[lammps-users] Thermal conductivity computations

Sreekant,

Why you do not have a linear temperature profile? How many time steps are you doing during NVE? What is the size of your molecular domain?

Please copy your scripts.

Regards, Javier


Javier,
I attached the scripts to my previous email - please see them attached to this email also.
I have also attached an Excel file which shows the temperature profiles. Using a time step of 0.001 picosecond, I have run almost 500000 timesteps during the NVE run. Also, the domain size is about 27.1 by 27.2 by 81 Angstroms, with the temperature gradient computed along the 81 Angstrom direction.

Thanks,
Sreekant

— On Wed, 3/4/09, Javier Goicochea <goicox@…24…> wrote:



> From: Javier Goicochea <goicox@…24…>
> Subject: Re: [lammps-users] Thermal conductivity computations
> To: [email protected]
> Date: Wednesday, March 4, 2009, 5:54 AM
>
> Sreekant,
>
> Why you do not have a linear temperature profile? How many time steps are you doing during NVE? What is the size of your molecular domain?
>
> Please copy your scripts.
>
> Regards, Javier
>
> <br>> ------------------------------------------------------------------------------<br>> Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA<br>> -OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise<br>> -Strategies to boost innovation and cut costs with open source participation<br>> -Receive a $600 discount off the registration fee with the source code: SFAD<br>> http://p.sf.net/sfu/XcvMzF8H<br>>
>
> <br>> _______________________________________________<br>> lammps-users mailing list<br>> [email protected]<br>> https://lists.sourceforge.net/lists/listinfo/lammps-users<br>> <br>>

|

NVTstep.input (607 Bytes)

NVEstep.input (505 Bytes)

silicon-tempprofile-run1.xls (331 KB)

Hi Sreekant,

I think your domain size is too small and you’re introducing size effects in your calculations. The mean free path of phonons in Si is ~ 250 to 300 nm so in your domain you actually have reflections of phonons at the boundaries, which can explain a much lower thermal conductivity.

Jaime

Jaime,
I have done some Boltzmann-equation based modeling in the past. The conclusions about mean free path really seems to depend on the type of phonon transport model we want to go with (i.e. simple gray model with single phonon group velocity and relaxation time, or more rigorous models accounting for phonon dispersion and polarization, and frequency-dependent relaxation times). That said, I will try larger domain sizes just to be sure there are no size effects. Javier was also suggesting running all simulations for longer time durations.

Thanks,
Sreekant