Turbulent boundary layer resolution
 

Hey there Foamers. I am doing Large Eddy Simulation of turbulent natural convection boundary layers, and would like to ask if any of you experts can tell me the "conventional" methodology as to how many grid cells are required to resolve the boundary layer. Someone told me 35, and I just think that is too low. I currently have roughly 60 grid cells within the boundary layer.

So can anyone advice and possibly refer me to some references that discuss this particular issue.

Best regards,
Deji

Hi Deji,

I'm no expert but my advice would be not to focus on the number of cells but rather their size and distribution.

First cell should have a . Then use a geometric series with a factor of no more than 1.1.

I.e the cell height of the second cell should be at most 1.1 times the first cell. Third cell height should be less than 1.1 times the second cell height and so on...

Perhaps even a factor no greater than 1.05 should be chosen.

You should also have a look at the spanwise and streamwise values.

Best Regards

Nicolas

Hey there Nicolas. Thank you so much for the feedback. :)

Deji