[Crank-Shaft]

Hey CFDOnline,

I have finally found my old login details and managed to return to the forums after a period of hibernation lasting almost 7-8 months.

I am currently working on the development of an in-house research code at a university, in an attempt to implement a DES solver. I have been trying to create meshes for simplified test cases for modelling and solver verification.

I have experience with ICEM however, I've never tried to output a structured, multiblock mesh with a CGNS format. The current domain consists of a cylinder of 100 mm diameter and the dimensions of the enclosure are 4000x2000 (LxH). This parasolid (.x_t) file is attached.

My main approach was a simple O-grid around the cylinder and then stretched blocks outwards to the outer boundaries. This gave me 12 blocks with a single cell width in the spanwise direction since it is meant to be a 2D simulation at this stage. The boundary conditions were defined at the outer surfaces and since the inflow and outflow consisted of 3 blocks each, each BC had to be shared between these blocks. I suspect this is where the errors are. I have tried simple subsonicInflow, subsonicOutflow and top/bottom/left/right inviscidWall.

I took 3 different approaches (ICEM, WB, WB export to ICEM) however, it seems like there is a persistent error with the .cgns output. I was checking this by trying to read it into Fluent. Checking the mesh file in Tecplot also revealed that VORFN blocks are being exported through and I have tried several times to prevent this, without much success. Only the ANSYS WB output in CGNS seems to contain the interfaces or subfaces between the adjacent internal blocks however the ICEM exports do not have this.

I would greatly appreciate it if someone with experience in ICEM multiblock meshing was able to provide a method to export the CGNS mesh so that the BC and blocks are well defined, for Fluent import. I look forward to your suggestions since this is actually rather urgent.

[Crank-Shaft]

I tried this with a 2D geometry with a circular wall BC in the middle of the domain.

Then, the boundary conditions were all assigned to the edges and the interior meshes
were all generated. However, I encountered the same problems when trying to export the meshes and open them in Fluent.

Very confused at this stage. Has anyone faced this problem with anything except CGNS format output?

[Crank-Shaft]

Hello everyone,

I am still struggling with this problem of exporting the domain correctly with a multiblock, structured mesh in CGNS format while keeping all the boundary conditions intact.

In order to assess and isolate the problem, I tried several different workflows including making the entire geometry in ICEM and importing the geometry, while relying on ICEM to block and mesh.

I have changed the domain slightly, mainly due to the updated position of the boundary conditions. The attached file contains the project files and geometry. I plan on trying out an O-grid around the cylinder, which when expanded out leads to 3 blocks being shared by the inlet and outlet BCs respectively.

Please check them out if possible and I look forward to your suggestions, especially if you are able to successfully export a mesh to Fluent through either the new 3.0 CGNS or the old 2.4 CGNS libraries.

Thank you

[Mdong]

hi ,i meet the same problem.If you have solved the problem,please give me a hand.Thanks!