Greetings, class of 22.53

- Stress, strain, elastic constants
- Thermodynamic derivatives (heat capacity, thermal expansion...)
- Microscopic particle representations
- Micro-canonical ensemble fluctuation formulas

The new MD code (version 1.1) for Argon features

- isothermal elastic constants
- thermal expansion coefficient
- heat capacities Cv, Cp

For details on thermal conductivity calculations read this paper (text and figures) and this manuscript.

To install the program, download 3D.tar to your home directory and then

- athena% tar xvf 3D.tar
- athena% cd 3D
- athena% make cleanall; make

- athena% ./argon < con

Details include:

After making sure the

- athena% cp con_liquid_short con
- athena% make run

What the above short run cannot provide are the transport coefficients: thermal conductivity and shear viscosity, which can only be accurately calculated with longer runs (1-2 hrs). If you could access private machines, try the following (

- athena% cp con_liquid_long con
- athena% make longrun &

- athena% top or
- athena% ps -e

You do not have to fully understand these files. The final results
will be clearly presented in *temp.out*. But in case you are
interested, look up the thermal conductivity papers referenced above
and play with the Matlab programs *heat_spectra.m* and *heat_corr.m*, which plot out the
data.

Our results compare well with the published liquid simulation results
of Ermakova *et al, J. Chem. Phys.* 102, 4942 (1995), using the
Woon's potential. The slight differences should be due to details in
implementation, such as how the potential cutoff is treated. In *argon*, for instance, we do not carry
out the analytical integration for the tail contribution since some
properties such as thermal conductivity cannot be similarly treated.

Now, how about some solid-state calculations? After all, the code's original purpose is to study these properties in great accuracy, especially the thermal conductivity. Try a 30K solid (con_solid_NEV) running under NEV mode,

- athena% cp con_solid_NEV con
- athena% make run

A general Parrinello-Rahman scheme not only relaxes the volume, but
also the shape of the simulation box, if a non-hydrostatic constant
external stress is applied. Our code is able to do that after a tiny
revision, but is not implemented here for aesthetic reasons, since
under non-hydrostatic *constant* external stress one cannot
define a thermodynamic potential for the system as in the hydrostatic
case, and there is no Hamiltonian dynamics with conservation laws.

Email: lij@matsceng.ohio-state.edu. This page has been accessed at least times since Oct.6, 1998.