I was playing around with the "analyze energy" command when I found that it gave me zero energy for systems which I assumed was supposed to give non-zero energies. I tested a small script as follows:
I was expecting a non-zero energy but it returns the following:
{ energy 0.0 } { kinetic 0.0 } { 0 0 nonbonded 0.0 }
If I change the lennard-jones cut-off to 3.1, however, I get a non-zero energy; I get the following output
{ energy -0.041592730503778354 } { kinetic 0.0 } { 0 0 nonbonded -0.041592730503778354 }
I dug up an old thread from 24 Aug 2007 and it says
For the future you can test your question more easily yourself with:
setmd box_l 10 10 10
part 0 pos 1 1 1 type 0
part 1 pos 1.5 1 1 type 1
inter 0 1 lennard-jones 2 2 1 1 0
analyze energy
If this doesn't give you a nonzero energy, please bug me again!
This does not, in fact, give me a nonzero energy; it gives me a zero energy. I played around with the feature some more and I think that the Lennard-Jones cuts off at 1 less than the cut-off value specified (or r_cut - 1). Is this really the way the potential works or did I do something wrong? I tried this on different machines and they gave me the same results. I am using ESPresSo 3.2.0.
--
Kue, Karl Thomas Y.