Help needed on swirling flow simulation

kpxx666

I'm working on the 3D swirling flow simulation in a combustor now. I use fluent5 as a solver. The unstructure mesh with tet element was used. The turbulence model I used is Reynolds stress model. The problem I get is it never convergent. After 1000 iteration, the residence curve becomes flat and never changes. Any one has good ideas? Actully, I need some general sugestions on the simulation of swirling flow.
    Thank you

kpxx666

No one can help me? sigh!

boling2000

I am sorry that I never use FLUENT and I do not know what is the difference between your situation with general cfd case. Did you take account the burning or the something special ?
As I know, the speed to compute internal flow is slower than external flow case.

kpxx666

In swirling flow, especialy in strong swirling flow, due to the vortex breakdown, a reverse flow zone will be generated. the flow is more complicated than general case. Another problem is that the turbulence in swirling flows is strongly anisotropic. So to get a good result, you have to use Reynolds stress model which increase the difficult of the numerical analysis.

boling2000

我曾经和别人一起做过S形管道内的涡流计算，里面出现了分离和再附着的过程，流动非常复杂，计算收敛很慢。