[gmx-users] g_dipole: dipole moment autocorrelation function

[David van der Spoel]

On 2011-03-31 09.25, Dommert Florian wrote:
> On Wed, 2011-03-30 at 20:23 +0200, David van der Spoel wrote:
>> On 2011-03-30 20.16, Nilesh Dhumal wrote:
>>> Thanks.
>>> How can I take average.
>> summing up and dividing by the number of sims.
>>>
>>> How much long I should run the simulation.
>> until the average converges.
>>>
>>> Nilesh
>>>
>>> On Wed, March 30, 2011 12:59 pm, David van der Spoel wrote:
>>>> On 2011-03-30 18.54, Nilesh Dhumal wrote:
>>>>
>>>>> Hello,
>>>>>
>>>>>
>>>>> I am trying to calculate the dipole moment autocorrelation
>>>>> function for my system (ionic liquid). I am using gromacs 4.0.7 version.
>
> Hello,
>
> I think you will get into some trouble when you try to calculate an
> autocorrelation function of property that is not continuous in time. In
> an ionic liquid you have contributions of the molecular dipole and
> contribution of the charged molecules to the dipole. The latter part is
> unfortunately not continuous due to jumping of the molecules over the
> PBCs. It depends what kind of property you want to calculate, but we
> experienced that simulations in the range of 50-100ns are required to
> reliably determine properties like the dielectric constant. The next
> point is that you have to save your coordinates and velocities quite
> often ( every 0.02ps we used)
>
> Check corresponding literature for further advice.

Thanks Florian, for pointing that out. I forgot about it in my previous 
mail. In order to compute the dielectric constant one could in principle 
apply constant electric fields at different strengths and then 
extrapolate to zero. Don't know whether that has been done.
>
> Cheers,
>
> Flo
>>>>>
>>>>>
>>>>> I run the simulation for 4 ns.  I run the following command to
>>>>> calculate the dipole moment autocorrelation function
>>>>>
>>>>> g_dipoles -f water.trr -s water.tpr -corr total -c
>>>>>
>>>>> The function is not geting converge to zero.
>>>>>
>>>>>
>>>>> I want to use this data for calculation of power spectra by fourier
>>>>> transfom of dipole moment autocorrelation function.
>>>>>
>>>>> Can you tell why its not geting converge to zero?
>>>>>
>>>>>
>>>> You have to simulate at least a few 10s of ns for such slow liquids to
>>>> converge. Alternatively you can average over many independent simulations
>>>> (10s).
>>>>
>>>>>
>>>>> Thanks
>>>>>
>>>>>
>>>>> Nilesh
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> 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 --
>>>> gmx-users mailing list    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. Can't post? Read
>>>> http://www.gromacs.org/Support/Mailing_Lists
>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>> --
>> 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
>
>


-- 
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