[gmx-users] Inconsistent non-bonded energy when using energy tables.

[Samuel Ntim]

Dear All,

I will be working with a system having Lennard-Jones interactions between some species and Buckingham interactions between other species. I know this is possible using tables.

As a first check to be sure I am doing the right thing, I set up a minimal system of two particles fixed in their initial positions and without charges. When I run a short simulation on this system with and without tables, I get similar non-bonded energy contributions to the total energy.

However, when I increase the number of particles to three, each equidistant from the the other two, I get a different contribution with tables, from without tables. Without tables, I get about three times the contribution in the previous set up (which is what I expect), but with tables the energy is only slightly different from that of the two-particle system.

I have attached the two simple systems and the files used in running the simulations in a .rar file here<https://www.dropbox.com/s/v9x7zb0tu4mkn5p/SIMPLE.rar?dl=0>.

I will be glad to have someone point me in the right direction. Thank you.


Best regards,
Samuel