Question about using 2 different Tersoff files

Dear all,

I have an error message when using two different tersoff files.

Combine eam , tersoff and remaining are coupled with LJ interactions.

I wrote the script :

group Fe type 1

group Mo type 2

group Cr type 3

group Y type 4

group C type 5

group B type 6

group Si type 7

pair_style hybrid/overlay tersoff eam/alloy lj/cut 8.0

pair_coeff * * tersoff Si.tersoff NULL NULL NULL NULL NULL NULL Si

pair_coeff * * tersoff BNC.tersoff NULL NULL NULL NULL C B NULL

pair_coeff * * eam/alloy FeMoCr.set Fe Mo Cr NULL NULL NULL NULL

pair_coeff 1 4 lj/cut 0.3970 2.6955 # Fe-Y

pair_coeff 1 5 lj/cut 0.1127 2.8605 # Fe-C

pair_coeff 1 6 lj/cut 1.2574 1.8355 # Fe-B

pair_coeff 1 7 lj/cut 0.0960 3.0735 # Fe-Si

pair_coeff 2 4 lj/cut 0.5006 2.8105 # Mo-Y

pair_coeff 2 5 lj/cut 0.1421 2.9755 # Mo-C

pair_coeff 2 6 lj/cut 1.5856 1.9505 # Mo-B

pair_coeff 2 7 lj/cut 0.1211 3.1885 # Mo-Si

pair_coeff 3 4 lj/cut 0.3874 2.7030 # Cr-Y

pair_coeff 3 5 lj/cut 0.1100 2.8680 # Cr-C

pair_coeff 3 6 lj/cut 1.2272 1.8430 # Cr-B

pair_coeff 3 7 lj/cut 0.0937 3.0810 # Cr-Si

pair_coeff 4 4 lj/cut 0.2990 3.0700 # Y-Y

pair_coeff 4 5 lj/cut 0.0849 3.2350 # Y-B

pair_coeff 4 6 lj/cut 0.9471 2.2100 # Y-C

pair_coeff 4 7 lj/cut 0.0723 3.4480 # Y-Si

pair_coeff 5 7 lj/cut 0.0205 3.6130 # C-Si

pair_coeff 6 7 lj/cut 0.2291 2.5880 # B-Si

I got the following error message:

All pair_coeff are not set (…/pair-tersoff.cpp:362)

I understand there is an error in writing pair_coeff of tersoff potential ??

If I remove the Si tersoff pair_coeff and I replace it by a LJ 7 7 pair_coeff command for Si-Si interaction instead, the program is running.

Is there conflict in hybrid/overlay of 2 tersoff pair_coeff command ?

Could you help ? Thanks a lot

Regards

Pascal

Dear all,

I have an error message when using two different tersoff files.

Combine eam , tersoff and remaining are coupled with LJ interactions.

[...]

pair_style hybrid/overlay tersoff eam/alloy lj/cut 8.0

pair_coeff * * tersoff Si.tersoff NULL NULL NULL NULL NULL NULL Si

pair_coeff * * tersoff BNC.tersoff NULL NULL NULL NULL C B NULL

pair_coeff * * eam/alloy FeMoCr.set Fe Mo Cr NULL NULL NULL NULL

pair_coeff 1 4 lj/cut 0.3970 2.6955 # Fe-Y

[...]

I got the following error message:

All pair_coeff are not set (../pair-tersoff.cpp:362)

that is the expected behavior.

I understand there is an error in writing pair_coeff of tersoff potential ??

yes, you have twice a statement of the kind:

pair_coeff * * tersoff ...

the second of those will wipe out the first. in your pair_style
command line you have only specified one instance of tersoff.

If I remove the Si tersoff pair_coeff and I replace it by a LJ 7 7
pair_coeff command for Si-Si interaction instead, the program is running.

Is there conflict in hybrid/overlay of 2 tersoff pair_coeff command ?

first of all, i don't understand why you need to use hybrid/overlay
and cannot use plain hybrid itself. using hybrid/overlay for the kind
of setup you have will just increase the chances of double counting
some interaction and thus having a bogus simulation. (gluing together
three manybody potentials is going to raise some eyebrows anyway).

if you need two instances of tersoff in your force field, you need to
specify tersoff *twice* in the pair_style line and then refer to them
as "tersoff 1" and "tersoff 2" in the pair_coeff statements. please
see the pair hybrid documentation for details. it is all explained.

axel.

Thanks a lot for your explanations.
Regards
Pascal
-----Message d'origine-----