MEAM potential

LAMMPS LAMMPS Mailing List Mirror
1

Hi all

I used the new MEAM-lmp. And I used the 2nn meam.
I find that for the single element, The results are the same as the Lee’s paper.
For alloys, the results are different from Doc Lee’s paper.
I just compare the lmp and Lee’s paper (“Modified embedded-atom method interatomic potential for the Fe–Pt alloy system”)
This is comparison:

For Fe3Pt-L12 structure, lmp Lee

C11 275 238
C12 183 184
C44 97 94

The following file is the FePt.meam file

nn2(1,1)=1
Cmax(1,1,1)=2.80
Cmin(1,1,1)=0.36
augt1=0
ialloy=2
mixture_ref_t=0
erose_form=2
attrac(1,1)=0.05
repuls(1,1)=0.05
nn2(2,2)=1
Cmax(2,2,2)=2.80
Cmin(2,2,2)=1.53
attrac(2,2)=0.05
repuls(2,2)=0.05

lattce(1,2) = ‘l12’
Ec(1,2) = 5.86
alpha(1,2) = 5.9717
re(1,2) = 2.7181
attrac(1,2) = 0.05
repuls(1,2) = 0.05
Cmax(1,2,1) = 2.80
Cmax(1,2,2) = 2.80
Cmax(2,2,1) = 2.80
Cmax(1,1,2) = 2.80
Cmin(1,1,2) = 1.53
Cmin(2,2,1) = 0.36
Cmin(2,1,2) = 0.84
Cmin(2,1,1) = 0.84

And the library file for 2nn-meam is

‘Fe’ ‘bcc’ 8. 26 55.847
5.1572 4.15 1.00 1.00 1.00 2.8636 4.29 0.56
1.00 2.60 1.80 -7.20 1. 3
‘Pt’ ‘fcc’ 12. 78 195.09
6.4956 4.92 2.20 6.00 2.20 3.9174 5.77 0.90
1.00 3.94 -2.20 3.84 1 3

So any suggestion?
Thank you very much.

2

This is a Q that Greg will have to answer.

Steve

3

Hi Zhiwei,

There are some differences between the way B.Y. Lee has implemented the MEAM potential for alloys and the default implementation in LAMMPS. However, you should be able to reproduce his implementation by selecting the right parameters. Attached is an example MEAM parameter file that reproduces his results for the Fe-C system with an L12 reference structure; it should be similar to your Fe-Pt case. Pay particular attention to the “erose_form” and “ialloy” parameters; these are definitely needed to reproduce Dr. Lee’s results. Finally, make sure you’re using an up-to-date version of LAMMPS, and please read the pair_meam documentation page carefully to make sure you understand what all of these new parameters are doing.

Regards,

Greg

FeC-para.meam (724 Bytes)