LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] How to oscillate a wall while thermostatting
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] How to oscillate a wall while thermostatting


From: Axel Kohlmeyer <akohlmey@...24...>
Date: Wed, 1 Nov 2017 10:09:29 -0400

those numbers are useless and irrelevant for the properties you want
to determine.
you have to determine the amount of energy *transferred* and in the
case of the moving wall, that requires properly dealing with the
center-of-mass bias.
it looks as if you first need to have a deep peek into a text book on
statistical thermodynamics and gain a better understanding what you
are doing.

axel.

On Wed, Nov 1, 2017 at 9:54 AM, saeed alborzi <alborzi.saeed6@...24...> wrote:
> Dear Professor Kohlmeyer
>
> As you said , I compared the energy and velocity order of magnitude of the
> wall with and without oscillation. With a 2 angstrom amplitude and 2 fs
> period and no thermostatting, the maximum wall velocity is about 600 m/s and
> kinetic energy is about 23000 kcal/mol . With a 300K nvt thermostat and no
> oscillation, maximum velocity of 700 m/s and kinetic energy of 7000 kcal/mol
> is obtained. I think the values are not neglectable in camparison to each
> other. therefore I think I need to consider both fixes
>
> On Wed, Nov 1, 2017 at 3:27 PM, saeed alborzi <alborzi.saeed6@...24...>
> wrote:
>>
>> let me refer to an article titled "Static and dynamic behavior of water
>> droplet on solid surfaces with pillar-type nanostructures from molecular
>> dynamics simulation"
>>
>>  at : http://www.sciencedirect.com/science/article/pii/S0017931014007455
>>
>> this is similar to my case except that I want to add heat transfer to the
>> simulation.
>>
>> I was thinking that if it was possible to add some amount to the surface
>> atoms' velocity as oscillating velocity while it is not rescaled by fix nvt
>> , it would be easy to implement. but the only problem is that the "velocity
>> set v_vx v_vy v_vz sum yes" command does not take a time-dependent value.
>>
>> I apologize if I could not explain my meaning clearly and strongly
>> appreciate your help.
>>
>> On Wed, Nov 1, 2017 at 3:08 PM, Axel Kohlmeyer <akohlmey@...24...> wrote:
>>>
>>> On Wed, Nov 1, 2017 at 7:31 AM, saeed alborzi <alborzi.saeed6@...24...>
>>> wrote:
>>> > it was just an example, not the real case.
>>> > the effect of frequency value will be investigated
>>>
>>> if the frequency is (much?) lower, then - as i have already pointed
>>> out - there is no need to do both at the same time, as the time scales
>>> are sufficiently decoupled.
>>>
>>> what you are asking is a technically complex and difficult process to
>>> implement correctly.
>>> i am not going to spend any more time and effort on this until i am
>>> convinced that the outcome is physically meaningful.
>>>
>>> axel.
>>>
>>>
>>> >
>>> >> with a 500GHz motion?? how is that possible?
>>>
>>>
>>>
>>>
>>> --
>>> Dr. Axel Kohlmeyer  akohlmey@...24...  http://goo.gl/1wk0
>>> College of Science & Technology, Temple University, Philadelphia PA, USA
>>> International Centre for Theoretical Physics, Trieste. Italy.
>>
>>
>



-- 
Dr. Axel Kohlmeyer  akohlmey@...24...  http://goo.gl/1wk0
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.