DEM on MOOSE

243 views
Skip to first unread message

Jordan Cox

unread,
May 24, 2016, 6:18:34 PM5/24/16
to moose-users
Hey MOOSE users,

I am a new user to MOOSE but have some experience in numerical methods and computer simulations.  I'm working on my dissertation and I would like to do some simulation work in pebble bed reactors (PBRs) that involves MOOSE.   But also I need to look into discrete element modeling (DEM) simulations.  For example to model pebble flow data back to MOOSE for geometry and then MOOSE solving for the navier stokes equation etc.  

I understand the difference of DEM and FEM but, because I am not experienced with MOOSE I do not know if it would be possible to build a MOOSE animal to tackle DEM.  I am starting this thread so that if anyone knows if this would be productive or even possible within the MOOSE framework, or if anyone is working on this now.

Thanks!

Andrew....@csiro.au

unread,
May 24, 2016, 6:25:26 PM5/24/16
to moose...@googlegroups.com

We here at CSIRO are quite interested in this type of thing for a different reason.  We model rocks that break.  For quite some time we’ve been pondering whether to combine DEM with moose, but just haven’t had time to do anything.  i’ll be interested in hearing other people’s replies.

 

a

 

 

Ph: +61 7 3327 4497.  Fax: +61 7 3327 4666
Queensland Centre for Advanced Technologies
PO Box 883, Kenmore, Qld, 4069 
 

--
You received this message because you are subscribed to the Google Groups "moose-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to moose-users...@googlegroups.com.
Visit this group at https://groups.google.com/group/moose-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/moose-users/d22d0d4f-3ea2-46f9-999c-6cf2da83b157%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Benjamin Spencer

unread,
May 26, 2016, 1:01:39 AM5/26/16
to moose-users
A while back we developed a pretty incomplete prototype implementation of DEM in MOOSE.  We ran the sphere packing algorithm in another DEM code and loaded the bonds between particles as truss elements into MOOSE and computed the elastic response of the network of bonds.  We didn't do anything to handle contact interactions between the particles, which as I understand is really what DEM is all about. If you're just interested in it for rock fracturing, maybe you're more interested in modeling bond breaking than the flow of granular materials.

-Ben

Andrew....@csiro.au

unread,
May 26, 2016, 1:05:00 AM5/26/16
to moose...@googlegroups.com

Thanks Ben, that all sounds quite interesting.  I wonder why you did it?  Yes, i think your last sentence is correct for me,

 

a

 

 

Ph: +61 7 3327 4497.  Fax: +61 7 3327 4666
Queensland Centre for Advanced Technologies
PO Box 883, Kenmore, Qld, 4069 
 

Benjamin Spencer

unread,
May 26, 2016, 1:55:18 AM5/26/16
to moose-users
Andy,

Our motivation for that was that we were trying to replicate what we did in this paper using another code in MOOSE:

http://dx.doi.org/10.1016/j.nucengdes.2014.05.049

We definitely could have done it -- we just ran out of time. If you guys are interested in pursuing DEM or other related methods, I'd love to talk.

-Ben

co...@oregonstate.edu

unread,
May 26, 2016, 2:11:27 PM5/26/16
to moose-users
Hey Ben,

I am very interested in creating a DEM package with MOOSE.  I am looking into the paper you cited.  But I would like to use MOOSE to estimate graphite swelling in pebble bed reactors (PBRs).  The graphite swell rate is a PDE that will produce a volumetric change of the graphite that will then affect pebble packing.  I would like to be able to be able to couple several physics of the PBR together but the DEM method is really necessary for pebble flow as how the pebbles apply a force to one another and then move as a response I have never heard of anyone modelling using FEM.  

In short I am tremendously interested in working on this.

Jordan

co...@oregonstate.edu

unread,
May 26, 2016, 2:19:10 PM5/26/16
to moose-users
What DEM Program did you use?  Was it LAMMPS?

Jordan Cox

Benjamin Spencer

unread,
May 27, 2016, 4:15:46 PM5/27/16
to moose-users
Jordan,
  We used a research DEM code developed here at INL.  I think it's reasonable to implement the bond-breaking part of DEM in MOOSE, but I'm not so sure about the pebble flow part of the problem. DEM codes typically use explicit time integration, while we're mostly based on implicit time integration in MOOSE. I think we'd have some serious challenges solving that problem with an implicit solver.
  I'm happy to discuss options. Contact me off-list and we can talk.

-Ben

Reply all
Reply to author
Forward
0 new messages