[gmx-users] Temperature coupling of two-phase system

[Erik Marklund]

Dear Jason,

MD thermostats are in some sense unrealistic to begin with, considering how energy is transferred to and from the atoms. Alas, that seems to be unavoidable for microscopic emulation of the macroscopic case. For systems where some degrees of freedom are very loosely coupled to others, like in two phase systems, you should couple the parts separately, i.e. tc-grps = CNT SOL in your case. If you have ions in your system you go with tc-grps = CNT SOL_and_ions.

Kind regards,
Erik

On 16 Sep 2014, at 02:47, 朱文鹏 <jasonzhu925 at gmail.com> wrote:

> Dear Gromacs users,
> 
> I am trying to simulate nanotube-based channel (CNT: 3872 carbon atoms)
> with water (SOL: 15758 H2O). The nanotube is position restrained during
> simulation, while water molecules are constrained by . I've used two kinds
> of tc-grps setting. One is "tc-grps = System". The other is "tc-grps = CNT
> SOL". For both cases, I calculated temperatures of each group by using
> g_traj. They were corrected by accounting the degrees of freedom (DOF) of
> CNT (5805) and SOL (94545). However, I am confused about the temperature
> results and not sure which one is correct.
> 
> For the "tc-grps = System", the temperature of water is reported to be 300K
> as I expect. But the CNT temperature fluctuates around ~66K. It is much
> lower than the target temperature. I think it is caused by the huge
> difference of atom numbers and mass between CNT and SOL. Is it correct?
> 
> For the case of "tc-grps = CNT SOL", the temperature coupling is more
> accurate. The temperatures of both groups can reach the target temperature
> (300K). But I am told that such a thermostat for every component separately
> is unrealistic.
> 
> I wonder which tc-grps setting I am supposed to use for temperature
> coupling. tc-grps = System, tc-grps = CNT SOL, or even tc-grps = SOL only
> (it is like using 300K water to warm up CNT). I would appreciate it if you
> have any suggestion.
> 
> Thank you very much for your attention.
> 
> Best,
> Jason
> -- 
> Gromacs Users mailing list
> 
> * Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!
> 
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> 
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a mail to gmx-users-request at gromacs.org.