Border conditions on Molecular Dynamics Simulation

[mielgosez]

I'm building a code for Molecular dynamics (in C).
I'm employing Lennard Jones Potential, periodic border conditions and I integrate Newton's equations by making use of Verlet Integration.
When I plot Time vs Energy graph, what I get isn't a constant! but when border condition are remove, then energy is conserved! What is happening?

I appreciate any suggestion, thanks​

The code is in the following link (border condition are in red):http://es.scribd.com/doc/62027867/Molecular-Dynamics

 

[Future Bruno]

-CodeThe periodic boundary conditions you are using may be introducing some numerical errors in the integration process. This can cause the energy to be slightly off from what it should be. To minimize this effect, you should make sure that the timestep you are using is small enough to accurately capture the motion of the particles. Additionally, you may want to consider using a more sophisticated integration method that is better suited for dealing with periodic boundary conditions.