[MachZero]

I am currently running a study where i am looking at particle deposition. Particles are very small, in the 1-10 micron range. I have gotten baseline results and everything seems reasonable. I am running a mesh sensitivity study where i refine my mesh and prism layers. I have two geometries, one is a small tube, and the other is a larger tube.

Small tube (running SST, but flow is laminar): What i am seeing is that, for the small particles on the order of 1 micron, the more I refine, the amount of deposition is reduced. It isn't converging, even with a very fine mesh. At 4 microns, things stabilize at the 1st refinement of mesh.

Larger tube (running SST, flow is turbulent): For 1 and 4 micron, particle deposition continues to decrease with mesh resolution, but at 6 micron it is stabilized through all meshes.


I ran a timestep and particle count sensitivity study, and both show that my results do not change with these factors.

What could be causing the continued decrease in particle deposition, without stabalizing. Am i going about this mesh sensitivity wrong? Or am i forgetting something?

Running in STAR-CCM+ 18.06.

[FMDenaro]

Originally Posted by MachZero

I am currently running a study where i am looking at particle deposition. Particles are very small, in the 1-10 micron range. I have gotten baseline results and everything seems reasonable. I am running a mesh sensitivity study where i refine my mesh and prism layers. I have two geometries, one is a small tube, and the other is a larger tube.

Small tube (running SST, but flow is laminar): What i am seeing is that, for the small particles on the order of 1 micron, the more I refine, the amount of deposition is reduced. It isn't converging, even with a very fine mesh. At 4 microns, things stabilize at the 1st refinement of mesh.

Larger tube (running SST, flow is turbulent): For 1 and 4 micron, particle deposition continues to decrease with mesh resolution, but at 6 micron it is stabilized through all meshes.


I ran a timestep and particle count sensitivity study, and both show that my results do not change with these factors.

What could be causing the continued decrease in particle deposition, without stabalizing. Am i going about this mesh sensitivity wrong? Or am i forgetting something?

Running in STAR-CCM+ 18.06.

Well, my opinion is that you should first start to assess a pure laminar flow as test-case. No models at all! Then, the model to simulate the particle dynamics depends on several physical factors, not only on the dimension. There are models from the simple lagrangian transport to the Maxey-Riley equations. Note that small grid size should be compared to the particle number in the volume. Are you considering shape and collision between particles.



Then, in case of turbulent flows, the small particle are very sensitive to the physical fluctuations. This should suggest that a DNS formulation should be used. Depending on the inertia, it could be possible to think a LES formulation.