Inserting particles at random positions

Submitted by mkoester on Tue, 10/19/2010 - 10:03

Hi everybody,

I am interested in a simulation of a big sphere containing a many smaller particles. I tried this simulation by adding particles of a defined radius on a regular lattice. The problem is that this reguar structure works a a kind of crystalline system with slip planes when I try to deform it.

However I'd like to try another approach by adding the particles without a lattice with the "pour command". Is there a way to use this command with a sphere as the input region (and not a cube or a cylinder)?

regards,
Martin Koester

Forums:

ckloss's picture

ckloss | Tue, 10/19/2010 - 10:26

Hi Martin,

>>"pour command". Is there a way to use this command with a sphere as the input region (and not a cube or a cylinder)?

There are current works to improve fix pour/dev to be able to use arbitrary regions - this will be released at some point.

What you can do for now is use the "regionexempt" keyword in fix pour/dev. You can use a box that encloses the sphere as fix pour/dev insertion region and then subtract your sphere from this box region. If you then insert this into the regionexempt, that should do what you want

Christoph