Re: [ESPResSo-users] NPT Isotropic with P3M activated

[Olaf Lenz]

Hi!

On 02/22/2011 11:12 PM, Salvador H-V wrote:
> Could anyone give me a hint, please. If it ia possible in the
> current version of Espresso to perform a NPT simulation with one or
> two dimensions of the box being fixed and using electrostatics with
> P3M activated?

Hanjo's remark in his mailing is not wholly true. In the current release
version, it is impossible, as P3M doesn't know non-cubic boxes. However,
the development version of ESPResSo has code for non-cubic P3M. I expect
to do the next release within the next month or so.

Note, however, that the NPT ensemble code doesn't support non-cubic
boxes (that's why it's called npt-isotropic). If you want to do a
non-isotopic NPT ensemble, you will have to use a hybrid Monte-Caro
scheme, where you implement non-isotropic volume moves in Tcl. This is
not very complicated and should work fine, however you might lose some
performance because you have to do the MC moves in Tcl. On the other
hand, you don't have to do the MC moves very often, so it is probably a
non-issue.

So, you will definitely need the development version of ESPResSo, or you
have to wait for the new release, but then it should be possible.

If you need a short intro how to use the development code, just contact me.

Olaf


-- 
Dr. rer. nat. Olaf Lenz
Institut für Computerphysik, Pfaffenwaldring 27, D-70569 Stuttgart
Phone: +49-711-685-63607

olenz.vcf
Description: Vcard