Generating samples of loose granular packing

What is the best way to control the packing density without changing the characteristics of the grading curve? In other words, how can I obtain a looser column?
Here are some ways which I though may work:

  1. Remove some particles from previously settled column randomly and the number of the particles can be estimated from the ratio of the smaller packing density to the denser one with respect to the total number of particles. But using this way I even end up with a greater packing density as 0.86 than the previous one 0.84怂
  2. when assigning the suspended particles in rows and columns, give limited spacing in y direction.
  3. increase the friction angle only in the settling stage?
  4. Directly changing the grading curve. Like with bad grading where the minimum and maximum size of particles are more approaching the greater the void ratio will be.

Attached is the radius,x and y of the randomly particles-removed column and its ps.loose_50000_3rd.txt (1.2 MB)

Iā€™d try the following steps:

  1. Start with a ballistic deposition step, but instead of giving enough time (and potential head) between settling particles, keep the height of release as short as possible.

  2. Give a higher friction value for the particles, so they are more likely to be locked in.

  3. Delete 5 - 10% of the particles and let it reach equilibrium.

  4. Remove any surface particles, which are not part of the loose packing.

1 Like

Thank you Krishna! I will try now. Here is another issue, I want to increase the friction, there are mu, murf, mumg, mumb in our code. I only increase mu but it did not work. Could you please help me with this?

You can increase mu, which is the grain grain friction. murf is rolling and others are for interaction with the sidewalls, you can increase mum and mumb to increase side wall friction

Delaney, G. W. (2017). Structure Generation and Analysis from the Loosest to the Densest Random Packings. In EPJ Web of Conferences (Vol. 140, p. 06029). EDP Sciences.

is a good paper for generating loose granular samples