simulating compression of particles

Submitted by kuhsteak on Wed, 01/09/2019 - 13:10

Hi everyone,

i am working on a liggghts simulation of a plastic extruder for my master thesis.
On my adventure with DEM i noticed that the soft particle approach wants to have only small overlaps between particles (< 0.5%d).
In my real life case the particles are getting deformed in their way through the extruder. Wich leads to some questions:

1)How does a Youngs Modulus exceeding the limit of 5e6 Pa change my simulation. (in older versions i had to add "hard-particels: yes")

2)How can i simulate the compression thorughout the extruder without getting big overlaps?
- i found the idea in the forum to allow a certain overlap withoug inducing a force. Like in the link below. Has anyone a better idea?
- https://www.cfdem.com/forums/manage-overlap-particles

3)Without changing the contact Model i get bigger overlaps due to the restricted Space and the huge forces in the extruder. Are those results physically accurate? Where is the problem with bigger overlaps?

Thanks in Advance

Regards,

Thomas

Forums:

arnom's picture

arnom | Mon, 01/14/2019 - 11:41

Just briefly, the assumptions in DEM breaks down if your overlaps become too large. In the premium version of LIGGGHTS we have a model that actually reduces the radius of compressed particles. This model actually works quite well as standard DEM only allows for elastic deformation and in an extruder you certainly have plastic deformations.

If your Youngs Modulus is high, then your time step size will decrease along with smaller particle overlaps and/or higher forces.

I don't think the results are physically accurate as the simulation does not consider plastic deformation as noted above.

Kind regards,
Arno

DCS team member & LIGGGHTS(R) core developer

kuhsteak | Sat, 01/19/2019 - 08:57

for the reply Arno. I will try myself on a new contact model.

Kind regards,
Thomas

kuhsteak | Fri, 04/26/2019 - 07:03

Hi Arno,

i now implemented a new Model which changes the effective Radius in effect to a certain force. I also vary the constant factors of
the Hertz model due to the force applied.

After all i am not sure if that is the thing i should do. Could you give me a little hint how you do it in the premium version.
Are you actually changing the radius or do you only change it in in the force calculation?
What is happening with the mass preservation.

I would highly appreciate a small push in the right direction.

Thanks in advance,
Thomas