Personal tools
You are here: Home Forum Use Re: Hexahedron (i,j,k) algorithm fails

Re: Hexahedron (i,j,k) algorithm fails

Up to Use

Re: Hexahedron (i,j,k) algorithm fails

Posted by Abraham Vivas at February 23. 2020

Dear Salome users and experts,

I'm trying to generate an hexahedral  mesh for the boundary layer and wake of a blade. I subdivided it in a C-type topology and multiple blocks. I defined the node distribution on every edge, and the surface mesh looks perfect, but the Hexahedron (i,j,k) fails for one of the blocks with the following message:

"Invalid input mesh. 2D mesh is not suitable for i,j,k hexa meshing"

I attach a screenshot of the error message, a screenshot of a clipping showing the block that fails, and the python dump file:

https://drive.google.com/open?id=163L6XqY7ul5AyC0KrcT0pr9Y7isXWWFs

https://drive.google.com/open?id=1HAthrfo4zg0eNf01UJnZ_yZC4TQzPQ6m

https://drive.google.com/open?id=133gf9vnxlg4DevE-9DLrViC3laFhzKfN

Can anyone please have a look into it?

Many thanks in advance.

Best regards,

Abraham Vivas.

Re: Hexahedron (i,j,k) algorithm fails

Posted by Abraham Vivas at February 23. 2020
Update: I removed the third link (link to python dump file) because it pointed to a non-existing geometry. I add a link to a zip file with a working python script and the geometry file:
 

Re: Hexahedron (i,j,k) algorithm fails

Posted by SMESH expert at February 25. 2020

Hello Abraham Vivas

Your case reveals a bug in Hexahedron (ijk) algorithm. Thank you for reporting it. A fix will be available in v9.5.0.

Meanwhile I can propose the following workaround. Compute 2D mesh on geometry, then copy this mesh (menu Mesh > Copy Mesh), assign Hexahedron (ijk) to the copy mesh and compute this 3D mesh.

I modified your script to demonstrate what I mean.

Best regards

Edward

Attachments
Powered by Ploneboard
Document Actions