[gmx-users] NPT - density off

[Mark Abraham]

On 21/08/2011 2:08 PM, Elisabeth wrote:
>
>
> On 20 August 2011 21:20, Mark Abraham <Mark.Abraham at anu.edu.au 
> <mailto:Mark.Abraham at anu.edu.au>> wrote:
>
>     On 20/08/2011 8:02 AM, Elisabeth wrote:
>>     Dear experts,
>>
>>     I am intending to calculate the equilibrium density of a pure
>>     hydrocarbon at different pressures , at 425 K. The normal boiling
>>     point of the liquid is around 350 K. For pressures below 100 bar
>>     densities I am getting from NPT is in accurate. I start form a
>>     structure which is compressed to above 0.6 g/cm3 density but
>>     since temperature is high density goes than to the values below.
>>
>>     P = 50 bar > NPT rho=   0.344   experimental density ~ 0.54
>>     P    100 >     NPT rho=  0.43      experimental density ~ 0.55
>>
>>     densities become more accurate for P >100 bar.
>>
>>     P     500 >   rho=  0.56    experimental density ~ 0.61
>>
>>     I thought maybe you have some idea on how this inaccuracy can be
>>     improved.
>
>     These could reflect limitations in the model you are using (which
>     almost certainly wasn't parametrized upon data like this). Your
>     integration time step is twice as large as is commonly used in the
>     absence of constraints. Also, be sure you are measuring your
>     density only after equilibration, not an average that includes the
>     equilibration period - and describe that method so people here
>     know you're making such sensible measurements.
>
>
> Hello Mark,
>
> Thank you for your comment. Densities are equilibrated. I am not sure 
> what you mean by limitations in the model.

Parameters are not magic truth values. They're empirical values that 
have been observed to replicate some experimental or computational 
results - almost invariably room temperature and atmospheric pressure. 
It is not necessarily known how well they reproduce other conditions - 
increasing temperature and pressure may well lead to those parameters 
modelling some alternative reality. Taking a small test system and 
systematically trialling small increases in temperature and/or pressure 
from "normal" will tend to reveal such trends. Such results may indicate 
the need for different parametrization. This is typically expensive in 
computer and human time.

Mark

> I actually fixed the density and tried NVT using the experimental 
> density at 298 and found a full agreement between vaporization heat of 
> the liquid and that reported in the original OPLS paper. Once I try to 
> obtain the density using NPT this inaccuracy in densities appear. I 
> have been using 1 fs dt with berendsen barostat for 5 ns and collected 
> data over the last 1 ns ( there are about 2500 atoms in the system). 
> So now I am going to follow your instruction and try dt of 0.5 fs for 
> another 5ns along with cpt files from previous barostat trial 
> (berendsen). I will get back to you to report the results soon.
>
> BTW: I am using version 4.5.4
>
> Thank you,
>
>
>
>     Mark
>
>
>>     Thank you in advance for your invaluable help.
>>     Best,
>>
>>
>>     ;        Bonds
>>     constraints             = none
>>     constraint-algorithm = lincs
>>
>>     ;        Run control
>>     integrator          =  md
>>     dt                  =  0.001
>>     nsteps              =  5000000
>>     nstcomm             =  100
>>
>>     ;        Output control
>>     nstenergy           =  100
>>     nstxout             =  100
>>     nstvout             =  0
>>     nstfout             =  0
>>     nstlog              =  1000
>>     nstxtcout          =  1000
>>
>>     ;        Neighbor searching
>>     nstlist             =  10
>>     ns_type             =  grid
>>
>>     ;        Electrostatics/VdW
>>     coulombtype         =  Shift
>>     vdw-type            =  Shift
>>     rcoulomb-switch     =  0
>>     rvdw-switch         =  0.9
>>
>>     ;        Cut-offs
>>     rlist               =  1.25
>>     rcoulomb            =  1.0
>>     rvdw                =  1.0
>>
>>     ;        Temperature coupling
>>     Tcoupl              =  v-rescale
>>     tc-grps             =  System
>>     tau_t               =  0.1
>>     ref_t               =  425
>>
>>     ;        Pressure coupling
>>     Pcoupl              =  berendsen
>>     Pcoupltype          =  isotropic
>>     tau_p               =  1
>>     compressibility     =  3.5e-5
>>     ref_p               =  100
>>
>>     ;        Velocity generation
>>     gen_vel             =  no;yes
>>     gen_temp            =  425
>>     gen_seed            =  173529
>>
>>
>>
>>
>>
>>
>
>
>     --
>     gmx-users mailing list gmx-users at gromacs.org
>     <mailto: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
>     <mailto: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/20110821/872960df/attachment.html>