Re: [ESPResSo] Position-dependent Langevoin thermostat

[Christoph Junghans]

Hi Lorenzo & Olaf,

there is no problem in modifying the langevin thermostat as long as you 
keep the ration between noise and friction. In Espresso the noise and 
friction prefactors are saved in langevin_pref1 and langevin_pref2.
In theory you just have to change them.

From the implementation side this is more complicated, because in 
Espresso there is only one global friction constant for all particles.
So inside of the function friction_thermo_langevin, you need to know the 
number of neighbors, the position of the aggregate or whatever.

Cheers,

Christoph

Olaf Lenz schrieb:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi Lorenzo!
>
> Lorenzo Isella wrote:
> > There is a fundamental problem with this approach: once a monomer is
> > surrounded by other monomers, it should feel a different noise due to
> > the screening by its neighbors, but this effect is not accounted for
> > in Espresso. I refer to the case in which there is no shear, just
> > Brownian motion.
> > Nobody knows exactly what it should feel, but it would be (at least
> > for me) wonderful to have a feature in Espresso that allows me, while
> > keeping the overall system temperature around T, to have the noise on each
> > monomer in an aggregate expressed as a function of e.g. the number of
> > first neighbors of that monomer.
>
> The Langevin thermostat has been constructed as is to model the
> statistical NVT ensemble, while still being very simple.
> I think that if you would modify the Langevin thermostat in that way
> without exactly knowing what you're doing, you would mess up this
> statistical ensemble that your simulation lives in, and I wouldn't
> expect that any of the effects that you might see in such a simulation
> has any resemblance to what happens in nature.
>
> As for whether there might be useful alternatives for you, I think that
> others on this list have a better founded opinion on this matter.
>
> Olaf
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.4-svn0 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFI0PkstQ3riQ3oo/oRAn4XAKC48hivq1hClOZPfwhvy8jfOG1v7wCeJElD
> lUFUePZDMKtouByO8rTqUKw=
> =rqLD
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> ESPResSo mailing list
> address@hidden
> https://fias.uni-frankfurt.de/mailman/listinfo/espresso

--
Dipl.-Phys. Christoph Junghans
Max Planck Institute for Polymer Research
Theory Group
POBox 3148
D 55021 Mainz, Germany

Phone: +49 6131 379 335
Web: http://www.mpip-mainz.mpg.de/~junghans