[gmx-users] Exploding box (in one dimension) Semi-isotropic pressure coupling?

[Sander Pronk]

There *should* be no preference, but I seem to remember that one of the artifacts of using Ewald-style electrostatic treatment (such as PME) is that there may be dipole-coupling between periodic images, especially if one symmetry has already been broken. 
That doesn't mean you shouldn't use PME, though: you're a long way off from the thermodynamic limit, so you can't rely on fluctuations being slow. The situation you're describing will always happen, and with a small number of particles (i.e. less than you can see with the naked eye) it will happen sooner rather than later.

Sander


On Mar 30, 2010, at 18:00 , sapna sarupria wrote:

> Hi Sander
> Thanks for your response. Is there then a preference for z-dimension to expand (other than the way the system is setup in terms of semi-isotropic coupling). Would one see an expansion in any one direction if I had the same system setup with independent pressure coupling for each dimension (anisotropic)? 
> 
> Thanks,
> Sapna
> 
> On Tue, Mar 30, 2010 at 11:53 AM, Sander Pronk <pronk at cbr.su.se> wrote:
> It sounds like the normal thing that would happen if you have a system that has no shear elastic constant, like a fluid. 
> In that case, there are no restoring forces against growth of system size in one coordinate with a concomitant decrease in the other coordinates, so eventually this should happen (unless I misunderstand the nature of your system).
> 
> Sander
> 
> 
> On Mar 30, 2010, at 17:47 , sapna sarupria wrote:
> 
>> Hello all,
>> 
>> I have a simulation running for a hydrate (with CO2) and water in contact with each other. I run the simulation at 280K at which the hydrate melts and finally the system is just liquid water with CO2 in it. However, as the simulation proceeds the box begins to expand in the z-direction and decreases in the x-y direction considerably. Initially the box has 4.8x4.8x9.6 nm^3 dimensions. After the hydrate melts (the volume fluctuates more) and is fine for another 10 ns but after that the dimensions in the x-y direction decrease to 2.5 nm and z-dimension increases to 35 nm. At this point the simulation crashes since my cut-off distance are 1.2 nm. 
>> 
>> I am using semi-isotropic pressure coupling but the pressure in the x-y direction and z-direction are the same (=30.5 bar) and so is the compressibility (=4.5e-5). I am using Parrinello-Rahman pressure coupling and Nose-Hoover thermostat. The CO2 and water molecules are rigid and shake is used to maintain their geometry. The simulation time step is 2 fs. Can someone tell me why such a thing happens? I have pasted my mdp file for completeness. I am using Gromacs  4.0.5 and I run the simulations in parallel using 4 processors. Values related to domain decomposition etc are unchanged and the default values are used.
>> 
>> Any insights will be helpful. Thanks for the help.
>> 
>> MDP File:
>> title               =  CO2 hydrate + water      ; a string
>> cpp                 =  /lib/cpp                 ; c-preprocessor
>> dt                  =  0.002                    ; time step
>> nsteps           =  25000000            ; number of steps
>> nstcomm       =  10                           ; reset c.o.m. motion
>> nstxout          =  20000                    ; write coords
>> nstvout          =  20000                    ; write velocities
>> nstlog            =  25000                    ; print to logfile
>> nstenergy      =  500                   ; print energies
>> xtc_grps        =  OW_HW1_HW2_CO2
>> nstxtcout       =  500
>> nstlist             =  10                       ; update pairlist
>> ns_type          =  grid                     ; pairlist method
>> coulombtype   =  PME
>> rvdw                =  1.2                      ; cut-off for vdw
>> rcoulomb         =  1.2                      ; cut-off for coulomb
>> rlist                  =  1.2                      ; cut-off for coulomb
>> DispCorr          =  EnerPres
>> Tcoupl              =  Nose-Hoover
>> ref_t                 =  300
>> tc-grps             =  System
>> tau_t                 =  0.5
>> Pcoupl              =  Parrinello-Rahman
>> Pcoupltype       =  semiisotropic            ; pressure geometry
>> tau_p               =  1.0   1.0                ; p-coupling time
>> compressibility     =  4.5e-5  4.5e-5           ; compressibility
>> ref_p               =  30.5  30.5               ; ref pressure
>> gen_vel             =  yes                      ; generate initial vel
>> gen_temp            =  200                      ; initial temperature
>> gen_seed            =  372340                   ; random seed
>> constraint_algorithm = shake
>> constraints         =  all-bonds
>> 
>> 
>> 
>> Thanks a lot
>> Sapna
>> 
>> 
> 
> -- 
> 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/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/mailing_lists/users.php

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100330/8b5ca822/attachment.html>