# Temperature mismatch

Dear Users,

My simulation has atoms in a carbon nanotube (CNT) and atoms which are outside this CNT. The atoms in the CNT are in the NVE ensemble.

I presume the other atoms are just frozen as they are not in any fix.

The temp reported for the CNT atoms agrees with my thermostat (0.5K) but the default ‘temp’ reported in the log file is much much higher (e.g. 1500 K).

How could 250 stationary atoms cause the overall temp to be so much higher. To be honest I expected the ‘temp all’ and ‘temp for CNT group’ to be similar.

I would be grateful if anyone could explain this to me.

Daniel

……………………………………………………………………………………………………………

Daniel Mulvihill

Government of Ireland Research Fellow,

Department of Mechanical, Aeronautical, and Biomedical Enginering,

University of Limerick

Dear Users,

My simulation has atoms in a carbon nanotube (CNT) and atoms which are
outside this CNT. The atoms in the CNT are in the NVE ensemble.

I presume the other atoms are just frozen as they are not in any fix.

guessing is a bad strategy. you have to know. you can easily just dump
all information into a file and look at it in detail and make certain.

The temp reported for the CNT atoms agrees with my thermostat (0.5K) but
the default ‘temp’ reported in the log file is much much higher (e.g. 1500
K).

How could 250 stationary atoms cause the overall temp to be so much higher.

To be honest I expected the ‘temp all’ and ‘temp for CNT group’ to be
similar.

I would be grateful if anyone could explain this to me.

LAMMPS computes what you tell it to compute.

The temperature is calculated by the formula KE = dim/2 N k T, where
KE = total kinetic energy of the group of atoms (sum of 1/2 m v^2),
dim = 2 or 3 = dimensionality of the simulation, N = number of atoms
in the group, k = Boltzmann constant, and T = temperature.

so it is just applied. to either your CNT or the entire system. the
kinetic energy is, of course, 1/2 m v**2. now if you don't time
integrate some atoms, that does not imply that they have zero velocity
unless you explicitly set it. also, including additional atoms will
add degrees of freedom, so if you include atoms that are not moving,
the resulting temperature will be bogus as you have seen.

axel.

Thanks Axel,

Whats the best way to output info on what those unintegrated atoms are
doing velocity and temp wise?

My CNT contains 6000 atoms (these are integrated fix NVE) and no fix is
applied to the other 250 atoms. How can these 250 atoms have velocities
and temperature? I have not applied any velocities to them.
I also find it hard to see why the 'temp all' reading can be so high
(1500 K) compared to the CNT temp of 0.5 K. Even if these atoms have
small oscillations how can they affect the result so much?

Daniel

Thanks Axel,

Whats the best way to output info on what those unintegrated atoms are
doing velocity and temp wise?

write out a custom dump and look at it.

My CNT contains 6000 atoms (these are integrated fix NVE) and no fix is
applied to the other 250 atoms. How can these 250 atoms have velocities
and temperature? I have not applied any velocities to them.

how can you expect them to have any kind of consistent data, when you
don't initialize it.

I also find it hard to see why the 'temp all' reading can be so high
(1500 K) compared to the CNT temp of 0.5 K. Even if these atoms have
small oscillations how can they affect the result so much?

atoms that don't move have no oscillations, and the kinetic energy
does not depend on positions.

it is impossible to discuss these things properly only based on your say-so.
there are far too many ways to make mistakes, overlook important
details, or misinterpret inputs, documentation, or output.

unless you provide convincing proof that LAMMPS is at error, we have
to assume that the problem is on your side and thus it is your job to