Learning Forum

[karl.zammit]

Hi all,

UPDATE: Why does meshing work when the geometry is saved as .stp and not when in native .dsco/.pmdb/.scdoc/.agdb ??

---

I’m struggling to create even a coarse surface mesh in Fluent Meshing for what should be a straightforward case: half a wing in axial flow inside a symmetric cuboidal (wide) domain. I’ve meshed more complex geometries before without issues, so this one has me stumped.

Geometry & setup:
* Created entirely in ANSYS Discovery (no imports).
* Wing trailing edge deliberately blunted to avoid zero-thickness.
* Airfoil/wing faces ensured to be single continuous surfaces (no fragmented profiles).
* Domain is a simple box; half-model with a symmetry plane.
* Using Fluent Meshing (Watertight workflow and manual steps both attempted).

What I’ve tried:
* Discovery Repair: reports no problems.
* Fluent Meshing Geometry Check: no errors or warnings.
* Tried very coarse global sizing, relaxed curvature/proximity, max cell size on all faces.
* Ensured no free edges, duplicates, or tiny sliver faces as far as the tools report.

The problem:
* The surface mesh generation fails every time, even with extremely coarse settings.
* I consistently get errors such as "Breakout detected" and "Front could not be closed" at the tip during surface meshing.
* No preceding geometry/health warnings to point at a root cause.

Question:
Is there a preferred export route from Discovery for Fluent Meshing?

Happy to provide the files for you to try it out and give out any additional details. I’ve been at this for a few weeks and find it odd given the simplicity of the setup. Any ideas or checks I might have missed would be greatly appreciated!

Thanks in advance.

 

[Anilkumar]

Hi Karl,

Thanks for reaching out to Ansys Forum Support.

We would definitely recommend to use native formats as those will carry all the share topo, NS information which will be missing from neutral format like STEP. Could you please provide more details on the Dimensions? It appears to be issue with dimensions. Like the model size seems very large. Could you please try meshing the scaled down model of this? maybe 10 or 100 times scaled down model?

Secondly, try using standard unit system in the discovery instead of large unit system. Let me know on how this goes.

If it still fails then could you please share the transcript file details here? copy paste the details from geometry import to failure stage.

Thank you!

Anil