[Nick R]

Hi all,


I know this question has been asked before but I would like to get some confirmation (preferably in a reference if such a document exists!).

In a simulation I need yplus to be one and this corresponds to a high aspect ratio of 1000 near the walls unless I include more nodes in the normal direction which would cost me a huge computational time.

I've heard that aspect ratio is not that important in the near wall region. Is this correct? Would the value of 1000 still be acceptable?

Cheers,
Nick

[cfd_newbie]

Originally Posted by Nick R

Hi all,


I know this question has been asked before but I would like to get some confirmation (preferably in a reference if such a document exists!).

In a simulation I need yplus to be one and this corresponds to a high aspect ratio of 1000 near the walls unless I include more nodes in the normal direction which would cost me a huge computational time.

I've heard that aspect ratio is not that important in the near wall region. Is this correct? Would the value of 1000 still be acceptable?

Cheers,
Nick

I am replying to your question with my own understanding.
I think the aspect ratio of 1000 is high and apart from solver stability issues, you may end up not capturing the right physics. In order to resolve the boundary layer Y+ is not the only criteria but you have to place adequate number of cells near wall to capture the boundary layer well.

[Nick R]

Thanks for the comments. I already have enough points in the streamwise direction to capture the Boundary Layer but I was wondering if the normal (into the wall) is important cause these are the nodes that would effect the aspect ratio in my case.

[cfd_newbie]

Originally Posted by Nick R

Thanks for the comments. I already have enough points in the streamwise direction to capture the Boundary Layer but I was wondering if the normal (into the wall) is important cause these are the nodes that would effect the aspect ratio in my case.

Yes the normal points to the wall are damn important. your ratio of cells thickness in the boundary layer should be around 1.05-1.10 (So if you first cell thickness is x, then the next cell thickness would be 1.05*x and so on). If you problem is external aerodynamics you should have this growth of cells to about 50 layers. If you problem is internal flow you should have at least 10-15 layer thickness.

[Nick R]

I am talking about into the wall nodes (spanwise for example in external aerodynamics).

[cfd_newbie]

Originally Posted by Nick R

I am talking about into the wall nodes (spanwise for example in external aerodynamics).

Can you post a few pics of your mesh ?