[noodleluvmay]

Hello,

This type of question might have appeared quite a lot in the forums, but I could not really find a definitive answer for it, so I would greatly appreciate it if anyone could help with the required y+ values for the turbulence models in Fluent:

1. Does k-omega Standard model use wall function at all? If so, does it mean that the y+ range should be between around 50 - 500? If not, what kind of range would the y+ value require?

2. As far as I am aware, k-omega SST does not use wall function and solves viscous layer directly, but some people have said it would require a y+ value of 1, some have said 5-10, so I was wondering if there might be a more certain answer for it?

Thank you!

[vinerm]

As far as Fluent is concerned, each model uses Enhanced Wall Treatment. This is a two layer model and in reality is insensitive. So, it can work with all values up to 500 or so. However, is a low Re model, i.e., it's applicability is more in the region with low Re. So, it is recommended that the should be of the order 1, i.e., between 1 and 10.

[noodleluvmay]

Originally Posted by vinerm

As far as Fluent is concerned, each model uses Enhanced Wall Treatment. This is a two layer model and in reality is insensitive. So, it can work with all values up to 500 or so. However, is a low Re model, i.e., it's applicability is more in the region with low Re. So, it is recommended that the should be of the order 1, i.e., between 1 and 10.

Hello, thank you for the quick reply! I might not be understanding you here correctly, but what do you mean when you have said that k-omega will work for y+ up to 500, but recommended between 1-10?

Would the previous one (1-500) be applying for k-omega standard, and latter (1-10) for k-omega SST?

Also, would you mind me asking what kind of range of Reynolds number would usually be considered as low Re simulation for external flow?

Thank you very much!

[vinerm]

has got nothing to do with . So it doesn't care how fine or coarse the mesh is. It is the wall treatment. And the wall treatment used by all models in Fluent is EWT. And EWT is insensitive, i.e., it can work with high as well as low values of . However, if you wish to resolve the viscous sublayer, most likely because the objective demands that, you should maintain below 10.

is nothing but Reynolds number based on the mesh size. And when it is of the order 1, this implies that the viscous forces are comparable to inertial forces, i.e., the equations can resolve the flow-field without any special treatment. Therefore, it is supposed to be of order 1. Do note that order 1 does not imply 1, rather values up to 9.

[LuckyTran]

Standard k-omega does have wall functions that kick in around wall y+ > 100. However, it is against the philosophy of the k-omega model to be using wall functions in the first place. The wall function is there just to make the k-omega model usable for coarse meshes. The wall function for omega is also quite garbage and not as robust compared to wall functions in epsilon models.

This wall function also exists in k-omega SST which blends a k-epsilon with a k-omega solution. But because the blending favors k-epsilon far from boundaries and k-omega near boundaries, the wall function that exists for the omega equation practically has no effect.

In either case, the philosophy of using either the k-omega or k-omega SST model is to resolve viscous layers directly, so the recommendation is always going to be super low wall y+ when either of these models are involved.

Originally Posted by noodleluvmay

but some people have said it would require a y+ value of 1, some have said 5-10, so I was wondering if there might be a more certain answer for it?

What's much more important than achieving the target y+ on only the first grid cell is resolution. You want to pack 10/30/100 cells across the boundary layer, etc.