[gmx-users] David, adding new atom types to gromacs "shell model"

xieyh at hkusua.hku.hk xieyh at hkusua.hku.hk
Mon Mar 14 20:09:17 CET 2005

Dear David:

The following is one email you replied previously. 

I am a little confused with "give shell a mass and do normal dynamics". 

Let me guess, normally, in shell model, the shell particle is massless and
[water_polarization] item should be considered in sw.itp. However, if using
approximating method, we can appoint a small mass to shell particle. What I
wanna know is about the meaning of normal dynamics. 

In detail, if a mass is given to shell, then we need not consider
[water_polarization] any more? Is it right? On the contrary, I just need to
define [bond], [angle] etc items for describing the connection of shell
particle to dummies or other atoms? Could you help me to confirm myself?

Thanks your kind help for so long a time.


Xie YH
Hong Kong Univ.

On Wed, 2004-05-05 at 02:36, Julius Su wrote:
> >Hello,
> >
> >I had some questions about the shell model implemention in GROMACS --
> >
> >(1) is it possible to define new shell types?
>Can you make yourself more clear? Right now a shell is a point charge
>connected to another particle (or dummy position) with a spring. The
>shell is massless and its position is optimised at every time step.
>Obviously one can also give them a mass (e.g. 1 amu) and do normal

More information about the gromacs.org_gmx-users mailing list