[gmx-users] Npt equilibration of the membrane-mimicking, CCl4 layer
Thomas Schlesier
schlesi at uni-mainz.de
Fri Feb 10 17:51:45 CET 2012
Hi,
i second Justins seond idea (creating a small box of equilibrated CCl4
and then fill the simulation box via the -cs option).
Depending if you have other molecules in your system, make the
simulation box a little bit bigger, because you will get some holes. In
the subsequent NPT simulation these holes will vanish and your box will
shrink.
The first way has the problem, that if you have holes between some CCl4
molecules, no new molecule would fit there. Then you could make a short
equilibration, so that the holes gather to bigger holes and fill them
again with the -ci option. If you're lucky, you will reach after a very
long time the required density.
(In my case i used mesitylene as a solvent and after day of the
automated iteration (try to put molecules into the box, equilibrate, try
to put more molecules into the box, ...) i dismissed this approach.)
greetings
thomas
>
>
>
> James Starlight wrote:
>> Justin,
>>
>> 2012/2/6 Justin A. Lemkul<jalemkul at vt.edu<mailto:jalemkul at vt.edu>>
>>
>>
>> Some simple calculations using the desired density and the box
>> dimensions (to get the volume) will tell you exactly how many
>> molecules you need. If you only "suppose" you've got a reasonable
>> number, there are better ways to be sure ;)
>>
>>
>> In accordance to my calculations I need in ~ 1000 CCl4 to obtain 1.600
>> g/cm^-3 density with my box dimensins
>>
>> By the way way I try to define such system by
>>
>> genbox -ci ccl4.gro -nmol 1000 -box 8.6 6.5 3 -o ccl4_box.gro
>>
>>
>> I've obtain memory error :(
>>
>>
>> Is there any other way to define such system with the desired size and n_mollecules with the less computation demands?
>>
>
> Two options:
>
> 1. Add them incrementally to the box (100 at a time or so, until the desired
> value is reached)
>
> 2. Build a small box of solvent (a few hundred molecules), then use genbox -cs
> instead of -ci to use the small solvent box to create the one of desired size.
>
> -Justin
>
More information about the gromacs.org_gmx-users
mailing list