Settling sphere in box

Submitted by jpola on Thu, 07/28/2016 - 13:10

Hello,

I'm would like to share the case I was working on for several weeks. It is a simple case where the particle is released from some height in a box filled with fluid (settling_sphere.png). I was trying to compare CFD-DEM results with the one presented in:

"Simulation of Sedimentation of a Sphere in a Viscous Fluid Using the Lattice Boltzmann Method Combined with the Smoothed Profile Method" Suresh Alapati, Woo Seong Che and Yong Kweon Suh. Advances in Mechanical Engineering Article ID 794198

http://ade.sagepub.com/content/7/2/794198.abstract

The case is created to be working on 2 threads.
Mesh is quite small (40 x 40 x 64) but it can be easily changed in block mesh and I'm not using local dynamic mesh refinement.

Tests for Re = 31.9 and 11.6 are presented in the velocity_proflie.png. I've used a larger mesh there (60 x 60 x 120) to generate this chart.

What bothers me the most are the oscillations which are presented in the oscilations.png file.
Does anyone knows what is the source of those oscilations? As far as I know the the magnitude depends on the mesh size.

To find the source of this issue I looked at the relation between the particle center compared to mesh cell centres and mesh faces. I wanted to see what is the releation velocity profile and the position of the particle center on the mesh. Some fragment of that is presented on the picture "Velocity vs. Cells".
In this chart I plotted cell centres and cell edges of the mesh. The blue line presents the position of the sphere centre projected on the mesh. The red dots shows the points when mesh is at the cell centre. Similarly the second plot presents the velocity of the particle. The green dots presents the moments when sphere is at the cell centre. Recently I was thinking also that they are some parasitic currents which may show up in VOF kind of simulations.

What do you think about that? Does anyone of you also seeing such phenomena?

Regards,
Jakub.

Zheng Hu | Fri, 11/03/2017 - 18:51

Dear Jakub, Thanks for you sharing. I tried to run you example while met an error with "Residual data for U must be specified as a dictionary". Can you run this code directly without any changes? Should I change something so that the code can run correctly? Many thanks.