[lammps-users] Epsilon and sigma LJ parameters in CHARMM param files vs. LAMMPS data files

LAMMPS LAMMPS Mailing List Mirror
1

Hi,
I have a separate dialogue running about a simple simulation I'm trying to accomplish in LAMMPS (getting ethane to go to the liquid phase at the saturation temperature & pressure). One of the questions I was asked was whether my parameters in my data file (generated from the charmm2lammps tool) were correct.

One thing I have noticed was in the CHARMM parameter file the Lennard-Jones equation, for which non-bonded parameters are listed, is listed as:

!V(Lennard-Jones) = Eps,i,j[(Rmin,i,j/ri,j)**12 - 2(Rmin,i,j/ri,j)**6]

(...which presumably converts to E = epsilon [ (sigma/r)^12 - 2 (sigma/r)^6 ] )

In LAMMPS, the reference page lists the non-bonded interactions when using a CHARMM forcefield to be the traditional Lennard-Jones potential, which is given as follows:

E = 4 epsilon [ (sigma/r)^12 - (sigma/r)^6 ]

The charmm2lammps tool seems to leave the epsilon value alone, other than dropping the negative sign, found in parameter file (I would also expect a division by 4). Further, it performs the following equation on the sigma value:

$cols[3]*2.0**(-1/6)

I understand the reasoning behind the multiplication by 2 as in the parameter file sigma (which is referred to as capital 'r') is listed as:
Rmin/2: A, Rmin,i,j = Rmin/2,i + Rmin/2,j

...so the list value is divided by two in the parameter file and we're undoing that division. But why take the -1/6 power of it, when it's already listed as being inside the parentheses in the parameter file, as is the LAMMPS equation sigma?

Sorry if that isn't clear.

Also how does that get us from:
E = epsilon [ (sigma/r)^12 - 2 (sigma/r)^6 ]

to:
E = 4 epsilon [ (sigma/r)^12 - (sigma/r)^6 ] ???

This is vital to me being able to access if the generated parameters in my data file are correct.

Thank you in advance for your help!!

Cheers,
Jason R. Mick
WSU University
Graduate Research Assistant

2

Hi

hi jason,

,
I have a separate dialogue running about a simple simulation I'm trying to accomplish in LAMMPS (getting ethane to go to the liquid phase at the saturation temperature & pressure). One of the questions I was asked was whether my parameters in my data file (generated from the charmm2lammps tool) were correct.

One thing I have noticed was in the CHARMM parameter file the
Lennard-Jones equation, for which non-bonded parameters are listed, is
listed as:

!V(Lennard-Jones) = Eps,i,j[(Rmin,i,j/ri,j)**12 - 2(Rmin,i,j/ri,j)**6]

(...which presumably converts to E = epsilon [ (sigma/r)^12 - 2 (sigma/r)^6 ] )

here is your problem. the CHARMM Rmin is not the LAMMPS sigma. Rmin is
the minimum of the potential and sigma is the point where the potential
goes becomes zero. for a 12-6 lennard jones holds:
sigma = 2**(1/6) * Rmin
thus the conversion listed below. epsilon is unaffected.

i've been using the converter scripts before and compared them
agains NAMD (which reads CHARMM style directly) and the _do_ work.

thus in re your other simulation problems i have one question:
did you provide the proper simulation cell dimensions? the
problems you reported would match giving too small a cell.

cheers,
   axel.