[gmx-users] high temperature

[Justin Lemkul]

On 3/1/15 2:46 AM, mah maz wrote:
> Dear all,
>
> I have a CNT system. I have set the mdp file as follows, but the
> temperature output is logically more than what I expected even in the
> production step(around 500k). I have changed the mdp file several times but
> it doesn't work. What are the possible causes of problem?
> Thank you!
> dt                  =  0.0001
> tinit               =  0
> nsteps              =  2000000
> nstxout             =  1000
> nstvout             =  1000
> nstfout             =  0
> nstlog              =  1000
> nstenergy           =  1000
> nstlist             =  5
> ns_type             =  grid
> rlist               =  0.9
> coulombtype         =  cut-off

Plain cutoffs introduce serious artifacts.  It appears from the tc-grps below 
that you have just a CNT in vacuo?  If this is the case, you should be using 
infinite cutoffs (rlist=rcoulomb=rvdw=0)

> rcoulomb            = 0.9
> rvdw                =  0.9
> rvdw_switch         =  0.5
> vdwtype             =  switch

These cutoffs seem totally arbitrary to me; they probably aren't conserving 
energy well and you're getting heating from artifacts.

-Justin

> pbc                 = no
> integrator          = md-vv
> Tcoupl              =  nose-hoover
> tau_t               =  0.1
> ref_t               =  300
> tc_grps             = CNT
> gen_temp            = 300
> pcoupl               = no
>

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================