Particles in lattice disappear through bottom wall

Submitted by rudranarayan on Thu, 08/04/2011 - 21:04

I have been running many simulations with a Lattice of granular particles with different CAD meshes interacting with them. There has been one major problem that i have not been able to figure out. When my timestep is any higher then 10^-7 gravity will pull the particles right through the bottom ZWall for no apparent reason. This results in the loss of all my particles after a short period of time. And they don't just fall out of the simulation, when they hit the bottom zwall the entire row of particles just disappears. Taking advice from other forum posts i tried adjusting my Youngs modules and coefficient of restitution to numerous different values, none of this helping what-so-ever.
I have many different input files but all produce the same results when the timstep is raised above 10^-7. This input has a time-step of 10^-6 and the lattice spacing is .01, my normal spacing is .002 but .01 is sufficient to show the problem i run into (with .002 the same problem occurs).

Thanks for the help!

Forums:

raguelmoon's picture

raguelmoon | Fri, 08/05/2011 - 02:14

Hi
I am wondering why you apply granular wall to CAD after inserting particles. I guess you apply this before run:
fix bucket_wall all wall/gran/hertz/history 1 0 mesh/gran 1 cad1

Maybe this will solve your problem....

Ram

Ram

rudranarayan | Fri, 08/05/2011 - 02:38

Ram,

Im not exactly sure what you mean. I use fix cad1 all mesh/gran one.stl 1 .05 -2.5 -.9 1.9 0. 90. 0. to initiate the cad mesh in my simulation and then fix bucket_wall all wall/gran/hertz/history 1 0 mesh/gran 1 cad1 to tell LIGGGHTS cad mesh is a granular wall that will interact with the particles.

Regardless, removing anything that has to do with CAD meshes still results in the same problem. There is something wrong with the lattice itself (or bounding walls), not the import/use of CAD meshes.

Thanks for the response!

ckloss's picture

ckloss | Mon, 08/08/2011 - 05:01

The reason for losing particles is that your simulation domain is fixed and spans from 0 to 1 in z-direction, whereas your z-wall is at -0.001, which is outside of the domain, so it will not hold the particles back properly.

Christoph

rudranarayan | Mon, 08/08/2011 - 18:02

Thanks for the response Christoph!

This is what i thought initially but if i change my simulation to as follows:

region lat block 0 .1 0 .6 0 .1 units box
region reg block -.1 .5 -.1 .8 -.1 .5 units box
create_box 1 reg #name defined region box

fix xwall all wall/gran/hertz/history 1 0 xplane -0.001 0.101 1
fix ywall all wall/gran/hertz/history 1 0 yplane -0.001 0.601 1
fix zwall all wall/gran/hertz/history 1 0 zplane 0.0 .49 1

When setting the simulation i get this and the simulation will not run:

run 1
Setting up run ...
Memory usage per processor = 622.712 Mbytes
Step Atoms KinEng 1 Volume
0 782901 0 0 0.324
1 782901 nan nan 0.324
Loop time of 18.0693 on 2 procs for 1 steps with 782901 atoms

And as said earlier, using the same simulation as posted in my original question and a timestep of 10^-7 I dont get this same problem.