attempting to rescale 0.0. temperature

LAMMPS LAMMPS Mailing List Mirror
1

Hi All,

Below one is my input file and i am getting this error " attempting to rescale 0.0. temperature" .I introduced an Ar atom on top of a initially heated Ni surface. Right now, i am trying to calculate the velocity of Argon at 300 K from Maxwell-Boltzmann Distribution and below one is my input file. I would realy appreciate if someone tell me what is wrong with the input file.
Regards,
Rose
.

#Initialization

units metal
boundary p p p
atom_style atomic

#Atom Definition

read_data data.NiAr

#Force Field
pair_style hybrid eam lj/cut 2.5 lj/cut 2.5
pair_coeff 1 1 eam Ni_u3.eam
pair_coeff 2 2 lj/cut 1 0.01033 3.405
pair_coeff 1 2 lj/cut 2 0.08257 2.822

#Settings
timestep 0.0001
mass 1 58.6934
mass 2 39.948

neighbor 0.5 bin
neigh_modify delay 5

#Region Definition

region lower block INF INF INF INF INF 35.2
region upper block INF INF INF INF 46.2 INF
group lower region lower
group upper region upper
group boundary union lower upper
group mobile subtract all boundary

set group lower type 2
set group upper type 3

#temp controllers
compute mytemp mobile temp

velocity mobile create 300.00 45232 mom no dist gaussian ( I am getting the error here)

#Define fix
fix 1 all nve

dump myDump1 all custom 100 NiAr_velocity.dat id type vx vy vz
dump myDump2 all custom 100 NiAr_position.dat id type x y z
dump 1 all image 1000 dump.*.jpg type type

thermo 25
run 10000

2

Hi All,

Below one is my input file and i am getting this error " attempting to rescale 0.0. temperature" .I introduced an Ar atom on top of a initially heated Ni surface. Right now, i am trying to calculate the velocity of Argon at 300 K from Maxwell-Boltzmann Distribution and below one is my input file. I would realy appreciate if someone tell me what is wrong with the input file.

please activate your common sense (and a look into a text book on
statistical mechanics would not hurt either).

you *cannot* have a temperature *distribution* for a single atom. in
fact, the whole concept of temperature doesn't make much sense for a
single atom. under the assumption of equipartitioning, we do compute
temperature on small groups of atoms from their individual kinetic
energies, but the proper definition of temperature only works really
well this way in the limit of an infinite number of particles.

if you look through the LAMMPS documentation for how temperature is
computed, you'll see that LAMMPS by default will compute the number of
degrees of freedom per atom and subtract the number of translational
DoFs, as simulations under periodic boundaries are usually invariant
against translations. thus for a single atom, you end up with 0 DoFs
and thus cannot compute a temperature at all.

thus what LAMMPS is trying to tell you is that you are trying to do
something that doesn't make sense. you *can* assign a *specific*
kinetic energy to your atom, but it doesn't make any sense at all to
talk about a temperature for this atom and particularly not about a
temperature distribution (and BTW: a gaussian distribution is not a
maxwell-boltzman distribution, either).

axel.