Learning Forum

[btingthewind]

HI, recently I am generating  a mesh for a louvered fin geometry with ANSYS MESHING, the thickness of the fin is 0.1mm, and there is a boundary layer of  ten layers on the fin surface? However, when i finished the meshing, i found the minimum orthgonal quality  was as low as 0.02, obviously it cannot be processed in FLUENT. And various trials have been done by me to improve the quality, but they all did not work. So, can someone do me some suggestions to improve it?

BSET REGARDS

[Rob]

The starting point for this is to understand where and why you have poor quality cells: otherwise I could just reel off a long list of things that will work but not necessarily be causing you a problem. 

Use the quality graph and pick the lowest 2-3 columns to highlight the poor cells. Identify why they're bad, and then see what can be done to change this. Post a few screen shots too. 

[btingthewind]

Thank you sir.

All the bad elements present at the end of the louver, I've used the face sizing, edge sizing ,body sizing etc. but all did not work well. So ,what i can do next?

[Rob]

It's hard to tell from the images: do you have a thin face at the end of the louvres? What is it for?

[btingthewind]

THX,sir, maybe my discription is not clear, all the faces construct the LOUVERED FIN. The fin I am using is depicted in the following fig, may it help.

[Rob]

Depending on the application you may be able to use a thin wall in Fluent: we don't need to model metal thicknesses in most applications. This avoids the small face at the end of the unit. 

[btingthewind]

THX,sir.

To use thin wall in fluent is a good solution in some specific situations, but my work now has to consider the thickness of the fin, so thin wall is not applicable.

BEST REGARDS!

[Rob]

If it's for thermal properties read up on shell conduction. Otherwise you'll need to resolve the plate thickness which will increase the cell count. 

[btingthewind]

THX, sir.

The bad elements are enclosed by the red lines in the following pic, do you think what shoud i do to have them resolved? or what parameter in the inflation settings relate to the bad elements? 

BEST REGARDS!

 

[Rob]

The problem is caused by the transition from the inflation layer to the first cell. Mesh without the inflation and compare the effects. 

[btingthewind]

THX, sir,

I've tried what you suggested, the mesh without inflation layer can reach a  minmun aspect ration above 0.15,  but I am going to use KW SST in fluent, so there must exist inflation layers.

What can I do next?

BSET REGARDS

[Rob]

Then you need to do something to avoid the inflation terminating and giving poor pyramid caps. When you say you need to consider the plate thickness why is that? 

[rajkumar444]

Do meshing in ansys fluent. Polymesh will work for one fin mesh.