Sutton Ag EAM Example¶
This provides an example of using potable
to tabulate the Ag model given by Sutton and Chen in [1].
Potential Model¶
Where:
- \(E_T\) is the energy of ths system.
- \(r_{ij}\) is the separation between atoms \(i\) and \(j\).
- \(c\) and \(\epsilon\) are adjustable parameters specific to interacting species.
- Inside the square brackets the first term \(V(r_{ij})\) are the pair potentials.
- The second is the many body term: \(c \sum_i \sqrt{\rho_i}\). Where \(\rho_i\) is the electron density.
Pair potential form:¶
Where:
- \(a\) and \(n\) are potential parameters.
This must be multiplied by the \(\epsilon\) term from equation (12) above:
To make things easier later on, this will be re-expressed as:
This will allow this functional form to be written using the provided as.exponential potential-form.
Many body terms¶
Density function:¶
The density function is:
Again to allow the use of the as.exponential potential-form this will be re-written as:
Potable input¶
The potable
input for this model can be downloaded as Ag_sutton.aspot
and will now be described:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | [Tabulation]
target : setfl
#
cutoff_rho : 600
drho : 0.005
#
cutoff : 12.0
dr : 0.001
[EAM-Embed]
Ag : product(as.constant 2.5415e-3, as.sqrt -144.41)
[EAM-Density]
Ag : as.exponential 4681.013008649 -6
[Pair]
Ag-Ag : product(as.constant 2.5415e-3, as.exponential 21911882.787 -12)
|
Notes:
lines 1-8 [Tabulation]:
- lines 4,5: gives the resolution and extent of the function in
[EAM-Embed]
. - lines 7,8: defines resolution and extent of the tables generated for the
[Pair]
and[EAM-Density]
functions.
- lines 4,5: gives the resolution and extent of the function in
lines 10 and 11 [EAM-Embed]:
- Defines the embedding function.
- Note the use of the
product()
potential modifier to multiply the square root embedding function by the value of \(\epsilon\).
lines 13 and 14 [EAM-Density]:
- Describes the density function.
- The value of 4681.013008649 is obtained as \(a^m = 4.09^6\).
lines 16 and 17 [Pair]:
- Defines the pair potential component of the model.
- As above, the
product()
potential modifier has been used to multiply the function by \(\epsilon\). - Here the first parameter to the
as.exponential
form is \(a^n = 4.09^{12}\) = 21911882.78.
Making and testing the tabulation¶
To tabulate the potential download the aspot file
and run it through potable
potable Ag_sutton.aspot Ag_sutton.eam.alloy
A LAMMPS input file is provided to allow you to test the Ag_sutton.eam.alloy
file produced by potable. This input file can be downloaded here: Ag_sutton_fcc.lmpin
and will energy minimize the structure and then perform an NPT MD equilibration at T=300K. Frames will be dumped every 1000 timesteps (1ps) and dumped to a LAMMPS dump file named dump.atom
(this is suitable for visualisation in Ovito).
In terms of the table file the important part of the LAMMPS input is:
pair_style eam/alloy
pair_coeff * * Ag_sutton.eam.alloy Ag
This tells LAMMPS to accept a setfl
formatted file (pair_style eam/alloy
). The Ag
at the end of the pair_coeff
line says that LAMMPS should associate atom type 1 with the Al
species label in the table file Ag_sutton_eam.alloy
.
Placing both the LAMMPS and table file in the same directory run LAMMPS as follows:
mpirun lammps -in Ag_sutton_fcc.lmpin -log Ag_sutton_fcc.lmpout
Footnotes:
[1] | A.P. Sutton, and J. Chen, “Long-range Finnis-Sinclair potentials”, Philos. Mag. Lett. 61 (1990) 139 doi:10.1080/09500839008206493. |