[lammps-users] why Nan?


I'm trying to add two dummy atoms into the system. In the attached test
case, there are two
cyclo-Hexane molecules defined in the data file and two extra dummy atoms at

I used a fix nvt for all real atoms. I hope I can exclude the dummy atoms
for the simulation.
But I got 'nan' output for the energy output, vdW term. Can anybody find out
where is wrong?
And maybe show me how to add 'dummy' atoms in a proper way. Thanks!

data.test file:

What you are doing might work if you put the dummy atoms
at different positions. Else they will have r = 0 separation
and you'll get an INF or NaN when you compute with them.

You could also use neigh_modify exclude command to
exclude interactions with those atoms, or pair_style hybrid
and assign a "none" style to those interactions.