[kyles.sverige]

I have been having trouble on initialization and get an error:
"Divergence detected in AMG solver: ads-0."

I have tried adjusting the relaxation factors, but it did not work. I have been on the forums many times looking for a solution and some say this error appears if you have a bad mesh quality.

I have been trying unsuccessfully for a few days to get an unstructured grid of a multi-element airfoil to pass a mesh check in ANSYS Fluent. When I check the mesh quality in Pointwise it says the grid has a good quality, but I get several errors in Fluent. Fluent output is printed like this when I check the mesh:

Warning: non-positive volumes exist.

   minimum volume (m3): -4.975500e-02
   maximum volume (m3): -1.800502e-07
     total volume (m3): -5.991529e+01
 Face area statistics:
   minimum face area (m2): 3.727322e-04
   maximum face area (m2): 4.325760e-01
 Checking mesh..........
WARNING: left-handed faces detected on zone   3:     0 right-handed, 130360 left-handed.
WARNING: left-handed faces detected on zone   4:     0 right-handed,    39 left-handed.
WARNING: left-handed faces detected on zone   5:     0 right-handed,    78 left-handed.
WARNING: left-handed faces detected on zone   6:     0 right-handed,    39 left-handed.
WARNING: left-handed faces detected on zone   7:     0 right-handed,  2782 left-handed..
   Zone   2: 263658 face(s) with wrong node order (on 1000000)..............
Done.

WARNING: Mesh check failed.

To get more detailed information about the mesh check failure
increase the mesh check verbosity via the TUI command
/mesh/check-verbosity.
Fluent can also try to fix the mesh check problems via the TUI command
/mesh/repair-improve/repair.

Any thoughts on how to repair this mesh? The "repair" command in the TUI did not work.
Thank you in advance for any help.

[bmahnic]

Kyle, hi!

Have you managed to solve the problem and how? I have the same problem.

Regards, Mark.

[kyles.sverige]

Hello Mark,

I was able to solve the problem, but to be honest this was several months ago now, and it is not really clear in my memory still. I do remember when I was getting errors I had several fields; structured around the airfoil, and unstructured far field. I was able to use existing wind tunnel data for the tested airfoil to validate a method of generation for a fully unstructured grid. This resulted in a single field which then was imported to fluent and passed all checks. However, I still used the TUI commands to further refine the mesh to achieve a higher grid quality in fluent after import. If you have any more questions regarding don't hesitate to ask, and maybe I can screenshot a few things I remember doing to help you out.

Regards,
Kyle

[bmahnic]

Hi, I managed to solve the problem, but I do not remember how either. I have a filling that I forgot to set "length" units to mm or something alike.

Regards, Mark.