yade-users team mailing list archive

[Janek Kozicki <question619109@xxxxxxxxxxxxxxxxxxxxx>]

Question #619109 on Yade changed:
https://answers.launchpad.net/yade/+question/619109

Description changed to:
HI,

I am reposting my question here, so that everyone can benefit from the
answers.

To calculate concrete the sample made of spheres (with cohesion) has
some of these spheres overlapping. Since it is difficult to generate a
sample where none of the spheres overlap. And with some overlapping the
spheres want to "explode" until there is no overlap. Sometimes cohesion
prevents this, but this means starting calculations with non-zero forces
between spheres.

To mitigate this problem, the initial overlap between spheres is stored
[*] (on first iteration) and is treated as "zero" penetration depth
during tension & compression.

[*] "is stored" means that this change in code is implemented in my
local copy of yade.

My question is:
1. does yade already have such feature, so that I don't need to port it back to latest version of yade?

So far Feng Chen has answered my question:  http://www.mail-archive.com
/yade-dev@xxxxxxxxxxxxxxxxxxx/msg12350.html

> Regarding your question, as far as I know packmol (
> http://www.ime.unicamp.br/~martinez/packmol/home.shtml) can handle initial
> packing of spheres (or even complicated sphere clusters) by maintaining the
> minimum distance between the sphere centers using a BOX-QUACAN
> optimization, thus if you make the minimum distance greater than
> max(r_i)*2, you should have zero overlap between the spheres.

And PACKMOL is an interesting solution to the problem. Actually a better
one.

But the question still remains: shall I implement/port this feature to
yade from my local yade copy?

best regards
Janek

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.