Problems with Constraint function.

[DaniBajac]

Hello all. I'm trying to optimize the geometry of the H2O2 molecule freezing the dihedral angle, using the .CONSTRAINT keyword, but don't seems to work very well because, after I done the optimization with the dihedral angle in 90 degress, I calculate their redundant coordinates with the .FINDRE keyword and the angle changes! I attach the outputs of the calculation of the geometry and the .FINDRE calculation.

Thank you all!

[hjaaj]

You have uploaded the .FINDRE calculation with the converged geometry. Maybe redundant coordinate no. 12 is not the same in the initial geometry and the final geometry? Did you do the .FINDRE with the initial geometry and used that to select red. coord. no. 12?

Another thing, ZMAT is always read in Ångström, also if ANGSTROM is not specified. This is in contrast to the input for xyz cordinates, which are read in a.u. unless ANGSTROM is specified. Your initial bond distances are therefore only approx. half of what you want, and the values in the converged geometry (most evident in the O-O distance).