External biasing

Hello guys, this question is more of a conceptual understanding of Lammps. I have a simulation of box of length -15 to 15 and I am adding an external potentials of u(x)=-0.1x and v(x)=2(sin(x))^3-0.1(x). The problem that I am having is that when I use a dump command to output my atom positions and forces, I find the atoms only spanning in the negative coordinates. This is a problem for me since I want the forces acting on the particles found at both the negative and positive coordinates. I am required to use only a unit velocity in all directions. What can I do to get the required forces in both the negative and positive coordinates?

It is not clear from your description what you are trying to achieve and specifically how you are doing it. It would be best if you provide an input deck that can be run and modified to reproduce what happens and to explore why.

What do you mean by that?

Wong question. You need to figure out why atoms go where they go. That must be due some input of yours. Only when you understand the why can you look into the how. If you just suppress the symptom, you are not doing science but computer animation.

script2.in (5.6 KB)
Attached is the input script for the job. I am working with a one particle simulation and a temperature of 1 so I used velocity set to 1 in the x,y and z directions. The goal of the job is to output the forces acting on the particles and postprocessing the data to get the energy surface by integrating the forces.

There are multiple problems with your input.

  • it is near unreadable due to the excessive use of variables
  • it makes no sense at all to apply a thermostat to a single particle and temperature is ill defined in that case (what does it couple to and exchange kinetic energy with?). only kinetic energy is.
  • because you set a velocity vector of 1 1 1 your one particle will move in that direction modulated by the external potential
  • I don’t understand your purpose. Since you have an analytical external potential, you can integrate it analytically. What you claim to try to do is not likely to work this way.