how many points to resolve a boundary layer?
 

Does anybody know what is the minimal number of grid points to resolve a boundary layer? Is there any general method/rationale/experience to determine this?

Thanks,

Jiaying

Re: how many points to resolve a boundary layer?
 

Sorry, forgot to mention that I am going to use FVM with LRN k-e turbulence models.

The problem is three-dimensional, but I am asking how many points along the wall-normal direction needed to resolve a bounday layer, to calculate flow friction and heat transfer. Let's assume the geometry is not that complicate (say, a flat plate).

Thanks,

Jiaying

Re: how many points to resolve a boundary layer?
 

Hi,

According to the Star manual it is recommended to use at least 20 boundary layers with a y+ value of app. 1 near the wall.

Regards,

Volker

Re: how many points to resolve a boundary layer?
 

heat transfer - so the thermal boundary layer. while resolving the bl one should take care of the prandtl number - or not?

Re: how many points to resolve a boundary layer?
 

Please have a look at "numerical computation of Internal and external flows" by C. Hirsch. I think you will have a general idea.

Re: how many points to resolve a boundary layer?
 

Re: how many points to resolve a boundary layer?
 

Thanks, I'll search for that.