LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] Fluctuation in calculating energy and specific heat
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] Fluctuation in calculating energy and specific heat

From: Axel Kohlmeyer <akohlmey@...24...>
Date: Mon, 23 Apr 2018 06:43:30 -0400

On Sun, Apr 22, 2018 at 11:28 PM, Muhammad Saeed
<saeedphysics96@...24...> wrote:
> Dear LAMMPS users,
> I am working on "Thermal properties of Self-Assembly monolayers at the
> liquid-solid interface" and want to calculate adsorption energy, enthalpy
> and specific heat with NVT ensembles using LAMMPS. But high fluctuations
> occurs in the temperature (+- 30) and energy of the system during the
> simulations.

...and what version of LAMMPS are you using? and on what platform are
you running? the mailing list guidelines state, that those are crucial
pieces of information.

> Can someone please give me suggestions how to minimize these fluctuations?

some fluctuations are to be expected and depend on the system size.
there are multiple discussions on this subject in the mailing list
archives, and it should also be explained in any good text book on MD,
that for a system in equilibrium, the average over time is equivalent
to the average over the system.

> The input script for adsorption energy is

now, there are multiple problems with this script. please see the
in-line comments below.

> units           real
> atom_style      charge
> variable        t      equal    278
> variable        dt     equal    0.01
> variable        system string   combined
> read_data
> pair_style   reax/c NULL
> pair_coeff    * * ./ffield.reax.cho C H O

as it says in the header of the ffield.reax.cho file, this is a force
field for combustion, so it is a *very* bad choice for your project.

> velocity        all create \$t 87287
> neighbor        1.5 bin
> neigh_modify    once yes

not updating the neighbor lists for a liquid system is also a *very* bad idea.

> fix 1 all box/relax iso 0.0 vmax 0.001

fix box/relax has *no* effect during dynamics runs.

> fix 2 all qeq/reax 1 0.0 10.0 1e-6 param.qeq
> fix 3 all nvt temp \$t \$t 0.03

a choice of 0.03fs for tdamp is a *bad* choice with a time step of 0.01fs.

> timestep        \${dt}

a time step of 0.01fs is rather wasteful for the type of your system
at ambient temperatures. with a proper starting geometry and a
suitable force field, you should be able to run a stable, energy
conserving MD with 0.1fs and possibly even larger (0.2 - 0.25fs).


> thermo          100
> run 500000
> Kind Regards
> Saeed
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites,!
> _______________________________________________
> lammps-users mailing list

Dr. Axel Kohlmeyer  akohlmey@...24...
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.