Particle failure capability

Submitted by andybond13 on Thu, 11/07/2013 - 20:24

Is there a LIGGGHTS capability to take a particle under given conditions and split it / replace it with several smaller particles? I'm specifically looking for a solution within the "granular" atom-style world (but it could be a hybrid style).

I could have sworn I saw an example of this a few months ago, but I can't find it now. Was this the multisphere capability? (node/20 illustrates multiple particles bonded together, but I want one particle that splits into multiple independent particles).

Thanks!

richti83's picture

richti83 | Fri, 11/08/2013 - 13:50

Hi,
do you mean something like this:
http://ubuntuone.com/1P3F8zsdIoUq0P6YMVGoP6

unfurtunately the fix breakparticle/force was not converted to Liggghts 2.

However, there are some restrictions in 2.3.8 (see below) which will be relaxed over time:
+ fix breakparticle/force is not yet available

Chris: is there a plan to make this available in Liggghts 3 ?
In the mean time you can download Liggghts 1.5.3, there is an example (I made the video above with this example script)

I'm not an associate of DCS GmbH and not a core developer of LIGGGHTS®
but I have 7 years user experience using LIGGGHTS®, ParaView and LINUX

andybond13 | Tue, 11/12/2013 - 19:53

When you say this example script, was that a script given in the 1.5.3 distribution examples directory?

andybond13 | Tue, 11/12/2013 - 17:44

aha - breakparticle/force ! Yes, that's the capability I was remembering.

What's the approximate time-frame for the re-release of this feature (weeks/months/years)? Until then, I'll try using the Ligghts 1.5.3 version.

Thanks!

jtvanlew | Fri, 11/15/2013 - 00:49

I have been working on the same thing. And since I'm a terrible programmer I had to avoid c++, I used the python wrapper for lammps/liggghts. Here's an example of a compression I did to check my code:

http://www.youtube.com/watch?v=kBQb5AIC2Lo

I can talk more about it if that approach is intriguing to you.

jon

ckloss's picture

ckloss | Fri, 11/22/2013 - 16:58

Hi all,

>>I could have sworn I saw an example of this a few months ago, but I can't find it now.
yes, it's like Christian described:


However, there are some restrictions in 2.3.8 (see below) which will be relaxed over time:
+ fix breakparticle/force is not yet available

>>Chris: is there a plan to make this available in Liggghts 3 ?
Actually not, at the moment we have other (more pressing) needs to
unify the famework for pair and wall styles.

But if someone is willing to collaborate on the breakage capability,
we can at least update the last status and develop some honest physics on top of that
"showcase"

@jon: cool video!!

Cheers
Christoph

andybond13 | Tue, 12/31/2013 - 00:19

Thanks, I'll have to explore this! A few questions off the bat:
-have you documented it?
-can stress be the criterion rather than force?
-can you provide guidelines for the fragment file?
-will this work in a large-scale/MPI environment?
Thanks! I look forward to playing with it.

rahulsoni | Tue, 01/21/2014 - 07:38

Sorry for delayed reply

-have you documented it?
Not yet
-can stress be the criterion rather than force?
I think yes. Just go to fix_breakparticle_force.cpp then a loop their changes the flag for particles that are going to be break based on force acting on it. Criterion can be changed to stress. Required properties can be called from atom-> (like other properties are being called in same loop). But I am not sure about the mathematical way what model will be used to calculate the stress within the available options.
-can you provide guidelines for the fragment file?
fragment file is available within the example of liggghts 1.5.3 otherwise download at https://www.dropbox.com/s/fblb0e7wypm7gy5/breakage.tar.gz
Other fragments file can be downloaded from the links provided in the fragment file itself
-will this work in a large-scale/MPI environment?
I think it should because rest all the environment is of latest verison of liggghts only.

--
Regards

Rahul Kumar Soni
Scientist, CSIR - IMMT, India