Hi,
I am using the fix heat command to provide heat to a system. However, I would like to know how does lammps provide the heat energy.
From the manual, I understand that the command will add non-translational kinetic energy to the atoms which conserves the aggregate momentum. However, my system is a atomic system. Hence, may I know what does non-translational kinetic energy refers to?
Finally, I would like to confirm that the eflux should be scaled with the number of atoms in the group means that the energy would be supplied uniformly to each atoms?
Thank you very much.
Hi,
I am using the fix heat command to provide heat to a system. However, I
would like to know how does lammps provide the heat energy.
From the manual, I understand that the command will add non-translational
kinetic energy to the atoms which conserves the aggregate momentum. However,
my system is a atomic system. Hence, may I know what does non-translational
kinetic energy refers to?
fix heat doesn't care about whether you have atoms or molecules.
heat is added by scaling the velocities after the center of mass
velocity has been removed hence no change to the total momentum.
Finally, I would like to confirm that the eflux should be scaled with the
number of atoms in the group means that the energy would be supplied
uniformly to each atoms?
no, it doesn't.
axel.
Dear Axel,
Thanks for the explanation on the fix heat. On the compute temp/region command, I understand how the formula works. However, may I know how Lammps deal with the boundary cases? For example if an atom is exactly on the boundary of the defined region, does Lammps include it in the calculate of temperature?
I wrote a C file which calculates the temperature based on the positions and velocities dump in the xyz file. However, the temperature calculated is slightly different from that of Lammps.
Thank you.
Dear Axel,
Thanks for the explanation on the fix heat. On the compute temp/region
command, I understand how the formula works. However, may I know how Lammps
deal with the boundary cases? For example if an atom is exactly on the
boundary of the defined region, does Lammps include it in the calculate of
temperature?
I wrote a C file which calculates the temperature based on the positions
and velocities dump in the xyz file. However, the temperature calculated is
slightly different from that of Lammps.
i doubt it is due to rounding, but please note that output of coordinates
always constitutes some kind of rounding and truncation, so "exactly on the
boundary" is ill defined.
in any case, you are probably not aware that the temp/region compute does
not remove any degrees of freedom by default. you can change this via
compute_modify. see the documentation for details.
axel.
Dear Axel,
From the manual, I understand that degrees of freedom can be remove from the computational of the temperature in the region using the extra option of the compute_modfiy. However, I am not very sure about the concept of degree of freedom for the system. Since I am just trying to verify the c file, I didnt include any process apart from the airebo pair potential which causes bonding to occur between the carbon atoms and a fix nve to let the system equilibrate.
I am sorry but I am not familiar with the concept of degree of freedom.
Thank you.
Dear Axel,
From the manual, I understand that degrees of freedom can be remove from
the computational of the temperature in the region using the extra option
of the compute_modfiy. However, I am not very sure about the concept of
degree of freedom for the system. Since I am just trying to verify the c
file, I didnt include any process apart from the airebo pair potential
which causes bonding to occur between the carbon atoms and a fix nve to let
the system equilibrate.
I am sorry but I am not familiar with the concept of degree of freedom.
then you better figure it out. this falls into the "you have to do your
homework" category. you can't compute temperature from an MD simulation
properly without understanding this. just grab a stat mech text book and
start reading.
axel.