[gmx-users] Problem to simulate silica melts and amorphous silica with BKS potentials

[David Waroquiers]

Hello again.

It's pretty clear now for me thanks. I still have a little problem due
to the unphysical trend of the BKS potentials for r --> 0 : it goes to
minus infinity. The atoms can thus collide in case of a high temperature
(providing them enough velocity to overcome the barrier) and I get the
following error which I suppose is connected to this problem :

"Range checking error:
Explanation: During neighborsearching, we assign each particle to a grid
based on its coordinates. If your system contains collisions or
parameter
errors that give particles very high velocities you might end up with
some
coordinates being +-Infinity or NaN (not-a-number). Obviously, we cannot
put these on a grid, so this is usually where we detect those errors.
Make sure your system is properly energy-minimized and that the
potential
energy seems reasonable before trying again.

Variable ci has value -2147483648. It should have been within [ 0 ..
8 ]"

Practically, at ambient temperatures of the order of 300 K, it doesn't
happen but I want to anneal first at 5000 K to get a melt SiO2 and then
quench it to 300 K.

The standard procedure to avoid particle colliding is to add a hard
repulsive part acting only on short distances. Is it possible to add a
repulsive part on short distances only (something like
Vrep(r)=a*(r-r0)^4 with r0 being the distance at which the barrier is at
its maximum) and how could I do it ? The potential energy should then be
V=Vrep(r) for r < r0 and V=V_BKS(r) for r > r0.

If it's not possible analytically, I assume I should use a user
specified potential function by making a table.xvg file with the x,
f(x), f'(x), etc ... as it is mentioned in the manual (p133 and p148) ?

David Waroquiers - PhD Student
Université Catholique de Louvain
Laboratoire de Physico-Chimie et de Physique des Matériaux
European Theoretical Spectroscopy Facility

Mark Abraham wrote:
> David Waroquiers wrote:
> > Hello
> > 
> > I'm trying to simulate amorphous SiO2 with the use of the BKS potentials
> > (van Beest, B. W. H.; Kramer, G. J. & van Santen, R. A. - Phys. Rev.
> > Lett., American Physical Society, 1990, 64, 1955-1958).
> > 
> > According to these potentials, I should use the Buckingham potentials
> > with user-defined parameters plus a coulomb part with user-defined
> > charges for each atom type. I don't need any other forcefield from
> > GROMACS.
> >
> > When I try to preprocess my topology file and get the .tpr binary file
> > in order to start the simulation, I get the following message in the
> > terminal :
> > 
> > processing topology...
> > Generated 0 of the 3 non-bonded parameter combinations
> > Excluding 0 bonded neighbours molecule type 'O'
> > Excluding 0 bonded neighbours molecule type 'Si'
> > 
> > I don't understand why my non-bonded parameters aren't generated ...
>
> "Generation" refers to the process described in (probably) chapter 4 of 
> the manual where all possible pairwise interactions between a set of 
> atom types are calculated from generic parameters characteristic of the 
> two atom types involved (see your [atomtypes] entries). An alternative 
> is to describe the pairwise interactions specifically, as you have done. 
> Thus GROMACS didn't need to generate any, and reports this fact.
> 
> Mark