I believe you used Google Translator to translate your question. However, I think it would be better to post your question in your original language then probably your problem would be clear to us.

For resolving the boundary layer it is preferable to use structured cells need the walls, but this does not mean you cannot use unstructured mesh cells. As long as your cell height (either structured or not structured) lies within the recommended Y+ region, I believe you are good to go. For wall functions, you can use unstructured mesh cells without problems, and you can have large sizes (Y+ up to 30 or even 100).

But in general, for CFD simulations, the structured grid is better and can be easily controlled.

Thanks Mr. Ahmed Osama Mohamed Elhassan Mahgoub

If the mesh is structured in all geometry, there is no need to use a boundary layer?

اگر در تمام هندسه مش ساختار یافته باشد،آیا نیازی به لایه مرزی نیست؟(برای جریان های آشفته)

Hi Amir Hossein Zakeri ,

In unstructured mesh, you can set the first layer height using the inflation option.

Also, in a structured mesh, you can set that using edge sizing and proper "number of divisions" and "bias factor."

Amir Hossein Zakeri

Perhaps you are confusing some concepts. There is always a boundary layer if you have a flow near a wall. But you could ask if I used structured mesh, do I need to use the inflation option in ANSYS Meshing? and the answer is: you can't because the inflation option is for unstructured grid only. In the inflation option, you can select the height of the first layer, or the total thickness, define a growth ration, and you have all the control parameters required. It is tricky though, because sometimes the inflation layer is compressed in order to avoid collision with layers from other surfaces.

But when using a structured grid, you have all control (which includes the size of the grid cells near the wall).

Thanks Mr. Ahmed Osama Mohamed Elhassan Mahgoub

Perfect,Perfect

I noticed exactly, thank you very much

Amir Hossein Zakeri Add to the PV answer, it is up to the turbulence model as well. To give you an idea, if you are using Reynolds stress or LES models, you don't need to consider the boundary layer, whereas, in the case of employing RANS models, the boundary layer is a must. So first determine your turbulence model and why are you going (have to) make use of it.

Regards

Hamed Majidian ,Model is k-epsilon standard and near wall treatment is enhanced wall function. My goal of considering this model is to reduce the cost of the calculation and reduce calculation time.

If we use LES models, does the simulation solving time increase?

Thanks

Amir Hossein Zakeri

absolutely, they are out of comparison. K-epsilon model normally is being used for high Reynolds flow, so in your case, if you are not going to implement BL meshing, it means the BL phenomena are out of interest of your simulation or has a less impact on the whole behavior of flow, particularly due to high Reynolds number. This is because, the high the Reynold number is, the less thickness is in the BL. As a result, it can justify why the BL mesh was not favored.

Regards,

I believe it would be better that you spare some time on the physical concept of the phenomenon rather than sticking to the CFD software. I would recommend watching

http://kashti.ir/%d8%af%d8%b1%d8%b3-%d9%87%db%8c%d8%af%d8%b1%d9%88%d8%af%db%8c%d9%86%d8%a7%d9%85%db%8c%da%a9-%d9%be%db%8c%d8%b4%d8%b1%d9%81%d8%aa%d9%87/

lectures 21 to 28 first.

Regards