Particles sticky on the auger and not flowing

Hello,

I want to simulate powder auger dosing of particles with size around 600-800 µm.
The problem is that the particles are sticking to the auger and not flowing down. I changed a lot the parameters of the friction, cohesion and rolling friction and still didn't work. What could be the problem?

Attached are some screen-shots of the input script and the problem occuring.

Could you please suggest some tips for a solution?

Attachment	Size
2018-03-07_08h56_42.png	86.79 KB
2018-03-07_08h56_06.png	37.95 KB
2018-03-07_08h10_59.png	389.65 KB
2018-03-07_08h01_39.png	18.21 KB
2018-03-07_07h57_13.png	319.74 KB
2018-03-07_07h54_08.png	87.19 KB
2018-03-07_07h51_39.png	68.37 KB
auger_dosing_2_materials.txt	3.94 KB
simplified_schnecke_d5.stl_.txt	2.34 MB
simplified_schnecke_d5.exclude.txt	0 bytes
boden_d5_1.stl_.txt	399.3 KB
boden_d5_1.exclude.txt	0 bytes

Forums:

LIGGGHTS® - User Forum

richti83 | Wed, 03/07/2018 - 20:06

pictures of script not very helpfull

Please could you provide your full script as text file ? We need to see some more facts (unit system, timestep, integrator, order of commands)
(you can edit your first post and attach the file with .txt file ending)

bilalelkassem | Thu, 03/08/2018 - 08:59

The full script as a text

The full script as a text file is attached up.

richti83 | Thu, 03/08/2018 - 17:32

some suggestions

Hi,

I tried you input script (without meshes) and your timestep seems to high for the used stiffness
WARNING: time-step is 43.129313 % of rayleigh time (../fix_check_timestep_gran.cpp:200) WARNING: time-step is 12.635861 % of hertz time (../fix_check_timestep_gran.cpp:208)
You should not exceed 20% of rayleigh time. (lower youngs modulus or use smaller timestep)

I don't have experience with very small, cohesive particles, I guess your high cohesionEnergyDensity could be one reason for the sticky particles, does it work with cohesion off ?

I don't know your meshes, but maybe the element exclusion leads to holes of the mesh and than the particles get into the auger and can't leave them ? Try to create a "clean" mesh using step files and netgen mesher

The rotational velocity is very very high, maybe you'll need a smaller timestep (when all steps above doen't help)

A last assumption is to switch to cgs unit system (centimeter,gram,second) to avoid numerical errors (but as there is no warning about volume expencity I don't really think this is the reason).

Are you from Germany ? As your meshes have nice 'denglish' names :-).

Good Luck,
Christian

bilalelkassem | Thu, 03/08/2018 - 18:54

Hallo Christian,

Hallo Christian,
Thanks for your response.
Yes, I am in Germany, and you?

I want to simulate SpheroLac 100 powder. It is not cohesive but I guess it is not possible to leave it 100 % without cohesion. I am just trying to validate the simulation with the real world case.

My mesh are good. I uploaded them up. If you run the simulation, please take care of the names of the stl files where they are somehow changed during uploading!
Could you please explain it for me how to calculate the rayleigh time and get the percentage?

richti83 | Thu, 03/08/2018 - 20:50

Hi,

Hi,
I'm in Magdeburg (University Magdeburg “Otto-von-Guericke”, Department of Materials Handling, Institute of Logistics and Material Handling Systems (ILM))

When it's not cohesive, I would try without cohesion but with higher friction [1]

When your mesh is fine, why do you need exclusion list read ?

About raylight time see: https://www.cfdem.com/media/DEM/docu/fix_check_timestep_gran.html
You have this command in your script, but you 'unfixes' it after 1 run - why?

Viel Erfolg, Christian.

[1] Wensrich, C. M. ; Katterfeld, A.: Rolling friction as a technique for modelling particle shape in DEM. In: Powder Technology 217 (2012), S. 409-417

bilalelkassem | Fri, 03/09/2018 - 10:26

Hi Christian,

I writing my PhD in Bosch in Stuttgart. I come from RWTH Aachen University.

What does ''cohesion off'' mean in pair_style command?
My mesh seems good in LIGGGHTS but when running the simulation an error comes and recommends me to use the exclusion list read. You can open it and check please.

Could I please get your email address to contact you?

Danke, Bilal

I took out the unfix.

richti83 | Fri, 03/09/2018 - 12:14

Hi Bilal,

cohesion off means, that there are no attractive forces in contact direction (like cohesion/adhesion)

The error means, in fact, your mesh is not good ;-). I guess you exported it from CAD and used standard tesselation of CAD program which is not good for DEM (that's why the error is thrown). A good mesh hast no wall thickness and consists only of equilateral triangles.

I don't want to post my email here as many spam bots are on this site, but you'll find it in my research profile (https://forschung-sachsen-anhalt.de/pl/richter-97115)

Greets to Stuggi Town,
Christian.