[CPMD-list] Segmentation fault for a molecular dynamics run

[Axel Kohlmeyer]

On Mon, 12 Apr 2004, Dai Ling wrote:

DL> Dear Kohlmeyer:

hello dai ling,

DL>  Thanks for you advice. However, my program still failed after running
DL> MD based on a previous wavefunction optimization. The input and output
DL> file are as shown below. In the output file, the temperature is scaled
DL> as expected, whlie the kinetic, Kohn-sham, Hamiltonian energy and the
DL> "DIS" (what is it?) all show increasing trend.  Any advice?

well, the energies (EKS, ECLASSIC, EHAM) are in fact _decreasing_.
it seems that your starting structure was quite a bit away from 
the equilibrium. the excessive potential energy has been converted
into kinetic energy, which was removed by rescaling the velocities.

unfortunately, during this process you again have gotten away from
the BO-surface. so you need also to occasionally re-optimize the 
wavefunction and remove ('quench') the fictitious kinetic energy 
from the electronic system.

so i suggest, you either do a geometry optimization before you
start the molecular dynamics run, or limit the maximum number of
md steps (MAXSTEPS) to something like 200 and then restart the
md while adding the keyword QUENCH BO (see the manual for details).

'DIS' is the (mean squared?) displacement from the initial positions. 

btw.: your output looks much more reasonable now.

	axel kohlmeyer.


DL>  
DL> Thanks in advance!
DL>  

[...]

-- 


=======================================================================
Dr. Axel Kohlmeyer                        e-mail: axel.kohlmeyer at rub.de
Lehrstuhl fuer Theoretische Chemie          Phone: ++49 (0)234/32-26673
Ruhr-Universitaet Bochum - NC 03/53         Fax:   ++49 (0)234/32-14045
D-44780 Bochum  http://www.theochem.ruhr-uni-bochum.de/~axel.kohlmeyer/
=======================================================================