fixed ground

Submitted by andrea.pasquali on Tue, 02/08/2011 - 17:27

Hi,
I'm new with LIGGGHTS. I tried the tutorials and they are very good but I found only granulars that move from an inlet to grond (or to outlet).
How can I set instead fixed granular on a plane like a ground?

Thanks for any help

Andrea

Forums:

ckloss's picture

ckloss | Tue, 02/08/2011 - 21:00

Hi Andrea,

I am not 100% sure if I understood the questions right - but if you want to define a wall by frozen particles, see here:

http://lammps.sandia.gov/bench/in.granular.txt
http://lammps.sandia.gov/bench/bench_granular.tar.gz

Note that these are LAMMPS scripts, so you have to replace the pair_gran and fix wall/gran commands by their LIGGGHTS counterparts (as in the LIGGGHTS tutorials)

Cheers,
Christoph

andrea.pasquali | Wed, 02/09/2011 - 10:44

Hi Christoph,
Thank you for your reply!
I'd like to change the "movingMeshGran" starting from fixed particles on the ground (not inserting them).

These are the lines of tutorial I think I have to change (I changed also some movements of bucket); I wrote:

#particle insertion
#region bc block -0.9 0.9 -0.4 0.4 -0.02 -0.0 units box
region bc block -0.9 0.9 -0.4 0.4 -0.5 -0.3 units box
group nve_group region reg
#fix ins nve_group pour/legacy 1500 1 1 vol 0.7 1000 diam 0.08 0.08 dens 2500 2500 vel 0. 0. 0. 0. -0.8 region bc
fix ins nve_group pour/legacy 1500 1 1 vol 0.7 1000 diam 0.08 0.08 dens 2500 2500 region bc
fix fre nve_group freeze
#run with particle insertion
#run 10000 upto
unfix ins
#run to let particle settle
#run 29999 upto
#use the imported mesh as granular wall
fix bucket_wall all wall/gran/hooke/history 1 0 mesh/gran 1 cad1
#run 30000 upto
#moving mesh
fix movecad1 all move/mesh/gran linear -0.5 0. -0.3 cad1 1
run 15000
unfix movecad1
#now drag the particles up
fix movecad2 all move/mesh/gran linear 1.5 0. 0. cad1 1
run 25000
unfix movecad2

I'd like to introduce particles (2 or more layers of particles) on the bottom wall and they have to be fixed, not dropping by gravity.
But when I see the result the particles overcome the bottom wall and it seems they work without gravity!

What am I wronging?

Thanks

Andrea

andrea.pasquali | Wed, 02/09/2011 - 12:42

Hi Christoph,
I wrote:

lattice custom 0.01 a1 1 0 0 a2 0 1 0 a3 0 0 1 basis 0.5 0.5 0.5
region reg block -1 1 -.5 .5 -.5 .5 units box
create_box 1 reg
region bc block -0.9 0.9 -.4 .4 -.4 -.2 units box
create_atoms 1 region bc
#create_atoms 1 box
neighbor 0.05 bin
neigh_modify delay 0
group nve_group region reg

But when I run I got:

Lattice spacing in x,y,z = 0.01 0.01 0.01
Created orthogonal box = (-1 -0.5 -0.5) to (1 0.5 0.5)
2 by 1 by 1 processor grid
Created 288000 atoms
288000 atoms in group nve_group
Importing STL file 'bucket.stl'
Solid body detected in STL file
End of solid body detected in STL file.
WARNING: Imported mesh contains skewed triangles with a size ratio larger 10
Mesh calculations running. This may take a while...finished!
Import of 28 triangles completed successfully!
Setting up run ...
ERROR on proc 0: Neighbor list overflow, boost neigh_modify one or page

What's it means?

Thanks

Andrea

ckloss's picture

ckloss | Wed, 02/09/2011 - 16:05

>>ERROR on proc 0: Neighbor list overflow, boost neigh_modify one or page
That means that there are too many neighbors. You should use the "set" command to set the particle diameter to some useful value.

Best,
Christoph

andrea.pasquali | Wed, 02/09/2011 - 16:07

Hi Christoph,
Yes I set the diameter and now is ok.
I posted a new question before your reply!

Thanks

Andrea

andrea.pasquali | Wed, 02/09/2011 - 16:06

Ok I solved my problems,
Another question: How can export the results (all) in vtk by pizza?
I know this procedure:
run pizza.py
d = dump("dump.case")
v = vtk(d)
v.many("case")

In my .vtk files I have only atom_type field. I'd like also to have velocity etc...

Thanks for any help

Andrea