The ABINIT Users Mailing List ( CLOSED )

[Prasenjit Sen <prasen@mri.ernet.in>]

Dear Joydeep,
     You cannot read in wavefunction by changing ecut as your
basis set (no. of planewaves) has changed. Instead, you can try
reading in the charge density from a previous calulation with smaller
ecut, and chain your alculations that way. Try looking at the input
variable getden.
     Hope this helps.

   With best wishes,
   Prasenjit Sen

 =======================================================================
  Harish-Chandra Research Institute     Phone: +91-532-2274307
  Chhatnag Road, Jhunsi
  Allahabad, UP 211019
       INDIA                            Fax:   +91-532-2568036
 =======================================================================

On Tue, 18 Jan 2005 joydeep@jncasr.ac.in wrote:

> Dear ABINITians,
>
> We are trying to do a self-consistent(iscf 5) calculation
> for a slab of
> acell 16.82045893906953590197  5.1502616       68.0301675998
> containing Ti Mo Al Au  layers (total of 23 atoms) occupying
> roughly bottom 1/3 in the z direction of the unit cell.
> .
> First we tried with ecut  40,tsmear 0.0015, ixc 11 (as we are using
> gga pspots), diemac 10000.0, diemix    0.4.
> But this  result in a very slow convergence and prolonged fluctuation
> of energy within a small region.
>
> As a way out we thought of doing a low ecut (10 or 5) calculation
> with a smaller unit cell (reducing z to 48 bohr) and
> then use the _WFK file as an input to a next calculation
> with ecut 40 and the original acell.
> The low ecut calculations do converge.
> But it seems  it is not possible to use this WFK file (using getwfk -1
> or irdwfk 1) for the second calculation because of the change in ecut.
>
> Is there any special option to facilitate this ecut switching ?
>
> We would greatly appreciate any suggestions that might help.
>
> Thanks in advance.
> Joydeep
>
>