Angle of repose

Hello,

I am trying to simulate the angle of repose test. On every case, particles roll over the lower surface : no cone is forming.
I use this parameter to modify my results without success :
- Coefficient of restitution particle–particle
- Static friction coefficient particle–particle
- Rolling friction coefficient particle–particle
- Coefficient of restitution particle–wall
-Static friction coefficient particle–wall
- Rolling friction coefficient particle–wall

Do I have to use cohesionEnergyDensity ?
Any help?

Rémy

Forums:

LIGGGHTS® - User Forum

richti83 | Tue, 02/20/2018 - 12:40

rolling friction

Did you enabled rolling friction for your particles and lower surface ?
pair_style gran model hertz tangential history cohesion off rolling_friction epsd2
fix walls all wall/gran model hertz tangential history rolling_friction epsd2 mesh n_meshes 1 meshes cad1

Do I have to use cohesionEnergyDensity ?

cohesionEnergyDensity is only needed when cohesion is NOT off (only for cohesive bulk material, but this material will never form a pile).

RemyB | Tue, 02/20/2018 - 15:44

- enabled rolling friction :

- enabled rolling friction :
I have done it but results still the same
- cohesionEnergyDensity
I use this because my material is very cohesive

I think this is the only way to form a pile. Right ?

richti83 | Wed, 02/21/2018 - 08:12

Could you explain your setup

Could you explain your setup in more detail? Which material properties have you tried allready (COR, friction, rolling friction, Youngs Modulus, Timestep, Geometry) ?

RemyB | Wed, 02/21/2018 - 11:32

Sure !

Sure !
My geometry is a cone with a lower diameter of 20mm located at 100mm from a disk.
This is my first case, so I use material property from littérature. There are values :
Young Modulus : 5e6
Poisson ratio : 0.45
Coefficient of restitution particle–particle 0.05
Static friction coefficient particle–particle 0.05
Rolling friction coefficient particle–particle 0.05
Coefficient of restitution particle–wall 0.45
Static friction coefficient particle–wall 0.65
Rolling friction coefficient particle–wall 0.1
The material has 2e-3m diameter, so I use a 1e-6 timestep.

I don't understand wich order is when you set parameters in the matrix.

richti83 | Wed, 02/21/2018 - 13:42

OK your friction is very very

OK your friction is very very low. As we use lower youngs modulus in DEM as in reality we have to calibrate the other parameters to get the same bulk solid behavior. [1][2]
I would suggest:
Y=1e7
Poisson 0.3
cor=0.2 (means 80% damping) for p-p and p-w contacts (it can be shown that the COR has less influence on AOR)
fric=0.3 .. 0.65
rolling friction 0.2 .. 0.7
CED (for cohesive bulk solid) 200000 to 600000

The order is for atomtypepair [1-1] [1-2] [2-1] [2-2] means p-p,p-w,p-w,w-w when 1=particle type, 2=wall type (obviously w-w doesn't exist but has to be filled in)

[1] C.M. Wensrich, A. Katterfeld: Rolling friction as a technique for modelling particle shape in DEM, Powder Technol., 217 (Feb. 2012), pp. 409-417
[2] T. Roessler, A. Katterfeld:Scaling of the angle of repose test and its influence on the calibration of DEM parameters using upscaled particles, Powder Technology
Volume 330, 1 May 2018, Pages 58-66

RemyB | Thu, 02/22/2018 - 09:22

Hi Richti,

Thank you for your advice.
I tried this on my case with your values but I had to reduce the CED. There is no flow from my insertion cone.

richti83 | Fri, 02/23/2018 - 08:46

Yes, sorry I forgot to

Yes, sorry I forgot to mention that CED depends on particle size distribution as overlaping area is inside the force formula.