[geth03]

hi all,

for couple of weeks i am trying to replicate a cyclone separator validation case:

https://www.sciencedirect.com/science/article/abs/pii/S000925091630121X?via%3Dihub

here, the kOmegaSST model is corrected by a curvature correction term to account for the swirling flow in the cyclone. i could implement that model and tested it already with another publication:

https://arc.aiaa.org/doi/abs/10.2514/6.1990-1484

the authors of the publication extended the outlets of the cyclone separator, so BC does not affect the internal flow as much as without the extensions. (see pic1). by doing that, the simulation matches the experiment pretty well (see pic2)

in my simulation, i even extended the outlets more. however, despite that, my simulation result is like the one from the authors without extensions. so the extensions did not change anything in my case. (see pic3)

i even changed the bottom outlet by adding a dustbin according to (Hoekstra, 2000) bc in the experiment a dustbin was used.

if anybody has experience with that i would be happy for any help!

kind regards