[gmx-users] Erik.Please help me! Some question about the NPT equirbrium CCl4 Box!

[Christoph Freudenberger]

Hi there,

PeiQuan Chen wrote:
> Dear gmx-users：
>   I am in the process of equilibrating a box of CCl4 in NPT!
>   But in this case, there must be something I am doing wrong
> because I don't get the right density(calculate density>exp 
> density). I construct a box that contained 216 molecules,and 
> perform Em on that box. Afther that I make a 500ps NVT 
> Dynamics on that to get a NVT equibrium box. Above all is 
> successful. But When I do a 1500ps NPT dynamics on the box 
> of NVT equibrium, It begins wrong with my system. The system
> potential decrease with times and the box vector begin to 
> decrease too(3.25878-->3.13426). Can anybody give me a hint how
> to get the right NPT equibrium box of the system?
[...]
> Does it seems something wrong of my files? And does the 
> oplsaa force field is not suitabe for CCl4? And how to get 
> a equibrium box of CCl4?
No, your procedure for setup and equilibration of the system
is ok. So are your files.
I suppose it has got to do with the forcefield parameters
themself.

You will never get perfect agreement with the exp. values.
One forcefield is suited best to reproduce on property, anotherone
will work better for something else.
Try to compare a few more properties.
How much is the density off the exp. value?
What about DHvap and self diffusion?

I have been playing with the topolgies for MeOH and DMSO as they
are implemented in gromacs a lot in the last week, finding both
models to give way to high densities and bad DHvap.
I have found new paramters for both models from the literature
giving much results now.

regards
-- 
Christoph Freudenberger
Abt. Organische Chemie I AK Siehl - Uni Ulm -Tel: ++49-731-502-2785