[molpro-user] why casscf not converge in an opt calculation in molpro

[Dr Seth OLSEN]

Hi,

MolPro automatically ramps up the convergence criteria when it sees 
that gradients are going to be calculated.  Just be happy you're not 
optimizing a conical intersection, in my experience that can cause a 
lot of frustration.  (10^-5 gradient and 10^-7 energy, I believe, 
which is 4 orders and 2 orders of magnitude tighter convergence than 
the default).

Cheers,

Seth

ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms

Dr Seth Olsen, PhD
Postdoctoral Fellow, Computational Systems Biology Group
Centre for Computational Molecular Science
Chemistry Building,
The University of Queensland
Qld 4072, Brisbane, Australia

tel (617) 33653732
fax (617) 33654623
email: s.olsen1 at uq.edu.au
Web: www.ccms.uq.edu.au 

ccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccmsccms




----- Original Message -----
From: Ding Xunlei <dxl at ustc.edu.cn>
Date: Thursday, July 7, 2005 7:29 pm
Subject: [molpro-user] why casscf not converge in an opt calculation 
in molpro
> Dear chemistry?
> 
> In molpro calculations with casscf, I found a puzzle.
> If omit the "optg" from the input file below, the casscf 
> calculation will converge.
> If use the "optg", the casscf calculation will not converge.
> If use "frequencies" instead of "optg", the casscf calculation 
> will also not converge.
> Why?
> 
> **** CO2 scan
> r1=1.2 Ang
> geometry
> C
> O1 ,C, r1
> O2 ,C, r1, O1,180
> end
> 
> gprint,basis,orbitals
> basis
> spd,c,vdz;c
> spd,o,vdz;c
> end
> hf
> casscf
> optg
> 
> 
> 
> 
>    Thank you for your help!
> 
> Yours sincerely,                         	
> Ding Xunlei
> 2005-07-07
> ______________________________________________
> 
> 
>