.
2. The toolbox does not generate the "wind inflow files". It creates the inputs necessary for you to run TurbSim yourself and generate them.

[RoshanThomas1699]

Thank you @rthedin , Few simple things i want to make sure. so, i can make use of the python scripts in example files and generate inputs, which calls the class FFcasecreation? also the scripts end like this

So i need to modify these? is there anything else that i should keep in mind or take care of?

[rthedin]

The example script located inside the examples directory is an example. Thus, it illustrates how to use the toolbox to help with the case setup. It is not meant to be executed all at once, and it will fail if you attempt to do so. In fact, we have the following comment in the middle of the example script: https://github.com/OpenFAST/openfast_toolbox/blob/ece7ca286c0109f4be8d60abbe24da77288a5bda/openfast_toolbox/fastfarm/examples/Ex2a_FASTFarm_TurbSim_driven.py#L172-L174

Since this is an example to new users, it is expected that new users execute it slowly, line by line, in case of issues. I'm not sure what your question is, and what you mean by "these". I think you could benefit from taking a step back and read the theory on why we need both a low-res box and one high-res per turbine and understand their requirement.

[RoshanThomas1699]

Thank you so much for your constant guidance, @jjonkman and @rthedin . I ran the provided example case (Python), and here are the results: When i ran the .fstf file it shown like

Is it okay to be like this, will it affect my results? or should I make the change for the domain? Also, the visualisation i got was

from+z axis views and on the second run for the same case just turning of yz and xz axis in simulation

the visualisation was like this. why the wind is disturbed in the field and turbine moving out of grid on second run

The files i used are mentioning below https://drive.google.com/drive/folders/1IU3WseXPZPuJUtji9L1SeksiOT9mUksf?usp=sharing

2. For wake added turbulence for predefined file. can i use the same available for any cases?

Thank you

[jjonkman]

Dear @RoshanThomas1699,

Regarding (1), there is no issue in FAST.Farm if the wake planes leave the back-side (maximum X) of the domain; you can ignore these warnings.

I'm not sure what you are doing, but I would not expect the high-resolution domains to be impacted by whether or not you visualize the XZ and YZ slices of the low-resolution domain.

Regarding (2), yes, the wake-added turbulence implementation was developed so that you could use the predefined isotropic turbulence boxes provided by NREL for all cases. That said, you could certainly generate your own isotropic turbulence and use that to see its effect.

Best regards,

[RoshanThomas1699]

@jjonkman Sir, I simulated the example case from the following link: Ex2a_FASTFarm_TurbSim_driven.py using the same .fstf file and inflow files (TurbSim-generated boxes).

However, when I ran the case twice using the same inputs (only change in the visualisation part), the visualization results appeared different each time.

Could you please let me know if:

This variability is expected,

Or if I might be missing something in the setup or execution?

I want to ensure I’m interpreting the results correctly and not overlooking any important settings or randomization aspects. Thank you.