[lammps-users] lost atoms when using table potential

LAMMPS LAMMPS Mailing List Mirror
1

Hi LAMMPS users,

When I used a table potential, some atoms are lost. From the output of thermo variables, I can’t see anything to cause it. Could anyone give me a suggestion? The following is part of output:

LAMMPS (22 Jan 2008)
Reading restart file …
orthogonal box = (-341.024 -327.738 -0.5) to (341.024 327.738 0.5)
2 by 2 by 1 processor grid
5500 atoms
500 atoms in group colloid
5000 atoms in group counterIon
Resetting global state of Fix 1 Style nvt from restart file info
Setting up run …
Memory usage per processor = 1.73258 Mbytes
Step Temp Press PotEng KinEng E_coul E_vdwl E_pair
155000 1 0.01064135 -0.32454778 0.99972727 -0.32438558 -0.00016220134 -0.32454778
155050 0.99999658 0.01064125 -0.32455643 0.99972386 -0.3243945 -0.00016193478 -0.32455643
155100 0.99999316 0.010641491 -0.32451056 0.99972044 -0.32434887 -0.00016169869 -0.32451056
155150 0.99998974 0.010641626 -0.32448292 0.99971701 -0.32432141 -0.00016151572 -0.32448292
155200 0.9999863 0.010641845 -0.32444333 0.99971358 -0.32428194 -0.00016139212 -0.32444333
…
…
183700 1.0010029 0.010657083 -0.32265364 1.0007299 -0.32252983 -0.00012380814 -0.32265364
183750 1.001016 0.010657055 -0.3226365 1.000743 -0.32251388 -0.00012261762 -0.3226365
183800 1.0010291 0.010657158 -0.32260134 1.0007561 -0.32247972 -0.00012162175 -0.32260134
ERROR: Lost atoms: original 5500 current 5498

my input script is:

2-d Colloidal anions with small counterions.

#initialization section

units lj #default
dimension 2
#processors 1
boundary p p p
atom_style charge

atom definition section by reading data files

##pair_style lj/cut/coul/cut 2.5 17.5
read_restart restart.155000.NPT
##read_data restart.data

LJ potentials #epslong sigma cutoff

pair_style hybrid/overlay coul/cut 100.5 table linear 1000

pair_coeff 1 1 coul/cut 100.5
pair_coeff 1 1 table U_col_col U_col_col
pair_coeff 2 2 coul/cut 10
pair_coeff 2 2 table U_ion_ion U_ion_ion
pair_coeff 1 2 coul/cut 60.247
pair_coeff 1 2 table U_col_ion U_col_ion
pair_modify shift yes

kspace_style pppm 0.0005

kspace_modify slab 3.0

#setting simulation parameters

neighbor 1.5 multi
neigh_modify check yes

define groups

group colloid type 1
group counterIon type 2

initial velocities

#compute colloid_temp colloid temp
#compute ion_temp counterIon temp
velocity all create 1.0 482748 #generate temperature by random seed
#fix 1 all nve
fix 1 all nvt 1.0 1.0 0.3 ###aniso 0.01 1.0 0.01 1.0 NULL NULL 1.2
fix 2 all enforce2d
###fix 3 all temp/rescale 100 1.0 1.0 0.05 1.0

Run

timestep 0.00003
thermo 50
thermo_style custom step temp press pe ke ecoul evdwl epair ### lx ly lz vol

dump 1 all atom 50 npt.lammpstrj
dump 2 all atom 10000 snapshots..lammpstrj
restart 10000 restart..NVT
log log.npt_cut
run 100000
write_restart restart.afterNVT.*

Have a nice day!

Dongsheng

2

Lots of things could go wrong in this simulation, since you use many features.
I'd try to isolate the problem by turning things off one at a time. E.g I'd
run on 1 proc, I'd print thermo every step (to see if it blows up right before
the lost atom), I'd try to run with a non-tabled potential, etc.

Steve

3

Hi Steve,

Thank you for your suggestion. I had tried LJ potential. It was fine.
I just try to output thermo variables every step. It looks like as following:

Step Temp Press PotEng KinEng E_coul E_vdwl E_pair
183842 1.0010402 0.010657154 -0.32258686 1.0007672 -0.32246628 -0.00012057966 -0.32258686
183843 1.0010405 0.010657153 -0.32258652 1.0007675 -0.32246596 -0.00012055478 -0.32258652
183844 1.0010407 0.010657153 -0.32258617 1.0007677 -0.32246564 -0.00012052991 -0.32258617
183845 1.2781292e+35 1.5719812e+33 4.9429525e+16 1.2777806e+35 -0.32246532 4.9429525e+16 4.9429525e+16
ERROR: Lost atoms: original 5500 current 5498

I have tried to run the program by one processor. I got the exactly the same thing. The second last step looks fine from the thermo output. Do you have any idea what is wrong?

All the best!

Dongsheng