atoms don't move

LAMMPS LAMMPS Mailing List Mirror
#1

Hi,

I added a new potential to LAMMPS, and I output the force and potential energy. I only have two atoms in the data file and they have zero velocities in the beginning. I run a NVE ensemble and the output force is nonzero, but the atoms do not move and the temperature is 0 all the time. What is wrong?

Lisa

#2

Hi,

I added a new potential to LAMMPS, and I output the force and potential
energy. I only have two atoms in the data file and they have zero velocities
in the beginning. I run a NVE ensemble and the output force is nonzero, but
the atoms do not move and the temperature is 0 all the time. What is wrong?

hard to say. please replace your new potential with a known potential
(e.g. lj/cut) and assign suitable parameters and see if the same thing happens.

if yes, please post that input and we can have a look.

axel.

#3

I have checked the tersoff/zbl potential and the temperature is not zero since the run starts. It seems that only my new potential has this condition.

2011/7/13 Axel Kohlmeyer <[email protected]>

#4

I have checked the tersoff/zbl potential and the temperature is not zero
since the run starts. It seems that only my new potential has this
condition.

ok, so the input works, thus the error has to be in
the code that you wrote. how did you check that
the forces are correct? by adding print statements
to your code or by writing forces to a dump file?
if only the former, please do the latter and check
if they match.

axel.

#5

I tried to dump the force, and it is zero for two atoms. But I put the atoms so close and when I directly print it in the code, it is around 5000 eV, how can it happens?

2011/7/13 Axel Kohlmeyer <[email protected]>

#6

I tried to dump the force, and it is zero for two atoms. But I put the atoms
so close and when I directly print it in the code, it is around 5000 eV, how
can it happens?

did you actually increment the forces into the atomic force arrays?

axel.

#7

Ok, I found the stupid probelm. The two atoms only have forces along the z direction, but I mistype the dz to be dy in the code, so the force is zero. The potential works fine for now. The total energy finally conserved. Thank you so much. I will check the bond angle and bond distance of a sample.

Lisa

2011/7/13 Axel Kohlmeyer <[email protected]>

#8

Hi,
My new potential can not be read by restartdata command. How can I setup it? I saw the pair_sw.cpp and pair_tersoff.cpp and did not find any setup for the restartfile. The error is in below:

Reading restart file …
WARNING Restart file version does not match restart2data version (The version of lammps is 7May 2011, I don’t know why the restart2data version is 6 Jan 2011)
restart2data version = 6 Jan 2011
ERROR: Unknown pair style watanabe

Lisa

2011/7/13 Lisa Lan 5a <[email protected]…24…>

#9

Hi,
My new potential can not be read by restartdata command. How can I setup it?
I saw the pair_sw.cpp and pair_tersoff.cpp and did not find any setup for
the restartfile. The error is in below:

Reading restart file ...
WARNING Restart file version does not match restart2data version (The
version of lammps is 7May 2011, I don't know why the restart2data version is
6 Jan 2011)
restart2data version = 6 Jan 2011
ERROR: Unknown pair style watanabe

you have to edit restart2data.cpp

just look for what is done with a suitably similar existing
potential and add and else branch for your potential where
needed. should be in two locations, if i remember correctly.

cheers,
   axel.

#10

Ok, it works. Thank you.

Lisa

2011/7/14 Axel Kohlmeyer <[email protected]>

#11

Thanks. I was wondering about this too.