Learning Forum

[Aras karimi]

Hi to all,

I am working on a 3D airfoil. Currently, I am conducting an independent mesh study to evaluate the dependence of the simulation results. My mesh is structured. The solution is done in steady state. My turbulence model is  KW-SST  with the following settings:

 

When I increase the number of grid cells to do the grid study, the message (stabilizing pressure coupled to enhance linear solver robustness) appears in the console and the residuals move hard and at a very slow speed. Is this a normal thing and does this problem always occur with the increase in the number of mesh cells?

Have I set the settings correctly?

Should I reduce the time scale factor?  And if I have to reduce it, to what number can I reduce it?

Should I lower its value at the beginning of the solution and increase its value during the solution?

Please let me know what I need to set up properly so that I can do a smooth study on the mesh.

 

Thanks in advance

Regards

 

[Federico]

Hello, 

does this warning go away with more iterations?

Also, how are you initializing your solution?

[Aras karimi]

No, this warning will not go away with more iterations. 

I initialize using hybrid.

Also, in the  'Reference Value'  section, I put the compute frame on the  inlet  boundary condition.

[Federico]

When you increase the number of cells, do you preserve mesh quality?

[Aras karimi]

Yes, the quality of the mesh is still maintained.

What is the meaning of this message?

I created 5 meshs to study the independence of the mesh solution. Time scale factor 1, which is the default of Fluent itself, was considered for all meshs. Only the first mesh, which had fewer cells, did not give this message, but meshs 2, 3, 4, and 5 give the message (stabilizing pressure coupled to enhance linear solver robustness).

When I reduce the time scale factor for meshs 2, 3, 4, and 5 to less than one, this message no longer appears. For example, I set the time scale factor to 0.7 for mesh 2 and no message appeared on the console. meshs 3, 4, and 5, which have more cells than mesh 2, required a time scale factor less than 0.7.

Is this behavior a normal behavior and for all problems with increasing the number of cells, the time scale factor should decrease?

[Federico]

The message means that your solution may be on the verge of diverging, so Fluent is introducing some mechanisms to stabilize it.

Reducing the time scale factor is another way to help with stability at the cost of convergence rate. If you're able to get a solution within an acceptable amount of computation time by reducing the time scale factor, then you can work with that.

But no, this is not "typical" when increasing the number of cells, unless the changes in the mesh result in worse mesh quality.

[Aras karimi]

Networks 2, 3, 4, and 5, which provide messages (stabilizing), have exactly the same quality as network 1, even some quality parameters of networks 2, 3, 4, and 5 are better than network 1.

I really don't know why I get the message (stabilizing) when increasing the number of grid cells. I am really tired.

What do you think I should do?

[Aras karimi]

Dear Federico,

Networks 2, 3, 4, and 5, which provide messages (stabilizing), have exactly the same quality as network 1, even some quality parameters of networks 2, 3, 4, and 5 are better than network 1.

I really don't know why I get the message (stabilizing) when increasing the number of grid cells. I am really tired.

What do you think I should do?

[Federico]

You can try reducing the under relaxation factors in Solution controls to help with stability.