[gmx-users] v-rescale and conserved energy

[Berk Hess]

Hi,

The integrator is not the issue. The cut-off setup is. 
You should use interactions that go smoothly to zero with a buffer region in the neighborlist.
Use for instance shift for LJ interactions. PME could be switched, but that is a much smaller
contribution. Use rlist about 0.3 nm larger than rvdw (and rcoulomb if you switch PME).
Note that your current cut-off's are overly long.

All this probably doesn't matter much for the surface tension.

PS spatially isotropic dispersion correction does not work for the surface tension,
you need to determine the dispersion correction to the surface tension analytically.

Berk

From: mstukan at slb.com
To: gmx-users at gromacs.org
Date: Wed, 1 Dec 2010 12:10:23 +0000
Subject: [gmx-users] v-rescale and conserved energy



Dear gromacs users, I am running NVT simulations for polarizable (shell particle) water model  (GROMACS version 4.5.1 and 4.5.3 (issue is same in both)) using v-rescale thermostat. While temperature, potential, kinetic and total energies behave quite well I am failing to obtain constant values for Conserved-Energy. It drifts quite significantly: 1000 kJ/mol/ns with average value 32500 kJ/mol.I have  noticed that this issue has been already addressed: http://oldwww.gromacs.org/pipermail/gmx-users/2009-May/041836.html So I tried to switch from  md (leap-frog)  to md-vv integrator, but then temperature went up enormously  According to http://www.mail-archive.com/gmx-users@gromacs.org/msg32395.htmlthis issue seems also be know , but I have not found if it  was fixed in 4.5.x version? Am I doing something obviously wrong? And should Conserved-Energy drift be a real problem if the purpose of the simulation is calculation of the surface tension?  My .mdp file is shown below. Thank you very much in advance.Mikhail ===============================integrator          =  md or md-vvdt                  =  0.001            ; time stepnsteps              =  20000000                 ; number of stepscomm_mode           =  Linear;Angular           ; Remove center of mass translation and rotationnstcomm             =  1003                     ; reset c.o.m. motionnstxout             =  10000            ; write coordsnstvout             =  10000                  ; write velocitiesnstxtcout           =  1000                  ; write coords to xtc-trajectory filenstlog              =  1000                     ; print to logfilenstlist             =  20                       ; update pairlistns_type             =  grid ;simple             ; pairlist method;================== Polarizable model parameters =======================emtol               =  0.001              ;the convergency criterion for maximum forceniter               =  50                ;maximum number of iterations for the shell particle optimization;=================  SHAKE/SETTLE Specification  ===============================shake_tol            = 0.00001;=================  Cutt off specification =============================pbc                 = xyz               ; periodic boundary conditionsoptimize_fft        =  yes                      ; perform FFT optimization at startcoulombtype         =  PMEewald_geometry      =  3dcrcoulomb            =  1.5 ;1.5                 ; cut-off for coulombrlist               =  1.5 ;1.5                 ; cut-off for nsvdw-type            =  cut-offrvdw                =  1.5 ;1.5                 ; cut-off for vdwdispcorr            =  EnerPres;=======================================================================Tcoupl              =  v-rescale ;        ; temperature couplingtc-grps             =  Systemref_t               =  298.15tau_t               =  0.1Pcoupl              =  no;=======================================================================gen_vel             =  no                       ; generate initial velocitiesgen_temp            =  300                      ; initial temperaturegen_seed            =  -1                       ; random seed ===============================================   
-- 
gmx-users mailing list    gmx-users at gromacs.org
http://lists.gromacs.org/mailman/listinfo/gmx-users
Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
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/Support/Mailing_Lists 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20101201/cc27fdf5/attachment.html>