https://linen.dev logo
Hi, over the last 1-2 weeks I have been doing some...
# general-help
c
Hi, over the last 1-2 weeks I have been doing some simulation runs with much more simplified setups. I used simpleFoam and no refinement cells. I varied the base resolution and turbulence models: kOmegaSST, SSG, LienCubic, LRR. For the secondary currents, the LienCubic model seems to give the closest results to the paper. I see secondary currents with upward movements and large vortices in the cross-sections. However, I don't observe vortex movements on the horizontal slice. Instead, there seems to be some "leaks" and discontinuity, where the flow seems to go to the points on the channel boundary at regular x intervals. When I checked the mesh, I saw some artefacts, and the leaks are right after these artefacts end. Other turbulence models show varying results. kOmegaSST shows weaker secondary currents/vortex structure in the cross-sections. Other models don't show coherent structures. There doesn't seem to be much difference in terms of mesh resolution, as in the pattern looks similar in the low- and high-resolution meshes. So I have some questions: 1. Why is there a significant difference in terms of flow results? I am primarily interested to know how to justify LienCubic showing the expected secondary currents pattern, and if I should use it with my own cases. 2. Is there any strategy to identify the cause of the artefacts, and have a cleaner mesh?