[lammps-users] external efield and temperature control

Dear all,

Thank you very much for always telling me many important tips for LAMMPS.
I could make lammps with openmpi, Axel. Thank you so much.

NowI would like to calculate non-equilibrium molecular dynamics, but I
have a trouble to select fix command options...
I am trying to calculate a bunch of molecules under very strong electric
First I used NVT ensemble but I got an error at the middle of iteration
as following.

"ERROR on proc 0: Bond atoms 20 23 missing on proc 0 at step 184356"

I thought that it is because there is a conflict between keeping
temperature constant and applied force by electric field, so I think I
cannot use NVE, too.
Actually I tried NVE, but I got an error as I expected.....

"ERROR on proc 0: Bond atoms 20 23 missing on proc 0 at step 88451"

It seems there are many options in fix command to manage it, but I am
still not sure which is the best for my calculation.
It might be very trivial questions in MD field, but if you have an
experience to run LAMMPS with efield options, I would be very glad if
you tell me anything you know.
Thank you very much.

Best regards,
Nanako Takahashi

If bond atoms are missing it is usually b/c the 2 atoms in the bond
flew far apart. You need to figure out why that happened.


Dear Nanako,
The reason that your simulation is failing is due to the fact that your bond length of a particular bond has increased a lot at certain timestep. Check out what could be the reason behind it. May be u are applying a huge electric field? Not sure though.

Dear Vikas and lammps users,

Thank you so much for your reply, Vikas.
Yes, I apply very strong electric field, and it seems the bond break because one of atoms in molcule is beyond calculation box for now. (I checked the coordinate of the atom.) And I use boundary command "f", so the atom is deleted, but other atoms are alive.....
I also discuss this in mail titled "Delete molecules at end of calculation box"
I would be appreciate you so much if you tell me any ideas.


Vikas Varshney wrote:

If one atom in a bond pair gets deleted by passing
out of the simulation box that
will be fatal. You will get the error you indicate.