Neighbor list overflow/ Atom types must start from 1

Submitted by Tamoghna on Wed, 01/30/2013 - 11:39

Hi!

I was trying a simulation with about 20000 particles of 0.003 m diameter. When I use the command for pour/legacy as:

fix ins1 nve_group pour/legacy 20000 2 10001 vol 0.4 100 diam 0.003 0.003 dens 2284 2284 region fac1
fix ins2 nve_group pour/legacy 20000 2 10001 vol 0.4 100 diam 0.003 0.003 dens 2284 2284 region fac2

I get error saying
Neighbor list overflow, boost neigh_modify one (neigh_gran.cpp)

When I use the command for pour/legacy as:

fix ins1 nve_group pour/legacy 20000 2 10001 vol 0.4 100 diam 0.003 0.003 dens 2284 2284 region fac1
fix ins2 nve_group pour/legacy 20000 2 10001 vol 0.4 100 diam 0.003 0.003 dens 2284 2284 region fac2

I get the error saying
Atom types must start from 1 for granular simulations

Could you kindly help me out with the reason??

Regards,
Tamoghna

AttachmentSize
Plain text icon output.test2_.txt450 bytes
Plain text icon output.test1_.txt678 bytes
Plain text icon in.test1_.txt1.7 KB
Plain text icon in.test2_.txt1.7 KB

Forums:

ckloss's picture

ckloss | Wed, 01/30/2013 - 17:48

Hi Tamoghna,

regarding issue 1: Neighbor list overflow, boost neigh_modify one means particles have too many neighbors. I suggest you simplify your simulation step-wise to track down the source of the issue.

regarding issue 2: the lowest atom type you use for inserting particles must be 1 . Have a look at the doc for fix pour/legacy for info how to change the atom type

Cheers
Christoph

Tamoghna | Thu, 01/31/2013 - 12:13

Thanks Christoph!!!

Issue 1 was due to the large skin distance I reduced it making it work fine :)

Issue 2 seems a bit strange as it makes parametric study a bit complicated (in case material number is a parameter!!!)

Regards,
Tamoghna

ckloss's picture

ckloss | Fri, 02/01/2013 - 12:34

>>Issue 2 seems a bit strange as it makes parametric study a bit complicated (in case material number is a parameter!!!)
No, this has nothing to do with material paramters. it's just a numbering convention. You can perfectly vary the material params

Cheers,
Christoph