achieve certain packing factor / volumefraction_region

Submitted by Silias on Thu, 05/03/2012 - 20:46

Hi there,

I want to achieve a certain packing factor within a given region in my granular bulk material.

I tried to achieve a packing factor of 0.6 (Volume of solid material within the refered region / volume of refered region) within the region reg_ins by using this fix:

fix ins all insert/pack seed 100001 distributiontemplate pdd1 insert_every 1000 overlapcheck yes all_in yes volumefraction_region 0.6 region reg_ins ntry_mc 1000

What happens in the simulation is:
This region ist the lowest part of a cylindrical container. Which means after the first 1000 steps, the particles which have been inserted first, have been settled on the ground and new particles will be inserted until the desired volumefraction_region of 0.6 is reached...
Instead of inserting the new particles on top of the settled particles of the first insertion or in spaces between particles where a new particle would have enough room without touching other particles, the particles are inserted at locations where they have a big overlap with other particles, which makes the particle bed explode...

What do I wrong to reach my desired packing factor of 0,6 in the refered region?

Thanks a lot,
kind regards,

Sebastian

ckloss's picture

ckloss | Mon, 05/07/2012 - 09:15

Hi Sebastian,

hard to say what happens w/o knowing your input script in detail. It might be also related to a small bug in overlapcheck, to be fixed in 1.5.3. Actually to reach a packing density of 0.6, I would use particle growth, such as in the packing example (or turn overlapcheck to no and let the packing relax with fix nve/limit). The overlapcheck does not scale very well for high packing fractions

Cheers, Christoph

Maryam | Thu, 05/31/2012 - 01:13

Does this mean that we can have a packing of deformed (overlapped) particles?

Maryam

ckloss's picture

ckloss | Mon, 06/04/2012 - 18:29

Hi Maryam,

no "overlapped" does not mean deformed. In the frame of DEM, a certain (small) overlap is allowed but bodies are (typically) assumed to be rigid

Christoph