Methods to generate a homogenous dense packing in EDEM

Yang Cui

Hi all,

I have some questions about generating a homogenous dense packing, where the solid concentration is very high (over 65%) thus the overlaps between particles are hard to avoid.  My simulation contains two different size of particles and the size ratio is about 5. When I try to generate such dense particle packing with "volume packing" and change the value of "Remove particle overlap" to a relative high value. I found the number of smaller particles is much lower than my expectation. I guess some small particles are generated inside large particles as the tolerance of particle overlap is actually larger than the diameter of small particles.  I haven't found an efficient way to generate such dense(volume fraction over jamming point), binary(size ratio are large), homogenous packing. 

Anyone can give some possible solutions for this case?

 

Best regards,

Yang

Ignacio Diez Alonso_21259

Hello Yang, 

May I ask, why are you setting the "Remove particle overlap" to a relative high value? As you say this setting is removing the small particles since the % used when checking if the overlap is too high is based on the larger particle radius making the contact. Reducing this value should help increase the number of small particles. Is there are reason why you would not want to use a lower "Remove particle overlap value"?

Hope that this helps.

Kind regards

Ignacio

Yang Cui

Ignacio Diez Alonso_21259 said:

Hello Yang, 

May I ask, why are you setting the "Remove particle overlap" to a relative high value? As you say this setting is removing the small particles since the % used when checking if the overlap is too high is based on the larger particle radius making the contact. Reducing this value should help increase the number of small particles. Is there are reason why you would not want to use a lower "Remove particle overlap value"?

Hope that this helps.

Kind regards

Ignacio

Sorry for my late reply.

The packing I want is a very high volume fraction, around 80%, for a binary system, which is far beyond the jamming point. In this case, overlaps between particles cannot be neglected. I tried to set the "remove particle overlap" to be 1% or 5%, but I found many particles are deleted in the generated packing from "VOLUME PACKING" function.

I also tried to start from a relative loose packing fraction and then compressed box (here I compress the box by making the top wall moving down to a specific value). In this case, because of the gravity and the size difference between particles, the homogenity was not very good along the Z axis. 

I found lammps provides some protocal like the energy minimization to achieve this. Not quit sure if there are other methods to generate the particle packing.