---

Hi,
I am simulating the flow between 2 rotating coaxial cylinders with radius of 0,25 and 0,20 m. I´ve defined my inlet, a ring at the bottom of the anular space, like a velocity inlet with 1m/s in axial direction, the outlet, a ring at the top, like a pressure outlet and the fluid like a reference frame which rotates at 1000rpm. I´ve used the Reynolds Stress Model to model the turbulence but I still have a lot of problems to get a converged solution. The residuals remain high and with oscillations, specially the continuity, because ich have a lot of reversed flow and the balance of mass flow doesn´t agree. Can anyone help me to loose this problem?!

Thank you very much

Xiana

Eine Antwort auf diesen Beitrag verfassen (mit Zitat/Zitat des Beitrags) IP

---

Hi Xiana,

is there a special reason to use the Reynolds-Stress model? The turbulence seems to be almost caused by the shear conditions and if your flow is almost axial than other turbulence models (such as realizable k-eps) should work as well and run more stable. Of course if you have large swirls (as in cyclones) you have to use Reynolds-Stress.
Maybe there is no need to use the reference frame model. You can also define a rotating wall by applying the angular velocity on the wall.
I sometimes make the experience that the solver has problems to converge, when the pressure outlet isn't on a physical isobaric surface, meaning that there is a pressure gradient in the outlet. Maybe you can arrange the outlet surface, that it is almost isobaric.
Maybe you should consider as well another initialization of the solution.
I hope this helps

Ulrich

------------------
Dr.-Ing. Ulrich Heck
ulrich_heck@dhcae.de
http://www.dhcae.de

Eine Antwort auf diesen Beitrag verfassen (mit Zitat/Zitat des Beitrags) IP