Particles can not stop fluctuate on the bed surface of a rotating drum

MFiX
#1

Hi guys,
I am trying to use MFIX-DEM for modeling a simple rotating drum. Most of my results, like particle velocity and angle of repose, look good, however, the particles on bed surface show different behaviors compared with other researchers’ work. Specifically, particles in Fig. 1 (mine) cannot stop fluctuate on the bed surface, but the phenomena are supposed to be like Fig. 2 (from literature). I have tried some adjustment, but the results are still not good.

  1. I have tried to adjust all parameters, only with very small Young’s modulus or restitution coefficient equal to 0.1 can make particles move less. While these parameters should not be so low.
  2. Smaller time step shows almost no improvement.
  3. With or without rolling torque, no help.
  4. Improve the quality of the geometry, no help.
  5. Settle down the particles for a longer time then rotate the drum, no help. After the drum rotated for 20 revolutions, particles still look like Fig.1.
  6. Change the contact model in MFIX, same phenomena.
    I have no idea what else can I do. Any suggestions would be very appreciated.
    image fig.1
    image fig. 2
#2

Do you have some experimental results to compare instead of comparing other simulation results of one instantaneous frame? And from fig 2, it seems the drum is rotating at 40 rpm, which is not a small rotation speed. If you look at the the paper of 2D DEM simulation of particle mixing in rotating drum: A parametric study, in figure 4, it shows at 40 rpm, the particles are not in the rolling regime, instead of cascading regime. So what are the properties of your particles, size, density etc, what is the geometry, cylindrical or spherical tumbler? 2D or 3D simulations? Please share more details of your simulation settings!

#3

Hi, Yupeng, thank you for your reply. I am trying to use the experimental results from the work of Parker el at.( D.J. Parker, A.E. Dijkstra, T.W. Martin, J.P.K. Seville, Positron emission particle tracking studies of spherical particle motion in rotating drums, Chem. Eng. Sci. 52 (1997) 2011–2022. doi:10.1016/S0009-2509(97)00030-4.) to validate my model. The Fig.2 is a simulation result from a literature in which the researcher successfully validated their results with Parker’s experiment. In my work, the 3D cylindrical drum has a diameter of 100 mm and length of 16 mm with periodical boundary condition, inside which the particle has density of 2500 kg/m3 and diameter of 3 mm. So the drum is in rolling regime. I have also did small rotational speed, but the particles are still fluctuate.

#4

Can you share your case settings?

#6

I am wondering why did you set DES_ET_INPUT=0.0 and DES_ET_WALL_INPUT=0.0?

#7

I showed you only one case setting. Actually, whatever the tangential restitution coefficient is, the particles are always fluctuating.

#8

Hi Yupeng,
Apologize for bothering you again. But if you are interested, I want to tell you that the reason for this problem is that the grid size is less than 2particle diameterRLM_factor when cell-linked (or grid-based) particle neighbor search algorithm is used. I just learned this from MFIX code. Anyway, thanks for the kind help. Cheers.

1 Like
#9

Hi，wenjie
I’d like to ask you some questions about rotating drum.
Does this model need to be built in MFIX itself？The official case has spinning drums, but it is limited.
Bother you！