contact model for overlappped part of a clump

Asked by Guo, Chang

Assuming such a situation, we build a clump with overlapped particles and get parameters between particles like kn and ks based on the linear contact model. If a contact occurs at an overlapped zone, this contact seems to be stiffer, and this stiffness is proportional to the number of overlapped particles. Is it right? How to solve this problem?

Question information

English Edit question
Yade Edit question
No assignee Edit question
Solved by:
Guo, Chang
Last query:
Last reply:
Revision history for this message
Karol Brzezinski (kbrzezinski) said :

Hi Guo,

Yade is really well-documented. I recommend reading the Guided Tour, DEM Formulation, and User's Manual. There is a section about clumps as well [1].

You can find there that "In Yade (class Clump), clump members behave as stand-alone particles during simulation for purposes of collision detection and contact resolution, except that they have no contacts created among themselves within one clump. "

Hence, there are no interactions between the clump members. Clump members' arrangement is rigid.



Revision history for this message
Guo, Chang (guopolyu) said (last edit ):

Dear Karol,

Thanks for your reply.

I didn't express myself clearly.
Assuming such a situation, there is a single sphere, A, and a clump, B, with two spheres, C and D. C and D have an overlapped zone. Normal stiffness between each pair of particles is kn . There is a contact between A and B and the penetration depth is delta. When this contact isn't at the overlapped zone, the normal force should be delta*kn. However, if this contact is at the overlapped zone, the normal force would be calculated twice, that is, 2*delta*kn. It looks unphysical.

Best Regards,

Revision history for this message
Vasileios Angelidakis (vsangelidakis) said :

Hi Guo,

What you describe is one of the limitations of the multisphere modelling approach. Indeed, this technique introduces an artificial amount of contact stiffness anisotropy at the particle level.

Thankfully, the linear model of yade considers the stiffness to be proportional to particle radius, and so smaller subspheres within your clump should have smaller effect in terms of additional force. If this issue is critical in your application, you can either 1. look into different clump generation techniques, where you can control the amount of overlap introduced or 2. change to a different modelling technique for non-spherical particles.

Hope this helps,

Revision history for this message
Guo, Chang (guopolyu) said :

OK. Thank you very much.