3D sliding meshes
Hi everyone, I am struggling with a 3D sliding meshes simulation of mixer utilizing a rotating rod. I set unit in micrometer in workbench DM.
I have done simulation in 2D successfully. It's a single rod rotating in a channel. 2 velocity inlets inject 2 different liquid which have same property as liquid water, pressure outlet...use species transport model I set a static zone and the other is a moving zone. (the rod is in the moving zone) the moving zone is a cylinder, which means the static zone has 3 faces interact with the moving zone. these 6 faces (plus the faces in the static zone) are set to be interface. (I made the setup similar as 2D simulation) starting fluent, I set the 6 faces as 3 pairs in mesh interfaces setting. however, when running the simulation everything goes well expect that the fluid couldn't enter the moving zone. in the result contours, I checked the velocity and mass fraction, it showed that the data in static and moving zone were not connected well, the moving zone did rotate but the interfaces acted as "walls" to block the liquid to flow into the moving zone. |
Hi,
Just to be sure, you did match the interfaces under `mesh interfaces' right? (otherwise the simulation probably wouldn't run anyway...) Your problem sounds curious, I don't see a particular reason why it wouldn't work. Did you try MRF as well? Does that work? |
Quote:
I created the name selection in Mesh cell Is it correct to match the 6 interfaces as 3 pairs? the solution was converged but solved 2 zone separately. I noticed in the boundary condition window there were several "walls" and I couldn't see them in display meshes option. ( I identified the walls belong to static zone) |
There are quite often some `empty' walls in the b.c.'s, shouldn't be a problem.
With 3 pairs of interfaces, do you mean a pair `top', `bottom' and `side' or something else? If you mean something else, it may be a problem, because every interface should be pared with every interface it can slide along I think. But in the top, bottom, middle case that should work. A simple but easy to overlook thing may be a wrong definition of the rotation axis? And I think you should try MRF - it shouldn't take much time, and if that does work, you may rule out some potential issues already. |
Quote:
When I tried simulate with the species transport model on, I encountered the same problem again... But the way what is the difference between MRF and sliding mesh? It seems MRF works in steady state, sliding in transient, but the result are not the same? |
http://www.tu265.com/di-49858dd39a41...3a674e0d87.png
this is what I got, please help! |
The green spot does indeed look odd. But it doesn't seem to extend vertically? Or can we not see the vertical extent at this point?
|
All times are GMT -4. The time now is 05:34. |