[gmx-users] mdrun in vacuum diverging? -- possibly due to nopbc?

[David van der Spoel]

Yousung Jung wrote:
> Dear more experienced users,
> 
> My system is a water/methanol mixture with 100 and 300 molecules for 
> water and methanol, respectively. I equilibrated the system with pbc for 
> 500 ps at 300K in vacuum. Then, I ran the actual simulation at 235K in 
> vacuum WITHOUT pbc to simulate a droplet of the mixture. I see a few 
> water and methanol molecules escaping "the system" (or, the droplet) 
> during the simulation.
> 
> The problem that I am struggling with for this system is, after about 
> 400 ps of the actual dynamics at 235K, temperature starts fluctuating 
> badly, getting worse until the simulation stops at around 600 ps. When I 
> visualize the trajectory, I see the system (a droplet in vacuum) 
> rotating and translating at around the same time as the temperature 
> begins fluctuating. I did use comm_mode = angular to eleminate rot. and 
> trans. modes.
> 
> I reduced the time-step to 0.5 fs, but the problem still exists (the 
> system translating significantly in vacuum). Is it possible that this is 
> due to the fact that the center of mass is used for removing 
> translational and rotational motion and a few escaped molecules are 
> causing this problem? (If those ones that are escaping are causing the 
> problem, I guess I could use pbc with a "big" box.) Assuming this is the 
> case, I have two questions:
> 
> 1. Is there a way to fix this rotational and translational artifact 
> without using pbc for my system?
> 
First make sure your system does not have angular velocity and set 
comm_mode = linear
Second run in double precision without cut-off
Third, set shake tolerance to 1e-8 or increase lincs_order
Using the things above you can still use 2 fs time steps.


> 2. If I have to use pbc, what would be the most economic combination in 
> terms of box size and etc?
> 
> Thanks very much for your help!
> 
> Yousung
> 
> ps. I am attaching my mdp file below, and an xvg file for temperature 
> analysis.
> 
> 
> ----------------
> 
> integrator          =  md
> tinit               =  0.0
> dt                  =  0.001
> nsteps              =  1000000
> nstcomm             =  1
> nstxout             =  1000
> nstvout             =  1000
> nstfout             =  1000
> nstlog              =  1000
> nstenergy           =  1000
> nstlist             =  0
> energygrps          =  System
> ns_type             =  simple
> pbc                 =  no
> comm_mode           =  angular
> 
> ; Constraints
> constraints         =  all-bonds
> constraint_algorithm = lincs
> 
> ;Treatment of Vdw and Elctrostatics
> vdwtype             =  cut-off
> coulombtype         =  cut-off
> rlist               =  0
> rvdw                =  0
> rcoulomb            =  0
> 
> ; Temperature coupling is on
> Tcoupl              =  berendsen
> tau_t               =  0.1
> tc-grps             =  System
> ref_t               =  235
> 
> ; Pressure coupling is off
> Pcoupl              =  no
> 
> ; Generate velocites is on at 235K
> gen_vel             =  yes
> gen_temp            =  235
> gen_seed            =  173529
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php


-- 
David.
________________________________________________________________________
David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
Dept. of Cell and Molecular Biology, Uppsala University.
Husargatan 3, Box 596,  	75124 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++