Hello, everyone,

Right now I am solving a cyclone problem and getting some troubles.

1) It is very difficult to get converged results. Even if I get a converged result, I still doubt if it is correct because the result shows that there is no reverse flow occuring at one of the outlets.

2) According to the Fluent manual, I should use RSM turbulent model to simulate this case(for strong swirling flow), but this model fails to give me a converged result. While RNG and realized model give me a converged result.

3) Some funny things happened. Sometimes I got a converged result with reverse flow at one of the outlets and sometimes not.(I used the same boundary conditions)

Bye the way, my current computational grids are tetrhedral elements( It is hard to use hex elements to generate the meshes for this case). Could this a problem?

Your advices and suggestions are highly appreciated.

1. Convergence : It is very natural that It is very difficult to get converged results for cyclone simulation. You may have couple of trials.

2. Turbulence model : It is very difficult to get 'Rankine vortex', which is common phenomenon of cyclone, by RNG model. So, many people recommend RSM.

3. Mesh : Frankly speaking, I do not exactly know why hexa is recommended. They say 'PRESTO' is prefered for cyclone simulation. Then hexa should be used. Anyway, it was difficult. But it is possible to generate hexa elements for cyclone modeling.

Sincerely, Jinwook

Hi,

Usually this problem is caused by using bounday conditions not correct represent physical model. Please check it carefully.

Thank you for both of you for the response.

So could you give me some advices how to set the correct boundary conditions for this case?

Thanks a lot!

how did you describe the air core boundary conditions?