ESyS-Particle

Considerable particle overlap after deformation

Asked by Dian

Hi all,

I was doing a simple experiment that a 2D multi-layer stack is deformed via horizontal compression. The particles were significantly overlapped after finite amount of compression. I thought that all particles would be almost rigid (i.e., fixed shape, diameter, etc) in DEM modeling, although a very tiny amount of overlap is allowed to generate normal force. But in my experiment, the overlap can be observed clearly. I suspect that this problem is due to improper initialization of some parameters, such as density, youngs modulus, etc. Any comments on it? Thanks.

Best,
Dian

Question information

Language:
English Edit question
Status:
Answered
For:
ESyS-Particle Edit question
Assignee:
No assignee Edit question
Last query:
Last reply:
Revision history for this message
Dion Weatherley (d-weatherley) said : 		#1

Hi Dian,

Significant overlap of particles (I call it "pan-caking") is usually fixed by increasing the normal stiffness of particle-pair interactions. Depending on which interactions you are using, try increasing either normalK or youngsModulus. You may need to increase it by as much as 100x or more. Bear in mind that you may also need to decrease your timestep increment to maintain numerical stability. The timestep increment (dt) is proportional to 1/sqrt(K) where K is the normal stiffness. If you use youngsModulus (E), K = E.Rmax.

Cheers,

Dion

Revision history for this message
El Vierde (vaccari-andrea) said : 		#2

Hi Dion,

do I understand correctly that if I want to model steel spheres (E=200GPa) with maximum radius of 1 m, then I should set K to 200e+9 while, if the particle where 10cm then K=20e+9?

I've read the forum answers about simulation units and real units and I still have not quite grasped it.

I'd like to use esys to roughly simulate sinkholes with a very basic model:
- top cohesive layer bonded to sidewalls supported by
- bulk of free particles contained in a hooper-like mash with
- a small hole at the center of the bottom retaining wall from where particle flow out of the simulation after a short distance (BTW, can I actually let particle disappear? I set a distance of 2*maxRadius below the hole.)

I would really like, eventually to put cohesive clay-like properties to the top layer and sand-like property to the free-particle so that as the support is remove the top will bend first then crack and simulate a top layer collapse.

It's just a rough model but I'd like to start easy to understand how all the components of the simulation interact.

I've few issues that I'm trying to address:
- define the proper material properties in terms of simulation unit (I'll go back and read the forum entries again)
- tag particle after compactation under gravity. Even using small tolerances in gengeo, the resulting particle block is still porous and I wonder if there is a way to tag the top layer after the particles had settled and then modify the cohesion properties.
- I guess for a safe simulation time I should use 1/sqrt(Kmax) where Kmax is the largest of K in my setup.

And lot of other questions, but I'll dig the forum first...

Thanks for the great tool!!!

Cheers,
Andrea

Revision history for this message
El Vierde (vaccari-andrea) said : 		#3

I forgot to mention that I'm starting with a 2D simulation...

Can you help with this problem?

Provide an answer of your own, or ask Dian for more information if necessary.

To post a message you must log in.