search

Exawind / amr-wind

AMReX-based structured wind solver
https://exawind.github.io/amr-wind
Other
111 stars 82 forks source link

"Invalid particle" error when adding refinements to simulation #857

Closed jfrederik-nrel closed 10 months ago

jfrederik-nrel commented 1 year ago

Hi there, I am running AMR-Wind with OpenFAST, and getting an error that seems related to the refinements I do around the OpenFAST turbine. The error I'm getting is:

Step: 14403 dt: 0.05 Time: 7200.1 to 7200.15 CFL: 0 (conv: 0 diff: 0 src: 0 ) terminate called after throwing an instance of 'std::runtime_error' what(): ParticleContainer::locateParticle(): invalid particle. MPT ERROR: Rank 12(g:12) received signal SIGABRT/SIGIOT(6). Process ID: 196504, Host: r2i2n29, Program: /home/jfrederi/projects/amr-wind/build/amr_wind MPT Version: HPE MPT 2.23 11/19/20 04:21:56

This error seems related (identical?) to the issue reported here: https://github.com/Exawind/amr-wind/issues/415 However, this is an old issue and it doesn't offer a solution, as it's supposed to be resolved in newer versions of AMR-Wind.

I'm using two refinement boxes around my turbine. When I remove the second level, the simulation runs. I've also tried using cylinders instead of boxes, but this results in the same error. My simulation input file is attached below. My AMR-Wind header is shown below:

                              AMR-Wind (https://github.com/exawind/amr-wind)

  AMR-Wind version :: e0ffbb6-DIRTY
  AMR-Wind Git SHA :: e0ffbb62eb60d70d74b02ed2a2587551face2aa2-DIRTY
  AMReX version    :: 23.01-26-ga232be4e275b

  Exec. time       :: Mon Jun 12 17:21:09 2023
  Build time       :: Feb 15 2023 16:21:48
  C++ compiler     :: GNU 8.4.0

  MPI              :: ON    (Num. ranks = 64)
  GPU              :: OFF
  OpenMP           :: OFF

  Enabled third-party libraries: 
    NetCDF    4.7.3
    OpenFAST

If anybody can help me resolve this issue, it would be much appreciated!

simulation.zip

jfrederik-nrel commented 1 year ago

@lawrenceccheung, I wanted to give you an update on this error. As we discusses on Wednesday, I've tried decreasing the step size from 0.05 to 0.02 and even 0.01. However, this does not resolve the error, which still occurs after the second time step. I'll now try updating AMR-Wind to the latest version, but if you have any other suggestions, please let me know!

jfrederik-nrel commented 1 year ago

@lawrenceccheung I have now also tried updating AMR-Wind (and OpenFAST) to the latest version. The problem unfortunately still persists.

github-actions[bot] commented 1 year ago

This issue is stale because it has been open 30 days with no activity.

jfrederik-nrel commented 10 months ago

Update on this issue, it appears to still be present for some of the cases that I am running.

We have managed to pinpoint the problem to refinements being to close to the lower boundary of the field. However, simply raising the refinement zone is unfortunately not an option as this would mean part of the rotor swept area would fall outside of the refinements. For the IEA 15MW turbine that I am running, the blade comes to 30 meters from the ground, while I need to raise my refinements to 50 meters to be able to prevent this error.

If anyone knows a solution/workaround for this, I'd be very thankful!