[Wien] non-rectangular cell

[L. D. Marks]

Oddly enough, NEWT is "working" in a sense, but not really solving the
problem you asked; PORT should work in principle, although it might not
work.

What NEWT is doing is minimizing the forces for fixed atoms 1, 14-17,
tending to move along the steepest descent direction (with a bit of
conjugant gradient-like component mixed in). It's not smart enough to know
that the energy can go up because by minimizing the forces on atoms 2-13
it is creating high-energy configurations (density etc) on atoms 1, 14-17.
What you really have is a problem where the off-diagonal terms of the
Hessian are large. PORT generates an estimate of these, so should be able
to handle it although there is a chance it will want to move atoms 1,
14-17 (you have to try).

The "correct" way to minimize your problem is to apply contraints to atoms
1, 14-17. Constrained optimization is a whole different ball game.

On Sun, 20 Jun 2004, Mahbube Hortamani wrote:

>   We calculated the same system in rectangular cell with the same R0, RMT, RKmax,
> Kmesh, which cutoff of  RKmax and Kmesh were
>   checked for the best value in rectangular cell.
>   For pure Si,  we chose the RMT=2.0 and it is enough also from result of calculation
> in the rectangular cell, we found  2.1  is enough. You
>   should know this system in rectangular cell  (with the same input) has not any
> problem.
>
> .....
>   thank you for your attention,
>   Mahbube.
>
>   Peter Blaha wrote:
>
>     I just noticed that your R0 for Mn is fairly large and not the WIEN2k
>     default. maybe this results in wrong forces, where zero force does not
>     correspond to E-min.
>     Another point: Mn as 3d element should have a larger sphere than Si
>     Just another silly check: Are you sure you converged with -fc and have the
>     valence forces in case.scf_mini.
>
>

-----------------------------------------------
Laurence Marks
Department of Materials Science and Engineering
MSE Rm 2036 Cook Hall
2225 N Campus Drive
Northwestern University
Evanston, IL 60201, USA
Tel: (847) 491-3996 Fax: (847) 491-7820
mailto:L-marks at northwestern.edu
http://www.numis.northwestern.edu
-----------------------------------------------