[gmx-users] Supercritical CO2

[V.V.Chaban]

The point is that any metastable state in numerical simulations
requires fine-tuned Hamiltonians.




On Fri, May 1, 2015 at 2:47 PM, Alex <nedomacho at gmail.com> wrote:
> With opls-aa, incorrect densities and diffusivities were calculated entirely
> away from the critical point for several liquids.
>
> Alex
>
> On May 1, 2015 11:40 AM, "V.V.Chaban" <vvchaban at gmail.com> wrote:
>>
>> This problem is natural. Even small change in (T,P) in the
>> supercritical range results in a huge change of density and, hence,
>> all other properties.
>>
>> You need either to alter (T,P) or to adjust the force field model
>> (both change will be slight, but will bring large changes of density).
>>
>> If you expect any phase separation in these setups, take care of the
>> Gibbs rule.
>>
>>
>>
>>
>> On Thu, Apr 30, 2015 at 6:38 PM, Alex <nedomacho at gmail.com> wrote:
>> > Only other suggestion I have is that 5 ns production runs are probably
>> > not
>> > long enough. Granted, we needed diffusion data, so I believe I had 50 to
>> > 100 ns long simulations, but even for the barostat to settle, it may be
>> > a
>> > good idea to have simulations significantly longer than 5 ns.
>> >
>> > Alex
>> >
>> > On Thu, Apr 30, 2015 at 3:14 PM, Alex <nedomacho at gmail.com> wrote:
>> >
>> >> Unfortunately, I had a pretty terrible experience trying to simulate
>> >> fuels, fuel mixtures, and CO2 dissolved in fuels with Gromacs (using
>> >> OPLS-AA) at high pressures and/or temperatures. It is my understanding
>> >> that
>> >> at the moment this is isn't the area where these forcefields are
>> >> seriously
>> >> tested.
>> >>
>> >> I have no results for other forcefields, but for OPLS-AA it was
>> >> unpublishable and came out as an internal note at the institution. If
>> >> curious, see
>> >> http://nvlpubs.nist.gov/nistpubs/TechnicalNotes/NIST.TN.1805.pdf
>> >> In fact, if you get any results that correspond to experimental data,
>> >> I'll
>> >> really appreciate it if you let me know. :)
>> >>
>> >> Alex
>> >>
>> >> On Thu, Apr 30, 2015 at 2:04 PM, Jarrett Lee Wise <jwise6 at uwyo.edu>
>> >> wrote:
>> >>
>> >>> Hello all-
>> >>>
>> >>> I am trying to run simulations containing supercritical CO2 and oil
>> >>> asphaltenes.  I have tested various force fields (TraPPE flex, TraPPe
>> >>> rigid, Charmm, EPM2, Cygan, and Zhang), but all of my results give
>> >>> either
>> >>> extremely low densities (~200 kg/m^3) or extremely high densities
>> >>> (~1600
>> >>> kg/m^3) when the density should be around 630 kg/m^3.
>> >>>
>> >>>
>> >>> I equilibrated a system of 5000 CO2 molecules with Nose-Hoover and
>> >>> Berendsen for 5 ns, then did a production run using Nose-Hoover and
>> >>> Parrinello-Rahman for another 5 ns using 1 fs time step.
>> >>>
>> >>> I have found that tau_p plays an important factor.  A relaxed tau_p
>> >>> gives
>> >>> low density while a vigorous tau_p gives a high density and crashes
>> >>> using
>> >>> Parrinello-Rahman pressure coupling.
>> >>>
>> >>>
>> >>> Has anyone else encountered similar problems or have any advice?
>> >>>
>> >>> Here is a copy of my MDP file.  Any advice is greatly appreciated.
>> >>>
>> >>> Thanks
>> >>>
>> >>>
>> >>> Jarrett Wise
>> >>> PhD Student
>> >>> Petroleum Engineering
>> >>> University of Wyoming
>> >>>
>> >>>
>> >>> title           = CO2 Density Run
>> >>> integrator      = md
>> >>> nsteps          = 5000000
>> >>> dt              = 0.001
>> >>> nstxout         = 0
>> >>> nstvout         = 0
>> >>> nstfout         = 0
>> >>> nstxtcout       = 20000
>> >>> nstenergy       = 100
>> >>> nstlog          = 1500
>> >>>
>> >>> constraint_algorithm = lincs
>> >>> constraints     = H-Bonds
>> >>> lincs_iter      = 1
>> >>> lincs_order     = 4
>> >>>
>> >>> cutoff-scheme   = Verlet
>> >>> ns_type         = grid
>> >>> nstlist         = 100
>> >>> rlist           = 1.4
>> >>> rcoulomb        = 1.4
>> >>> rvdw            = 1.4
>> >>> nstcalcenergy  = 10
>> >>> comm_mode = linear
>> >>>
>> >>> coulombtype     = PME
>> >>> pme_order       = 4
>> >>> fourierspacing  = 0.16
>> >>>
>> >>> tcoupl          = Nose-Hoover
>> >>> tc-grps         = System
>> >>> tau_t           = 5.0
>> >>> ref_t           = 320
>> >>>
>> >>>
>> >>> pcoupl          =  Parrinello-Rahman
>> >>> pcoupltype      = isotropic
>> >>> tau_p           = 3.0
>> >>> ref_p           = 520.0
>> >>> compressibility = 1.45e-4
>> >>>
>> >>> pbc             = xyz
>> >>> continuation    = yes
>> >>> gen_vel         = no
>> >>> gen_temp       = 320
>> >>>
>> >>>
>> >>> --
>> >>> 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.
>> >>>
>> >>
>> >>
>> > --
>> > 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.
>> --
>> 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.