[laolartet]

Dear members

I'm trying to mesh a stirred tank, but I'm having some problems,in this moment I am considering only the fluid and one impeller (a rhuston turbine), however the final reactor have: 2 impeller, 8 “accessories” (4 from the bottom and 4 from the top), 1 sparger and 4 baffles, the first image is a first approximation, I know I must simplify it.

I’m working in Design Modeler and because of the complexity of the model I started with only one agitator and the tank, then I generated the “fluid”, and after that volume for MRF with “slice material” thenI extracted the volume of the impeller with “slice material”and “suppres body”. (Second image)

I tried meshing (ICEM ansys meshing) with hex dominant (External volume or fluid) and tetra (internal MRF), but when I execute it on fluent it doesn’t converge, I tried also meshing just tetra and it improved but still doesn’t converge.

How can I improve the meshing?
Am I creating the volume properly?

MRF or Sliding mesh? I work with MRF but I read that by the number of accessories is better work with sliding.

Thanks in advance

Arturo

[jotac]

Dear laolartet,

have you solved your problem?

I'm doing something similar, but I'm facing many problems to mesh my geometry with hexahedral elements..

Thanks in advance for anyone that can help.

J

[ztdep]

dear friends:
"did not converge" may not caused only by the mesh system, but it has a great effects on the solution process.
I think you geometry is relatively regular, just make a multiblock grid systems with icemcfd.
the grid number can be decreased greatly, and the solution process can be more stabilized.

tried to use more robuster solver such as the fully coupled solver in fluent.