[pi120]

Hi,

Anybody knows a good rule of thumb for maximum aspect ratio allowed in the cells which are in the first prism layer?
(I am computing the Aspect ratio as the ratio between the 'Target surface size' and the height of the first prism layer)

Exceeding a certain value should lead to problems in accuracy and convergence, if I remember well. But what is such value?

Best
Pietro

[Peresvet]

Originally Posted by pi120

Hi,

Anybody knows a good rule of thumb for maximum aspect ratio allowed in the cells which are in the first prism layer?
(I am computing the Aspect ratio as the ratio between the 'Target surface size' and the height of the first prism layer)

Exceeding a certain value should lead to problems in accuracy and convergence, if I remember well. But what is such value?

Best
Pietro

The value is that your life won't be long enough to find it. As for the CFD: I have been looking for 5 years with the help of 600 cores and cannot find the point of using it at all.

[ashokac7]

Originally Posted by pi120

Hi,

Anybody knows a good rule of thumb for maximum aspect ratio allowed in the cells which are in the first prism layer?
(I am computing the Aspect ratio as the ratio between the 'Target surface size' and the height of the first prism layer)

Exceeding a certain value should lead to problems in accuracy and convergence, if I remember well. But what is such value?

Best
Pietro

Yes. It depends. What is the complexity of physics, where in the domain etc. Ideal number is 1:1. But there is no specific number to distinguish the accuracy levels. It's the gut feeling. These studies should be done at starting. For some cases we would be putting 20-30 prism layers in 1mm. Then obviously you wont be able to maintain even 1:5. And nowadays softwares are quite robust. You may be good enough solution for bad meshes. Hope this helps.