calculate pressure

Dear users
I want calculate pressure of fluid in nano channel
my nano channel is simple cubic
i define three section at first,middle and end of channel in the flow direction
when i calculate pressure at three section of channel without any external force, magnitude of pressures are not equal
what is wrong?

dimension 3
boundary p p p
units real

atom_style atomic
neighbor 2 bin
neigh_modify delay 1

create geometry

lattice fcc 5.04 orient x 1 0 0
region box block 0 261.504 0 56.344 -24.516 24.516 units box
create_box 3 box
create_atoms 1 box

mass 1 39.94
mass 2 107.86
mass 3 39.94

LJ potentials

pair_style lj/cut 8.52
pair_coeff 1 1 .2372 3.41
pair_coeff 1 2 .39744 3.12
pair_coeff 1 3 .2372 3.41
pair_coeff 2 3 .39744 3.12
pair_coeff 2 2 8.5 3.12
pair_coeff 3 3 .2372 3.41
#count region
region pressureip block 0 3.33 8.172 48.172 INF INF units box
region pressureop block 50 53.33 8.172 48.172 INF INF units box
region pressureod block 258.174 261.504 8.172 48.172 INF INF units box
variable vol equal 6531.06

define groups

region res1 block 0 3.33 INF INF INF INF units box
region res2 block 17.33 20.66 INF INF INF INF units box
region res3 block 34.66 38 INF INF INF INF units box
region res4 block 52 55.33 INF INF INF INF units box
region res5 block 69.33 72.66 INF INF INF INF units box
region res6 block 86.66 90 INF INF INF INF units box
region res7 block 104 107.33 INF INF INF INF units box
region res8 block 121.33 124.66 INF INF INF INF units box
region res9 block 138.66 142 INF INF INF INF units box
region res10 block 156 159.33 INF INF INF INF units box
region res11 block 173.33 176.66 INF INF INF INF units box
region res12 block 190.66 194 INF INF INF INF units box
region res13 block 208 211.33 INF INF INF INF units box
region res14 block 225.33 228.66 INF INF INF INF units box
region res15 block 242.66 246 INF INF INF INF units box
region res16 block 258.174 261.504 INF INF INF INF units box
region lowers block INF INF 0 1 INF INF units box
group lowers region lowers
region uppers block INF INF 55.344 INF INF INF units box
group uppers region uppers
region pump block 0 50 8.172 48.172 INF INF units box
group pump region pump
region lowerd block INF INF 1 8.172 INF INF units box
group lowerd region lowerd
region upperd block INF INF 48.172 55.344 INF INF units box
group upperd region upperd
group walld union lowerd upperd
group walls union lowers uppers
group wall union walls walld
group flow subtract all wall
set group wall type 2
set group pump type 3

initial velocities

compute mytemp flow temp/com
compute mytemp2 flow temp
compute walldt walld temp
velocity flow create 100 482748 mom yes rot yes dist gaussian temp mytemp2
velocity walld create 100 482748 mom yes rot yes dist gaussian temp walldt
fix 1 all nve
fix 2 walld temp/rescale 20 100 100 5 .5
fix_modify 2 temp walldt

velocity walls set 0.0 0.0 0.0
fix 3 walls setforce 0.0 0.0 0.0
#fix 4 flow addforce .1 0 0 region pump
compute pressure1 flow stress/atom mytemp
compute pressure2 flow reduce/region pressureip sum c_pressure1[1] c_pressure1[2] c_pressure1[3]
compute pressure3 flow reduce/region pressureop sum c_pressure1[1] c_pressure1[2] c_pressure1[3]
compute pressure4 flow reduce/region pressureod sum c_pressure1[1] c_pressure1[2] c_pressure1[3]
variable pressure5 equal -(c_pressure2[1]+c_pressure2[2]+c_pressure2[3])/(3vol)
variable pressure6 equal -(c_pressure3[1]+c_pressure3[2]+c_pressure3[3])/(3
vol)
variable pressure7 equal -(c_pressure4[1]+c_pressure4[2]+c_pressure4[3])/(3*vol)

Run

timestep 2
thermo 100000
thermo_style custom step time c_mytemp c_mytemp2
#run 200000
dump 1 all atom 100000 re3.flow
fix 5 flow ave/spatial 1 190000 200000 y lower 3.5 vx norm sample ave running units box region res9 file vel1.perofile
fix 6 flow ave/time 1 190000 200000 v_pressure5 v_pressure6 v_pressure7 file pressure2.txt
run 20000000

This script is very complex, whereas your problem is either very simple, or maybe non-existent. I did not see anything obviously wrong, except that your LAMMPS programming style (variable naming conventions, protocols for defining regions, etc.) is very haphazard and therefore error-prone. One cause for concern is the very large size difference between the 3 types. If they are all initialized on an FCC lattice, you are going to have large force imbalances once you start the dynamics. This suggests to me that you have not spent enough time testing your basic physical model and instead you jumped straight in with a very complicated set up. I suggest you take a few steps backwards and do some testing with a much simpler script. Also, the three pressures (named 5, 6, and 7!) will never be exactly the same, due to thermal fluctuations.

Aidan