[recepky]

Hi,

I have a problem that contains a cyclone geom. and has lots of porous jumps in the center.

I meshed the entire geometry with hexa mesh and I ended up with a good quality mesh, only some complex sections of the geometry has tetrahedra mesh.

I solve the problem using Reynold Stress Model, without porous jumps, but when I set the porous jumps and set the constant values for it and run calculation after 50 iteration solver become diverged. What I am doing wrong?

[A CFD free user]

Originally Posted by recepky

Hi,

I have a problem that contains a cyclone geom. and has lots of porous jumps in the center.

I meshed the entire geometry with hexa mesh and I ended up with a good quality mesh, only some complex sections of the geometry has tetrahedra mesh.

I solve the problem using Reynold Stress Model, without porous jumps, but when I set the porous jumps and set the constant values for it and run calculation after 50 iteration solver become diverged. What I am doing wrong?

what do you mean porous jump?
I worked on a cyclone with RSM turbulence model. It's somehow difficult to get a good converged solution, but don't know anything about porous jump? Why is it used for? Maybe you need to start with a simpler turbulence model such as k-e then, after a few iterations go to RSM. Even, you need to start an unsteady simulation. Anyway, I'll be glad to hear about porous jump. thank you

[recepky]

Porous jump is a boundry condution that can be used for any adjacent walls, porous jump is 1D solution equvalent of 3D porous zone.

[oj.bulmer]

Are you sure you have put the right permeability and C2 values for porous jump? Also, any specific reason for RSM model?

I have so far observed that Realizable model works best with the porous jumps, amongst all other 2-equation turbulence models. Perhaps, if you really want RSM, you can start with realizable model, have a decent convergence, and then switch to RSM.

OJ