[natantyt]

Hi

I have a packing of particles with overlaps enclosed by a box which forms the fluid region, I have used ansys automathic meshing with refinement leads to poor mesh with skewness over 0.9 and I can't get convergence in Fluent doing flow analysis . I have tried multizone failed due to input parameters. Do you have any idea of steps to get quality mesh on this complex geometry?

Sample picture below

[sarangdalne]

Try using patch conforming or patch independent methods in combination with the "On: Curvature and Proximity" advanced size function.
I am not sure which version of ANSYS you are using, if you have the latest then try using cutcell or tetra methods also. But, in that case you have to be cautious about the element size you give as your mesh resolution will be very much dependent on it. Also, if you just want to mesh the fluid region try the concept of virtual bodies and mesh groups.
Let me know if you specific questions.
Thanks,
Sarang

[natantyt]

Thanks for the reply,

I've used the patch conforming with proximity and curvature, I still get skewness of above 0.9 for around 100 cells out of 5 million. It is not possible for Fluent to convert them skewed mesh to polyhedral, so I need to localy manipulate them , do you have any suggestions. Refining makes the number of cells very high which will increase the computational cost very much.