[gmx-users] md of only water - obtaining different pressure in the x , y and z directions

[André Farias de Moura]

Hi,

pressure is an intrinsically macroscopic property (both in size and time),
which we may compute for such small systems during such short simulations,
but it does not behave ideally under these conditions. Remember that P and
V are thermodynamically coupled variables and that liquid compressibility
is low, so small volume fluctuations arising from pressure coupling to an
external bath will produce wild instantaneous pressure fluctuations.

If pressure matters for you for some reason, systems need to be much larger
and time needs to be much longer. How larger and longer should they be?
Take a look at your pressure fluctuations in a pressure-time plot or its
histogram and notice that it spreads over hundreds or thousands of bars, so
statistically meaningful averages can only be obtained if you manage to
decrease that fluctuations, but fluctuations decrease as 1/SQRT(N), so
increasing your system by a factor of 100 should decrease your pressure
fluctuation by  a factor of 10.

It is hard to afford such large systems, so most people usually do pressure
block averaging over long enough simulations.

I hope it helps

Andre

On Tue, Mar 8, 2016 at 7:19 AM, Diez Fernandez, Amanda <
amanda.diez10 at imperial.ac.uk> wrote:

> Hi,
>
> I am running a simulation which contains only water (using the spc216.gro
> coordinates and TIP3P model) in an NPT ensemble.
>  I get different values for the average pressure in the x y and z
> directions.
> I am running the simulation for up to 10s in a 4nm side cube unit cell.
> I am using a Nose Hoover thermostat and a Parrinello-Rahman barostat
> applied isotropically, with target pressure 1 bar.
>
> The averages after 1ns are:
>
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
>
>
> -------------------------------------------------------------------------------
>
> Temperature                 299.999    0.00098    6.07906 0.00504098  (K)
>
> Pressure                     1.3174      0.066    350.061   -0.14372  (bar)
>
> Box-X                        4.0396    0.00013 0.00874048 -5.7234e-06  (nm)
>
> Box-Y                        4.0396    0.00013 0.00874048 -5.7234e-06  (nm)
>
> Box-Z                        4.0396    0.00013 0.00874048 -5.7234e-06  (nm)
>
> Pres-XX                     1.11668          2     485.28   0.288761  (bar)
>
> Pres-YY                   -0.762405          1    427.866    -6.2632  (bar)
>
> Pres-ZZ                     3.59792        2.4    410.546    5.54328  (bar)
>
>
> The averages over the 10 ns are :
>
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
>
>
> -------------------------------------------------------------------------------
>
> Temperature                     300      8e-05    4.00951 0.000535848  (K)
>
> Pressure                    1.22566     0.0067    275.993 -0.0289005  (bar)
>
> Box-X                       4.03929    5.1e-05 0.00823878 0.000126042  (nm)
>
> Box-Y                       4.03929    5.1e-05 0.00823878 0.000126042  (nm)
>
> Box-Z                       4.03929    5.1e-05 0.00823878 0.000126042  (nm)
>
> Pres-XX                     1.29387       0.73    385.989  0.0898956  (bar)
>
> Pres-YY                     1.35691       0.74    378.841    3.13618  (bar)
>
> Pres-ZZ                     1.02621       0.84    377.022   -3.31278  (bar)
>
>
>
> Why is the average pressure in the in each direction different?
>
> And why is the drift of the XX direction so much lower than in the YY and
> ZZ directions?
>
>
> Many thanks!
>
> --
> 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.
>



-- 
_____________

Prof. Dr. André Farias de Moura
Department of Chemistry
Federal University of São Carlos
São Carlos - Brazil
phone: +55-16-3351-8090