Fluids

Fluids

Trouble importing mesh from GMSH

    • ari9748
      Subscriber

      Hello everyone, hope you all are doing fine. I need some help regarding a problem I am facing. I wished to import a 2D mesh from GMSH(.msh file format) and run the simulation in Fluent. But when I try to edit in the setup section it runs into an error and which is follows:

      Quote:

      (DP 0) Update of the Setup component in Fluent failed: Exception of type 'Ansys.Fluent.Cortex.CortexNotAvailableException' was thrown.

      I also noticed that by default the simulation is considering the dimension as 3D and I don't have any option to change it. Can anyone throw some ideas why am I running into this kind of problem?


      P.S- I am using Ansys Fluent 19.2 release.

    • Keyur Kanade
      Ansys Employee
      If it is 2D, make sure that your geometry and mesh is in XY plane.
      Also use latest version 2022R1. Fluent 19.2 is very old build.
      You can open Fluent in standalone mode in 2D.
      Please go through help manual for more details
      Regards Keyur
      How to access Ansys Online Help Document
      Guidelines on the Student Community

    • ari9748
      Subscriber
      Thank you for your generous response. Right now I am trying it with 19.2 version and would try with the latest version later. As you adviced I ran the fluent in pareallel standalone mode --->2D----> File----->Read mesh but encountering some problem still which is as follows:
      Building...
      mesh
      ==============================================================================

      Node 0: Process 36428: Received signal SIGSEGV.

      ==============================================================================

      Node 1: Process 43864: Received signal SIGSEGV.

      ==============================================================================

      Node 2: Process 5736: Received signal SIGSEGV.

      ==============================================================================

      Node 3: Process 50588: Received signal SIGSEGV.

      ==============================================================================
      MPI Application rank 0 exited before MPI_Finalize() with status 2
      Done.
      The fl process could not be started.
      Can you please provide some suggestion how I can troubleshoot this problem?
    • Keyur Kanade
      Ansys Employee
      Please check if geometry and mesh are in XY plane.
    • ari9748
      Subscriber
      I have double checked. It is in x-y plane and a 2 D model for sure. Also in 2022 R1 version it gives the same problem.
    • Rob
      Ansys Employee
      Can you confirm it's a 2d mesh and not a 3d mesh that's one cell thick? Will it read into Fluent Meshing?
    • ari9748
      Subscriber
      I have double checked. It is in x-y plane and a 2 D model for sure. Also in 2022 R1 version it gives the same problem.

    • Rob
      Ansys Employee
      And in Fluent Meshing?
    • ari9748
      Subscriber
      Yes I have tried. Now I tried with a simple geometry i.e a rectangle and when fed into fluent meshing it is showing:
      Warning: no nodes readanalyzing boundary connectivity...done.
      >
      Meshing>
      But here in GMSH the statistics is showing this:

      I have also uploaded the .msh file.
      Till now not reached to any satisfactory state.

    • Rob
      Ansys Employee
      Staff aren't permitted to open or download any files or attachments - it's covered in https://forum.ansys.com/discussion/23093/why-ansys-employees-dont-download-attachments

      I had a quick look at GMSH and the text I saw suggested it was 3D only, happy to be proved wrong on that but please double check.
    • ari9748
      Subscriber
      I have even tried other file format like (.nsh) Nastran Bulk data but same issue. Is there any other alternative to that?
    • Rob
      Ansys Employee
      Can you read the geometry (ie what was read into GMSH) into SpaceClaim?
    • ari9748
      Subscriber
      Hi Rob, thanks for your response again. Actually, I am trying to keep my mesh intact i.e what I obtained from GMSH. So, after lot of research I have found a way, firstly I exported the mesh from GMSH to Openfoam and then openFoam to Fluent as .msh file. I was finally successful in reading the same mesh file(i.e obtained from openfoam) in Fluent 2D as well as 3D. Now, what I saw the domain is having a single cell thickness and I am wondering will it be possible to run a 2D simulation in that regard? If so can you please suggest me a b.c for the front and back face(in openfoam it was kept as empty patch) whereas I am trying pressure far-field b.c for inlet and outlet. I have attached the pic how the mesh looks like.

    • Rob
      Ansys Employee
      OK, that sounds like the GMSH export isn't respecting the Fluent mesh requirements. Given it's one cell thick it's 2.5d and isn't quite 2d but doesn't show as 3d. I'll flag to DEV in case we need to update any readers.
      Moving back to your model, in 3d. Make the z boundaries symmetry and set the perimeter faces to suit the model. Assuming it's an airfoil (from the domain shape) pressure far field is a good starting point. Walls for the bit in the middle.
    • Rob
      Ansys Employee
      And a very quick response - we think GMSH isn't exporting a true 2d mesh. CFX also uses a 1 cell thick 2d approximation so the terms may have been muddled in GMSH. Fluent's 2d approach is true 2d in that there is no z coordinate.
    • ari9748
      Subscriber
      Thank you Rob. My issue is now solved. I have followed the GMSH---> Openfoam---> Fluent procedure and it worked well for my case.
    • QingFeng
      Subscriber
      Hello ari, I also met with the same problem of transferring Gmsh mesh file into Fluent case, but after drawing a C-grid type mesh for an airfoil by Gmsh, when I try to transfer it into foam format in a tutorial airfoil2D case, the transformation failed. But in Gmsh the mesh looks fine. I checked the volume, surfaces, and the physical boundaries, they are set correctly. Do you know how to deal with this kind of problem? Thank you!

    • Rob
      Ansys Employee
      That mesh doesn't look good - the jump in cell size is horrendous.
Viewing 17 reply threads
  • You must be logged in to reply to this topic.