[gmx-users] User tabulated potential and VdW Modifier

[Kiesel, Matthias]

Hello everybody,

I sent a message a week back with the same title and I hope this gets put into the right thread now (did not know how to respond directly). I actually did some more testing and I think I found out how everything works and I'm going to leave this here in case someone else ever wonders.

vdw-modifier = potential-shift does not shift a user specified potential, this can be seen from the substantial energy drift in the NVE simulations in my first message, which was eliminated by feeding a user list with shifted potential.

The modifier, however, adds and additional term to the LRC, namely the correction for the "lost" potential energy inside the cutoff sphere (see Frenkl and Smit). This correction is  not calculated with the user tabulated potential, but rather with the LJ potential (and even using this I wasn't completely able to reproduce the value, I'm 3kJ/mol short, which is appr. 0.2%).

>From the behaviour of the potential modifier I used, I would also guess the other modifiers do not work and gromacs overrides whatever I choose with "None".

kind regards,

Matthias