CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   define "yplus" in custom field function in fluent (https://www.cfd-online.com/Forums/fluent/227042-define-yplus-custom-field-function-fluent.html)

Mutlaq May 15, 2020 09:25

define "yplus" in custom field function in fluent
 
when I want to compute y+ in fluent it computed for the first layer only. so how can I define y+ in CFF I Know the law of y+=y*ut/kin.visc

but what the Y means in fluent how can I define

anyone can help?

vinerm May 15, 2020 09:35

Yplus
 
y is just coordinate normal to the wall. The challenge is not with y but with u_\tau. This depends on wall shear stress and is not easy to define for cells that are not adjacent to the wall. The major challenge is in separating wall shear from other shear layers.

Mutlaq May 15, 2020 10:01

Quote:

Originally Posted by vinerm (Post 770613)
y is just coordinate normal to the wall. The challenge is not with y but with u_\tau. This depends on wall shear stress and is not easy to define for cells that are not adjacent to the wall. The major challenge is in separating wall shear from other shear layers.

I got the y+=20 but in the spanwise direction as the no. of the layer that I put.

and plot the y+ vs u+ in the same direction but that doesn't mean anything because of that not in the normal wall direction.
I know that define y+ isn't easy but I need the code or formula of it because of all standard plots in books and researches are plots with y+

vinerm May 15, 2020 10:07

Yplus plot
 
It depends on your objective. No one plots y^+ vs u^+ from CFD since the graph is actually input to CFD. If you want to do it, you have to write a UDF and it won't be easy. The graphs that you see in books is based on experimental data and there is no benefit of regenerating it using CFD. If you are trying to do it for verification, then you have to either come up with a way to separate wall shear from turbulence shear or do multiple simulation with different velocities so that both, y^+ and u^+, change for each case.

Mutlaq May 15, 2020 10:19

Quote:

Originally Posted by vinerm (Post 770627)
It depends on your objective. No one plots y^+ vs u^+ from CFD since the graph is actually input to CFD. If you want to do it, you have to write a UDF and it won't be easy. The graphs that you see in books is based on experimental data and there is no benefit of regenerating it using CFD. If you are trying to do it for verification, then you have to either come up with a way to separate wall shear from turbulence shear or do multiple simulation with different velocities so that both, y^+ and u^+, change for each case.

thank you very much.

the last point. Would you recommend using the periodic boundary conditionز
I actually used it but I need your opinion

vinerm May 15, 2020 10:22

Periodic BC
 
That depends on the simulation. Periodic condition exists because it is useful but the usage is based on whether the physics and geometry, both, are periodic or not.

Mutlaq May 15, 2020 10:31

Quote:

Originally Posted by vinerm (Post 770639)
That depends on the simulation. Periodic condition exists because it is useful but the usage is based on whether the physics and geometry, both, are periodic or not.

I used the periodic to get fully developed for the straight channel then I used the profile to get the outlet condition of periodic as an inlet condition for another channel that has the same mesh, inlet, and outlet area but has a wavy in the bottom wall. so I used periodically only to get fully developed for a small length

vinerm May 15, 2020 10:33

Good
 
That's an appropriate use of translational periodicity.

Mutlaq May 15, 2020 10:36

Quote:

Originally Posted by vinerm (Post 770643)
That's an appropriate use of translational periodicity.

so that good choice and not wrong?

vinerm May 15, 2020 10:39

Correct
 
Yes, that's correct choice.

Mutlaq May 15, 2020 10:40

Quote:

Originally Posted by vinerm (Post 770647)
Yes, that's correct choice.

Thank you very much

Mutlaq May 18, 2020 13:33

Quote:

Originally Posted by vinerm (Post 770647)
Yes, that's correct choice.

Mr. vinerm can I ask you about adaption mesh in fluent if I need to use it or no?

vinerm May 18, 2020 15:23

Adaption
 
Mesh adaption is process to refine (or coarsen) the mesh in specific regions. If you already have the mesh that can resolve the required fields, then you don't need to worry about mesh adaption. Mesh adaption is one of the least used approach in most of the simulations.

Mutlaq May 19, 2020 00:04

Quote:

Originally Posted by vinerm (Post 771008)
Mesh adaption is process to refine (or coarsen) the mesh in specific regions. If you already have the mesh that can resolve the required fields, then you don't need to worry about mesh adaption. Mesh adaption is one of the least used approach in most of the simulations.

thank you for reply

Mutlaq May 27, 2020 10:10

Quote:

Originally Posted by vinerm (Post 771008)
Mesh adaption is process to refine (or coarsen) the mesh in specific regions. If you already have the mesh that can resolve the required fields, then you don't need to worry about mesh adaption. Mesh adaption is one of the least used approach in most of the simulations.

Hi Mr. vinerm I need to ask you how can I calculated CLF number any mesh size I need to use with y or x or z

delta t= (CFL*mesh size/velocity)

vinerm May 27, 2020 11:39

Cfl
 
For CFL, you need to use a scale and not exact value of x, y, or z size of cell. Usually, cube root of volume is used, provided the mesh is hex. If it is tet, then you can use cube root of (V_{cell}\sqrt{72})^{(1/3)} as length scale.

Mutlaq May 27, 2020 20:20

thank you very much

Mutlaq June 1, 2020 06:45

https://drive.google.com/file/d/1Lhn...ew?usp=sharing
Quote:

Originally Posted by vinerm (Post 772202)
For CFL, you need to use a scale and not exact value of x, y, or z size of cell. Usually, cube root of volume is used, provided the mesh is hex. If it is tet, then you can use cube root of (V_{cell}\sqrt{72})^{(1/3)} as length scale.

Hi Mr. Vinerm I want to ask you about this type of mesh if the mesh good or no, Because as I mentioned earlier I used a periodic inlet profile take it from a straight channel and another reason the mesh in someplace is very close in the streamwise direction

https://drive.google.com/file/d/1Lhn...ew?usp=sharing

vinerm June 2, 2020 09:38

Mesh
 
I am afraid I cannot access Google Drive. So, better to attach a few pictures here in the post.

Mutlaq June 2, 2020 10:07

1 Attachment(s)
Quote:

Originally Posted by vinerm (Post 773026)
I am afraid I cannot access Google Drive. So, better to attach a few pictures here in the post.


this is a mesh


All times are GMT -4. The time now is 11:41.