[gmx-developers] Wall interaction with tabulated potential

[Christoph Junghans]

2016-02-27 1:12 GMT-07:00 Jakub Krajniak <jakub.krajniak at cs.kuleuven.be>:
> On 2016-02-05 09:57, Sudharsan Pandiyan wrote:
>> Dear Gmx developers,
>>
>> I am trying to simulate a wall option using tabulated potentials
>> option. But it gives different LJ potential values when I compare 12-6
>> potential with the tabulated potential (where the table was generated
>> using same sigma and epsilon values that were used for 12-6 potential).
>>
>> I intend to modify my table potential later so I wanted to make sure
>> that it produces correct result. But my test shows that there is a
>> difference between 12-6 and tabulated potentials for wall interaction.
>> Could you please explain what is the difference and how can I get the
>> correct potential energy for tabulated potentials?
>>
>> PS: In the manual, its written that both 9-3 and 10-4 are integrated
>> over the surface area and 12-6 potential was applied directly with the
>> z-distance. But how the tabulated potential is represented? (section
>> 7.3.20)
>>
>>
>
> Hi,
>
> I've looked into the code and I might find the solution for this problem.
>
> I've run NVE simulation for very simple LJ systems of two particles
> and the single particle with the wall and I compared the energies and the
> trajectories.
>
> Tests:
> 1. two LJ particles at z-distance 1.2 (parametrize LJ interaction):
> -0.235941 kJ/mol
> 2. two LJ particles at z-distance 1.2 (interaction from table6-12.xvg):
> -0.23594 kJ/mol
>
> and wall (after fix):
> 3. particle at distance 1.2 from the wall (wall-type 12-6): -0.235892 kJ/mol
> 4. particle at distance 1.2 from the wall (wall-type table table6-12.xvg
> ): -0.235893 kJ/mol
>
> the data from 10000 steps. The simulation files are here:
> https://www.dropbox.com/s/ryr0x7q974gefu6/experiment.zip?dl=0
>
> I don't know if you are using GitHub pull request so I attached patch to
> this message.
> It's also on github branch:
> https://github.com/MrTheodor/gromacs/tree/fix_wall
Uploaded as: <https://gerrit.gromacs.org/#/c/5684/>

Christoph
>
> Best regards,
> Jakub
>
> --
> Jakub Krajniak
> KU Leuven, Dept. Computer Science lokaal: 01.41
> +32 477 68 84 04 / +32 16 37 39 92
>
>
> Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm for more
> information.
>
> --
> Gromacs Developers mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-developers_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-developers or
> send a mail to gmx-developers-request at gromacs.org.



-- 
Christoph Junghans
Web: http://www.compphys.de