[lammps-users] potential energy zero : pair_style

Dear Lammps users,

Version : September 2021

I am performing Langevin simulation for a dimer (ethane) at 300K in barrier crossing. The pair_style

interaction is modeled using LJ potential and the entire system is kept in the external double well potential.
However, I am getting the potential energy values as zero which should not be.

This has been discussed in the mailing list, and the issue might be in the pair_style implementation.

In my simulation, I have written as follows for ethane dimer :

pair_style lj/cut 2.5
pair_coeff * * 4.418 230.1 2.5

The computation that calculates the potential energy of the dimer using “compute” command also gives
zero which again confirms that there is no interaction between two monomers.

Kindly let me know which one I should check further to fix this issue. I should be able to track down but couldn’t.

Input script : dimer.in
input data : temp.dat

dimer.in (1.35 KB)

temp.dat (293 Bytes)

Please note that computer programs are “blind” to conceptual errors and thus just blindly apply input that is provided without checking if this makes sense.
Unfortunately, not much of what your input contains makes sense conceptually, starting from doing an MD simulation for just a single pair of atoms.
Further issues:

  • you choose reduced units but then enter values that look more typical for other choices of units
  • your LJ parameters make no sense at all with a sigma of over 200 and a cutoff of 2.5. as mentioned before, you are missing the point of using reduced units.
  • your LJ potential is wiped out by the pair style table.
  • you claim you have an external double well potential, but obviously that is not the case, as you represent it with a pair-wise potential. that makes no sense at all. an external potential would be implemented using fix addforce, for example
  • if you run your input, you should be seeing in the log file that there is not neighbor list request for lj/cut, also the info all command can provide you useful information about your input.

It looks to me that you need some proper in person tutoring in how to conduct simulations and work with software in general and how to confirm that what you input is actually applied and makes sense. if memory serves me well, you’ve been struggling with this for quite a while, so perhaps it would be better to approach this differently and get a helping hand involved.