[lammps-users] about velocity command

Dear lammps-users and developers

The simulation box has three groups: low group, flow group and upper group as followed. The all group are in NVE ensemble.
The velocity command is used to make the upper group have the stream velocity v=2.0 along the x direction. However, The stream velocity of upper group decreased gradually.

step velocity(upper group)
3000 0.893662
4000 0.851187
5000 0.824726
6000 0.80896
7000 0.792684
8000 0.770081
9000 0.748362
10000 0.732631

IN script:

dimension 3
boundary p f p
units metal
atom_style atomic

create geometry

lattice fcc 5.8662
region box block -0.3 46.92 -1 68 -0.3 46.92 units box
create_box 2 box

mass 1 39.94
mass 2 195.1

potentials

pair_style hybrid eam lj/cut 10.215
pair_coeff 1 1 lj/cut 0.0104 3.405
pair_coeff 1 2 lj/cut 0.0208 2.940
pair_coeff 2 2 eam Pt_u3.eam

define groups

region 81 block INF INF 8 59 INF INF units box
create_atoms 1 region 81 units box

lattice fcc 3.9239
region 1 block INF INF 0 7.8 INF INF units box
create_atoms 2 region 1 units box
group lower region 1

region 2 block INF INF 59.2 67 INF INF units box
create_atoms 2 region 2 units box
group upper region 2

group boundary union lower upper
group flow subtract all boundary

initial velocities

compute temp_lower lower temp/com
compute temp_upper upper temp/com
compute temp_flow flow temp/com

velocity flow create 100.0 482748 units box temp temp_flow
velocity lower create 100.0 281375 units box temp temp_lower
velocity upper create 100.0 281375 units box temp temp_upper

fix 1 all nve

fix 2 flow temp/rescale 10 100.0 100.0 0.02 1.0
fix 3 lower temp/rescale 10 100.0 100.0 0.02 1.0
fix 4 upper temp/rescale 10 100.0 100.0 0.02 1.0

Run

thermo 1000
thermo_style custom step atoms temp pe ke etotal
dump 1 all xyz 1000 kapitza.xyz
timestep 0.001

Couette flow

velocity upper set 2.0 0.0 0.0 units box # add the velocity of 2.0 to upper group

fix 7 upper ave/spatial 1 1000 1000 y 59.2 7.8 vx vy vz units box norm sample file vel_upper.profile

fix 21 lower momentum 10 linear 1 1 1
fix 22 upper momentum 10 linear 0 1 1
fix 23 flow momentum 10 linear 0 1 1

run 6000000

If you are integrating the top group (with initial vel = 2.0)
with fix nve, then presumably they are influenced by the
other groups which slow them down over time. If you
want them just to flow at the intital vel, then set the
force on them to 0.

Steve

2010/6/5 Tom <[email protected]>: