[Pw_forum] Options for constrained optimizations

[Zimmerman, Jonathan A]

Hi all,

I¹m a bit confused about the various constraint types that one can apply
during optimizations. I¹m trying to perform a generalized stacking fault
calculation wherein atoms in the top- and bottom-most layers will be fixed
in the direction of the fault plane normal (in order to limit relaxation
in this direction) and atoms in the top-half of the system will be
constrained in the shifting direction (a <112> direction for an FCC
lattice). You¹ll note that these two sets intersect and have some atoms in
common.

I¹m trying to use the Œbennett_proj¹ type of constraint for this
calculation, but it¹s not clear to me whether atoms (a) are constrained to
move ONLY along the vector direction specified, or (b) are constrained
FROM moving along the vector direction specified. If (a), is there a way
to restrict motion to a plane rather than a line? Is there a better way to
do this calculation?

Thank you,

Jon Z.
----------------------------------
Jonathan A. Zimmerman
Sandia National Laboratories
E-mail: jzimmer at sandia.gov
------------------------------------------