Negative Lambda 3 for Tersoff/ZBL

Dear users,

we recently encountered a problem while using a Tersoff/ZBL style interatomic potential
for the Fe-Cr-C system with the parameters from Henriksson et al.
[J. Phys.: Condens. Matter 25 (2013) 445401] which has a negative value for
lambda 3 (alpha(Cr,C,Cr)=-1.752 in their paper).

The input parameter validity check in pair_tersoff_zbl.cpp on line 188 and (duplicate) on line 194
forbids negative values for lambda 3. The obvious workaround we did was to remove the
restriction to positive lambda 3 values.

Nonetheless we would like to know whether there could be a reason for the restriction to positive values,
or if this fix can be included in the public lammps version, enabling the usage of the Tersoff/ZBL parameters of Henriksson.

Thanks in advance and best regards,
Dennis Rapp and Peter Binkele

Institut für Materialprüfung, Werkstoffkunde und Festigkeitslehre (IMWF)
Universität Stuttgart
Pfaffenwaldring 32
70569 Stuttgart
Tel: +49 (0)711 685 62586
Fax: +49 (0)711 685 62635
E-Mail: [email protected]
Web: www.imwf.uni-stuttgart.de
             www.simtech.uni-stuttgart.de

Dear users,

we recently encountered a problem while using a Tersoff/ZBL style
interatomic potential
for the Fe-Cr-C system with the parameters from Henriksson et al.
[J. Phys.: Condens. Matter 25 (2013) 445401] which has a negative value for
lambda 3 (alpha(Cr,C,Cr)=-1.752 in their paper).

The input parameter validity check in pair_tersoff_zbl.cpp on line 188 and
(duplicate) on line 194
forbids negative values for lambda 3. The obvious workaround we did was to
remove the
restriction to positive lambda 3 values.

Nonetheless we would like to know whether there could be a reason for the
restriction to positive values,
or if this fix can be included in the public lammps version, enabling the
usage of the Tersoff/ZBL parameters of Henriksson.

​thanks for the report. this is indeed a bug. according to aidan thompson
at sandia, it looks as if a bugfix to the (plain) tersoff pair style was
not propagated to several tersoff variants.
we have updated the code and submitted a pull request to the LAMMPS project
on github: https://github.com/lammps/lammps/pull/794​
​this should be included in the ne​xt patch release of LAMMPS.

best regards,
   aidan and axel.