[molpro-user] Optimization with constraint in z-matrix

Sebastian Marquardt sebastian.marquardt at epfl.ch
Mon Apr 5 11:20:25 BST 2010


Dear Molpro-list,

I wanted to do a geometry optimization with a constraint in a z-matrix.
The molecular architecture should be something like
		3
	1		2
where the bond angle between atom 1, 2 and 3 depends directly on the
optimised distance of atom 1 to 3 r(13) [=r(23)]. The aim was to observe
the movement of 3 over a range of different r(12). Due to the reason,
that the direct insertion of the needed trigonometric function in the
zmatrix resulted in an error (I don't know why, because with standard
operators it works), I tried to switch, after giving a zmatrix as input,
to internal coordinates to use the constraints in optg but it crashed as
well.
In the end I started not to use this constraint and to optimize also
over the angle, although it is in my opinion absolutely senseless, since
if one knows the distances r(12) and r(13)=r(23) the triangle is
completely determined and the angle is fix. These more or less full
active optimizations crashed as well, but as I guess due to other
reasons (error msg: "No gradient found").

Now, could you help me to find out, how it is possible to introduce
constraints in a z-matrix?

Best regards,

Sebastian Marquardt




More information about the Molpro-user mailing list