[gmx-developers] SD temperature coupling groups

hess at sbc.su.se hess at sbc.su.se
Tue Jun 30 09:14:38 CEST 2009


Hi,

This functionality only works for the global thermostats,
not for SD.
We could easily implement it for SD and we should.

Berk


> Hi GROMACS developers' group!
>
> I've been having some problems (blowing up) when I tried to simulate a
> molecule such that one part is coupled to a thermal bath using SD, and
> another part is not coupled to any bath. Effectively, the uncoupled part
> has infinite "relaxation time" (tau-T) which is the same as zero coupling
> (gamma); however, in the *.mdp file it is unrealistic to put in the value
> infinity for the tau-T parameter. The 4.0 manual specifically says, in
> section 7.3.14, "Temperature coupling":
>
> tau t: [ps]
> time constant for coupling (one for each group in tc grps), 0 means no
> temperature coupling
>
> However, when I tried putting in zero I got the following (debugging
> information log-file)
>
> line 267: SD const tc-grp 0: b nan c inf d -inf
> line 268: SD const tc-grp 1: b 2.08438e-14 c 8.33021e-11 d -2.5e-07
>
> where group 0 had tau-T[0] = 0 and group 1 had tau-T[0] = 1 ps . This
> shows that when I put zero in the temperature coupling timescale (for SD
> this is not exactly a first-order relaxation time), the SD parameters blow
> up.
>
> Note that when I put tau-T[0] = 1000 or tau-T[0] = 10000, the simulation
> ran properly.
>
> Where in the code can I check that what is written in the manual is,
> indeed correct?
> (i.e., I expect an if statement saying that if the tau-T value of some
> temperature coupling group is zero, then take the inverse of that to be
> effectively zero).
>
> Thanks in advance,
>
> --
> Inon   Sharony
> ינון     שרוני
> +972(3)6407634
> atto.TAU.ac.IL/~inonshar
> Please consider your environmental responsibility before printing this
> e-mail._______________________________________________
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-developers-request at gromacs.org.




More information about the gromacs.org_gmx-developers mailing list