Fix recenter command in cascade simulation

Hello, all.

I tried to make cascade simulation which contain several cascade in 1 simulation using loop.

Because the defect calculated by OVITO become too many as cascade increase, I guess this is come from system move due to cascade.

So I tried to use fix rencenter command before and after cascade to fix the system and compare defect.

But when I tried this, too many defects are derived even when comparing first cascade dump file.

Do you have any idea why this problem happen?

#-------------------------- Equlibration -----------------------------
reset_timestep  0
timestep 0.001

velocity all create ${TempEql} 4928456 rot yes dist gaussian

fix a all npt temp ${TempEql}  ${TempEql} 100 iso 0.1 0.1 100

thermo 1000
thermo_style custom step temp etotal press vol density

run 50000

dump           dumpNPT all custom 1000 dumpNPT*.data id type x y z
velocity       all create ${TempEql} 491307 dist gaussian

fix            1 all nvt temp ${TempEql} ${TempEql} 0.1                                 #Relax at 300K

thermo       1000
thermo_style custom step dt time vol lx ly lz press pxx pyy pzz pe ke etotal temp

minimize   1e-8 1e-8  10000 10000

run            ${NstpEql}
undump         dumpNPT

unfix     1

fix          center all recenter INIT INIT INIT units box
run 0

dump dumpinit all custom 1000 dumpPKA.0 id type x y z
reset_timestep 0
timestep 0.001

run 0

unfix center
undump dumpinit

This is script before cascade

#  ------------------------ PKA SIMULATION ----------------------------
velocity     PKA   set ${pkaVx} ${pkaVy} ${pkaVz} units box

fix          myNVE all nve
fix          adtdt all dt/reset 1 1.0e-6 1.0e-3 0.01 units lattice
fix          shell outeratom temp/rescale 1 ${TempEql} ${TempEql} 2 1

compute      interT interatom temp
compute      outerT outeratom temp

thermo        1000
thermo_style  custom step dt time pe ke etotal temp c_interT c_outerT
thermo_modify line yaml format none

#dump         DumpPKA all cfg 1000 DumpPKA${i}_*.cfg mass type xs ys zs id
#dump_modify  DumpPKA element Ni Co Cr

run 40000

displace_atoms all move ${rshiftx} ${rshifty} ${rshiftz} units box

fix          center all recenter INIT INIT INIT units box
run 0

dump           dumpPKA all custom 1000 dumpPKA.${i} id type x y z
run 0

unfix        myNVE
unfix        adtdt
unfix        shell
unfix        center
undump       dumpPKA

write_restart restart.PKA_shot${i}

This is script for cascade simulation.

I compared dump file dumpPKA.0 and dumpPKA.1 for deriving defects.

Research is not a place for guesses. Either you can properly quantify it or not.

It sounds like you have not understood what it is that fix recenter does (it does not immobilize atoms, it is more like changing the point of reference for a trajectory from the coordinate system origin to the center of mass of the fix group). As such it will not stop any motion of your system, just subtract it. That is not really what you want in a cascade simulation scenario.

In case you can quantify a motion of the stationary part of your system, you can simply immobilize it by immobilizing a small subset of its model (e.g. the lowest layer.).

Generally speaking, the problem seems to be more a problem of how you analyze your data, set up your simulations, and your expectations rather than something that can be “fixed” on the LAMMPS side. That would make this an issue of your research and not a LAMMPS issue and thus off-topic for discussions about LAMMPS.

Okay. I will try to solve this problme… Thank you for your advice.