[lammps-users] Problem about velocity profile

Dear Lammps users,
I want to simulate an electroosmotic flow with Nacl solutions between two parallel walls .when I calculate the velocity profile ,I find that the sodium and the chlorine have significant fluctuations in velocity profile while the water molecular not ,can anybody help me ? what wrong with my input script file ?

jiaoyixiong

Thanks very much ,and here is my input script file :

Initialization

units real
dimension 3
boundary p p f
atom_style full
pair_style lj/cut/coul/long 10
pair_modify mix arithmetic
bond_style harmonic
angle_style harmonic
kspace_style pppm 1e-4
kspace_modify slab 3.0

Atom definition

read_data data_Solution.txt

group flow type 1 2
group na type 3
group cl type 4
group boundary type 5
group eof union flow na cl
group lower id <> 6594 7745
group upper id <> 7746 8897

Settings

neighbor 2 bin
neigh_modify every 1 delay 0 check yes page 1000000 one 20000
velocity eof create 300 57287

pair_coeff 1 1 0.1553 3.166 10.0
pair_coeff 2 2 0 0 10.0
pair_coeff 3 3 0.0148 2.58 10.0
pair_coeff 4 4 0.106 4.45 10.0
pair_coeff 5 5 0.584 3.39 10.0
bond_coeff 1 600 1.0
bond_coeff 2 1200 1.0
angle_coeff 1 60 109.47

fix 1 all nve
fix 2 eof temp/rescale 10 300.0 300.0 1 1.0
fix 3 flow rigid molecule
fix 4 eof efield 0.055 0.0 0.0
fix 5 boundary temp/rescale 2 300 300 2 1.0
fix 6 boundary spring/self 2400
fix 7 boundary recenter NULL NULL INIT units box
compute 1 eof temp/partial 0 1 1
compute 2 boundary temp

timestep 1
thermo 1000
thermo_style custom step temp press pe etotal vol tpcpu spcpu
thermo_modify flush yes
thermo_modify temp 1
min_style cg
minimize 1.0e-4 1.0e-6 100 1000
restart 200000 nacl.restart
fix jiao1 na ave/spatial 2 10000 100000 z center 0.2 z vx density/mass units box ave one file na_distribution210.txt
fix jiao2 cl ave/spatial 2 10000 100000 z center 0.2 z vx density/mass units box file cl_distribution210.txt
fix jiao3 flow ave/spatial 2 10000 100000 z center 0.2 z vx density/mass units box file flow_distribution210.txt

dump 1 all atom 10000 vmd.lammpstrj

Run

run 30000000

Dear jiaoyixiong,

I would guess this is simply related to the fact that you have much
less Na and Cl ions than water molecules, therefore much less
statistics and much more velocity fluctuations. In this case, the only
ways to reduce fluctuations I can think of are to average your
profiles on longer times, or to simulate bigger systems, with a larger
number of ions.

Best,
Laurent