crazy particle

Submitted by hansmannb on Fri, 03/21/2014 - 10:39

Hej guys,

does anyone know which parameter cause the madness of one particle which kicks all the others out of the world?
This problem appears when i scale down the radius of the particles.

this is my input script:

>>>
units si
newton off
boundary f f f

atom_style granular
atom_modify map array

region world block -10 10 -10 10 -10 10 units box
create_box 1 world

neighbor 0.001 bin
neigh_modify delay 0

communicate single vel yes

variable emod equal 5e+06
variable quer equal 0.45
variable stoss equal 0.7
variable reib equal 100
variable dichte equal 400
variable radius equal 0.04
fix m1 all property/global youngsModulus peratomtype ${emod}
fix m2 all property/global poissonsRatio peratomtype ${quer}
fix m3 all property/global coefficientRestitution peratomtypepair 1 ${stoss}
fix m4 all property/global coefficientFriction peratomtypepair 1 ${reib}

pair_style gran/hertz/history
pair_coeff * *

timestep 0.001
variable tvortimestep equal 1000
variable t equal step*dt

fix grav all gravity 9.81 vector 0 0 -1

variable insertx_minus equal -5.5
variable insertx_plus equal -3.5
variable inserty_minus equal -0.5
variable inserty_plus equal 0.5
variable insertz_minus equal -2
variable insertz_plus equal 1

fix pts all particletemplate/sphere 1 density constant ${dichte} radius constant ${radius}
fix pdd all particledistribution/discrete 1 1 pts 1.0
region factory block ${insertx_minus} ${insertx_plus} ${inserty_minus} ${inserty_plus} ${insertz_minus} ${insertz_plus} side in units box
fix ins all insert/pack seed 100001 distributiontemplate pdd vel constant 0. 0. 0. &
insert_every once overlapcheck yes all_in yes volumefraction_region 0.8 region factory

fix wall1 all wall/gran/hertz/history primitive type 1 xplane ${insertx_minus}
fix wall2 all wall/gran/hertz/history primitive type 1 xplane ${insertx_plus}

fix wall3 all wall/gran/hertz/history primitive type 1 yplane ${inserty_minus}
fix wall4 all wall/gran/hertz/history primitive type 1 yplane ${inserty_plus}

fix wall5 all wall/gran/hertz/history primitive type 1 zplane ${insertz_minus}
#fix wall6 all wall/gran/hertz/history primitive type 1 zplane 5

fix interg all nve/sphere

dump dmp all custom 100 post/dump*.chute id type type x y z ix iy iz vx vy vz fx fy fz omegax omegay omegaz radius

run 900

write_restart restart/restart.*
<<<

nasato | Fri, 03/21/2014 - 16:48

Hi,

I guess your problem is related to your time step. Remember you are integrating explicit in time, so watch the time step.
Use always time steps smaller than 20% of Rayleigh time step, otherwise you will have this type of behavior.
Your Rayleigh time step in this setup is 0.0021792 s. You should use maximum 20% of it - 0.00043584 s
You can also use fix check/timestep/gran in your input script to know your current Rayleigh time step. Check the manual for details of this command.
Cheers,

Nasato