[gmx-users] Negative pressure in an interface NVT simulation

[Surya Prakash Tiwari]

Thanks a lot David. The paper you mentioned was very helpful.

I have figured out the reason of me getting negative pressure in interface
simulations. This might be helpful to other researchers.

I was using dispersion correction DispCorr = EnerPres in my mdp file. When
this term is on for pressure, Gromacs adds a constant term P_lr (eq. 4.175,
gmx manual 5.1.3) to the pressure obtained from virial and kinetic energy
(eq. 3.22). The final pressure obtained in this way is not accurate for
vapor liquid interface simulations, because vapor-liquid system is not
homogeneous, and the average dispersion constant calculated <C6> is not
correct. Despite getting wrong pressures, one gets the accurate surface
tension because dispersion corrections added to the pressure tensors (Pxx,
Pyy, Pzz) gets cancelled out while calculating surface tension.

Instead of using cutoff for Lennard-Jones potential, I used Lennard-Jones
PME
​ built in Gromacs and found the correct pressures.​



​Thank you.​


Surya Prakash Tiwari

On Thu, Sep 29, 2016 at 2:16 AM, David van der Spoel <spoel at xray.bmc.uu.se>
wrote:

> On 27/09/16 16:02, Surya Prakash Tiwari wrote:
>
>> Hello again,
>>
>> Can someone take my question. You don't need to fully answer my question.
>> If you could just show me the direction, that would be more than enough.
>>
>> This is as it should be. Your system has a surface tension as you see
> that you can compare to experimental data if you wish. Since you have two
> surfaces you have to divide the number by 2 and correct for units.
> Check J. Chem. Theory Comput. 11 pp. 2938-2944 (2015) for an up to date
> discussion of surface tensions in simulations.
>
>
>
> Thanks in advance.
>>
>> Surya Prakash Tiwari
>>
>> On Wed, Sep 21, 2016 at 12:11 PM, Surya Prakash Tiwari <
>> sptiwari at gmail.com>
>> wrote:
>>
>> Dear Gromacs users,
>>>
>>> I am doing a water liquid-gas interface simulation in NVT ensemble (at
>>> 298
>>> K, 1000 SPC/E, box size is described in the end). The pressure calculated
>>> (after equilibration) using gmx energy shows negative values:
>>>
>>> Energy                      Average   Err.Est.       RMSD  Tot-Drift
>>> ------------------------------------------------------------
>>> -------------------
>>> Pressure                   -151.627       0.67    190.708    2.20173
>>> (bar)
>>> Pres-ZZ                    -31.7913     0.0072    262.697 -0.0293928
>>> (bar)
>>> #Surf*SurfTen               1151.56        6.5    1806.55   -21.4399
>>> (bar
>>> nm)
>>>
>>> Since liquid and gas phases are in equilibrium, shouldn't pressure (at
>>> least Pzz) be equal to the saturation pressure of water. A large negative
>>> value is not making sense. Can someone help me with understanding this.
>>> One
>>> of my colleagues used NAMD, and he said that he is getting the right
>>> pressure for the interface system!
>>>
>>> Though the pressures, I am getting, are negative; calculated surface
>>> tension value matches well with those available in the literature. It
>>> seems
>>> that something is getting cancelled out, and I am getting the correct
>>> surface tension. Am I right?
>>>
>>> I tried to look how the pressure is calculated in Gromacs.
>>> Mathematically,
>>> my pressures are negative because the virial energies are larger in the
>>> case of interface simulation compared to the case without interface,
>>> other
>>> parameters remain almost same.
>>>
>>>
>>> The starting configuration for the NVT interface simulation was obtained
>>> using the following procedure:
>>> 1. 1000 water molecules were equilibrated using NPT ensemble (298 K, 1
>>> bar).
>>> 2. Find the average box lengths using gmx energy tool.
>>> 3. Obtain a gro trajectory file from the xtc/trr file using:
>>> gmx trjconv -f md -s md.tpr -o conf.gro -pbc whole -b 1000
>>> 4. A snapshot configuration from the above gro file was chosen such that
>>> its box lengths matches with the average box lengths obtained in step 2
>>> (to
>>> get the pressure right).
>>> 5. Double the Z length of the gro file obtained in step 4. Use this as
>>> the
>>> starting configuration for NVT interface simulation.
>>>
>>>
>>> I have tried to explain as much as possible to get the help, but if I am
>>> missing some information, please let me know and I will provide that.
>>>
>>>
>>> Thanks a lot. I look forward for some help.
>>>
>>>
>
> --
> David van der Spoel, Ph.D., Professor of Biology
> Dept. of Cell & Molec. Biol., Uppsala University.
> Box 596, 75124 Uppsala, Sweden. Phone:  +46184714205.
> spoel at xray.bmc.uu.se    http://folding.bmc.uu.se
>
> --
> 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.
>