ABINIT Discussion Forums

Hello,

I am wondering how to do dihedral angle constraint in abinit ?

Thanks,

Vivek

Dear Vivek,

Well, nothing is coded to do this.
The most powerful constraint specifier that is present in ABINIT is based on the
wtatcon input variable and associate variables :
http://www.abinit.org/documentation/hel ... ml#wtatcon

If there is a reference that describe how you would like to have it done, you might
communicate it ... then we will see whether it is difficult ... and whether somebody
wants to do it ...

Best,
X

The best way of doing this would be a simple constraint on a primitive internal coordinate (dihedral) in the delocalized internals formalism:

choose as variables the primitive internal you are interested in plus all the others, but reduced to delocalized internals. In this way you have one variable to fix (like with iatfix), but you still have to make sure the rest of the variables are properly orthogonal.

The article by Anselm Chemical Physics Letters 335 (2001) 321-326 says they did this for cartesian coordinates, but give precious few details on how. The chemists must know.

This is unfortunately not coded yet, and the delocalized internals are not in production yet (at all - the relaxation is much slower than bfgs so useless).

Matthieu