Learning Forum

[JMA337]

Hi,

I keep getting the following error when creating a hybrid (tetra+prism) volume mesh for a complex geoemtry. The surface mesh has good max skewness (0.59), good min dihedral (40deg) and there are no free faces, self interceptions or any other issue. I am generating the volume mesh from the linux environment, starting from the surface mesh. Using the auto-mesh command, everything is fine (no stair stepping of the prism layers) until it gets to the pyramid generations stage, where the following error occurs:

Retriangulating adjacent zone int-box-in:875.

 

Remeshing overlapped thread int-box-in:875 ...

Deleting overlap on int-box-in:875 ...

Improving mesh on int-box-in:875 ...Done.New node zone boundary-node-35148 created for boundary nodes not in current domain (tetrahedral).

 

Generating Pyramids...

 

Creating pyramids

prism-side-35058 (6944 faces)

 

===============Message from the Cortex Process================================

 

Compute processes interrupted. Processing can be resumed.

 

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

 

*** ERROR: multithreading violation ***

You are running a multithreaded application but

you set MLX5_SINGLE_THREADED=1. Please unset it.

*** ERROR: multithreading violation ***

You are running a multithreaded application but

you set MLX5_SINGLE_THREADED=1. Please unset it.

 The tgrid process could not be started.

 

Warning: An error or interrupt occurred while reading the journal file.

Some commands may not have been completed.

 

 

Meshing> Halting due to end of file on input.

Transcript Stop Time:  14:46:27, 16 Jul 2024 

Total Transcript Time: 4 Minute 47 Second.

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

I've not yet had success solving this problem...Any suggestion on what the origin of this error may be? 

I've tried running the mesh generation on a single node to limit the internode communication, but that did not solve the issue. Reducing the number of prism layers did not make a change either.

[Keyur Kanade]

Can you please check if you can generate volume mesh without boundary layer (prisms).

 

Regards,

Keyur

How to access Ansys Online Help Document

Guidelines on the Student Community

Fluids Engineering Courses | Ansys Innovation Courses

 

[JMA337]

 

Hi Keyur,

Yes, I can confirm the volume mesh is generated fine without prism layers.

Kind Regards,

 

 

[Keyur Kanade]

Ok. Then it is prism issue. Please make sure that you have selected correct walls to generate prisms. Make sure that prism ends correct at the ends on other walls. Please check following image. 

 

Regards,

Keyur

How to access Ansys Online Help Document

Guidelines on the Student Community

Fluids Engineering Courses | Ansys Innovation Courses

 

[JMA337]

Dear Keyur,

Thank you so much. Indeed, there was a typo in my script by which prisms where not being greated on some walls, leading to the segmentation error. That has now been solved. 

Thank you!