[CPMD-list] (no subject)

[Juerg Hutter]

Hi 

I assume the problem is related to the small cell you are
using. Certainly, the points raised by the others are
also important, but I think that your problem is directly
related to a NOT converged real space Ewald sum.
The default in CPMD is to sum up the real space
contributions of the Ewald sum only within the
original supercell. If your cell is too small and you
change the cell size you will see contributions
of this. Check this assumption by running the
simulations with 

TESR
4

Of course to get "good" results with small cells
you should also have converged the cutoff (especially
if you change the cell size) and k-points are important
also. In addition I advice everybody to use more
than the default value for the spline interpolation 
points for the pseudopotentials.

SPLINE POINTS
3000

is a good start.

Juerg

----------------------------------------------------------
Juerg Hutter                   Phone : ++41 1 635 4491
Physical Chemistry Institute   FAX   : ++41 1 635 6838
University of Zurich           E-mail: hutter at pci.unizh.ch
Winterthurerstrasse 190
CH-8057 Zurich, Switzerland
----------------------------------------------------------


On Wed, 7 Aug 2002, Dr.Mehmet Cakmak wrote:

> Dear All,
> 
> I am a bit puzzle to calculate lattice parameter of Si.
> Please see the below table. When I take bigger lattice
> parameter, the energy is getting smaller and smaller. 
> There should be turning point which is going to be 
> theoretical lattice parameter of Si. So any idea what could
> be the reason? or do i know something wrong?
> 
>  E_tot       a_lat
> -------      -----
> -7.14         5.0      
> -7.31         5.1
> -7.36         5.2
> -7.39         5.3
> -7.43         5.4
> -7.72         5.41    
> -7.73         5.42
> -7.74         5.43
> -7.74         5.44
> -7.75         5.45
> -7.78         5.5
> -7.83         5.6
> -7.87         5.7
> 
> 	-mehmet
> _______________________________________________
> CPMD-list mailing list
> CPMD-list at cpmd.org
> http://www.cpmd.org/mailman/listinfo/cpmd-list
>