LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] Silicon bulk Melting Point
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] Silicon bulk Melting Point

From: Ray Shan <rshan@...1795...>
Date: Mon, 24 Jul 2017 21:15:21 +0000

Yes, the “only” proper way of estimating the melting point is by constructing a two-layered system with one layer being amorphous (or liquid) and the other being crystal.  Then run this coexistent structure at different temperatures.  If you see for a certain temperature the solid-liquid line extends into the liquid region then this temperature is below the melting point.  On the contrary if the solid-liquid line extends into the solid region then the temperature is above the m.p.  A series of temperatures would allow you to bracket the m.p.

Estimating m.p. using energy vs. temperature or volume vs. temperature curves are extremely inaccurate - regardless of the potentials used.


On 7/24/17, 1:49 PM, "Axel Kohlmeyer" <akohlmey@...24...> wrote:

>On Mon, Jul 24, 2017 at 4:41 PM, Hao Chen <haochen@...258...> wrote:
>> Thanks, Axel. Yes, you are totally correct. The melting can be dependent on
>> the size and setup of the simulations. So either you can try to change the
>> size or setup. Or you can find some paper which has those data for different
>> potential. Thanks.
>or you determine it "the right way (tm)", e.g. from a set of
>coexistence simulation at various temperature points around the
>expected melting point.
>like i was mentioning before, we have this issue being discussed on
>this mailing list regularly, since there are always people that need
>to be reminded of the implications of the difference in time and
>length scales between computer simulations and experiments.
>Dr. Axel Kohlmeyer  akohlmey@...24...
>College of Science & Technology, Temple University, Philadelphia PA, USA
>International Centre for Theoretical Physics, Trieste. Italy.
>Check out the vibrant tech community on one of the world's most
>engaging tech sites,!
>lammps-users mailing list