[lammps-users] About Poiseuille Flow's density and velocity distribution

Dear all,
I am new to LAMMPS ,when I begin to learn it ,I try to duplicate the work of others to understand the command ,
when I run the Sample from 《Numerical Simulation of Viscous Flow: A Study of Molecular Dynamics and Computational Fluid Dynamics 》by Jeremy Fried , I have some problems:
I use the fix ave/spatial command to get velocity and density information .but the velocity distribution looks higher and the density distribution looks lower than Jeremy’s work as we have the same input file. what wrong ?

Best regards,

here is the velocity distribution: Graph_ux.jpg
and the velocity distribution from Jeremy :Jeremy.jpg

the input file:

3D Poiseuille Flow

dimension 3
boundary p p p
atom_style atomic
neighbor 0.3 bin
neigh_modify delay 5

geometry

lattice fcc 0.45
region box block 0 10.0 -10.0 7.5 0 10.0 units box
create_box 2 box

create atomic lattices

mass 1 1.0
mass 2 1.0
region 1 block INF INF -10.0 -7.5 INF INF units box
region 20 block INF INF -10.0 -9.16667 INF INF units box
region 21 block INF INF -9.16667 -8.33333 INF INF units box
region 22 block INF INF -8.33333 -7.5 INF INF units box
region 2 intersect 2 1 box side out units box
create_atoms 1 region 2
lattice none
lattice fcc 0.85
create_atoms 2 region 1
group wall region 1
group first region 20
group second region 21
group third region 22
group flow subtract all wall

potential calculation

pair_style lj/cut 1.12246
pair_coeff * * 1.0 1.0 1.12246

initial velocities

compute mobile flow temp
compute stationary wall temp
fix 1 all nve

Poiseuille flow

minimize 1.0e-3 1.0e-5 100 1000
fix 2 flow addforce 0.1 0.0 0.0

wall

fix 7 wall temp/rescale 1 0.722 0.722 0.0001 1.0
fix 5 wall spring/self 57.1
fix 20 first recenter NULL INIT NULL
fix 21 second recenter NULL INIT NULL
fix 22 third recenter NULL INIT NULL

run

timestep 0.0001
thermo 1000
thermo_modify temp mobile

dump 1 flow custom 2500000 dump2.flow y vx
dump 2 all atom 50000 flow2.lammpstrj
fix ans flow ave/spatial 2 5000 5000000 y center 0.1 y vx density/number file fix2.dat

run 10000000

Graph_ux.jpg

Jeremy.jpg

Who knows? You're comparing LAMMPS to some
problem in a book - any line in your input script could
be different than what they did in the book. What is
reaonably certain is that LAMMPS is doing exactly
what you asked it to do. So you'll need to figure
out what is different.

Steve

2010/5/22 jiaoyixiong <[email protected]>: