Dear Ehsan,

I just went over you fdf file very quickly and find some points, which may 
cause this problem:
1. you did not define any specific supercell however you turn the cell 
relaxation on. Is it necessary to relax the super cell? even if it is necessary 
you can first relax the atomic coordinate and after some relaxation turn the 
cell relaxation on.
2. kgrid_cutoff = 0 ?? , may be it is better to use monkhorst-Pack scheme to 
generate a suitable kgrid mesh.
3 DM.NumberPulay = 0 ?? may be it is better to use the history of steps to 
reach better convergence.
4. Is 100 Ry Meshcutoff enough for a converged result ? Just test higher 
cutoffs wether it improves the relaxation.

Regards,
Ebrahim

--- Ehsan Shakerzadeh <[email protected]> schrieb am So, 15.5.2011:

Von: Ehsan Shakerzadeh <[email protected]>
Betreff: [SIESTA-L] Any help is really appreciated
An: [email protected]
Datum: Sonntag, 15. Mai, 2011 06:57 Uhr

#yiv2062109214 p {margin:0px;}

Dear Siesta Users,
I want to do a geometry optimization on a boron-nitrogen nanotube. 
I have tried approximately one month and unfortunately, my efforts were not 
successful.
After 1000 CG step the relaxed geometry was not achieved, and the tube 
destroyed during optimization.
I think something is wrong with my input file. The input file is attached. I 
rely appreciated anyone who look at this input file and give me any suggestion.
Thanks a lot in advance
Best regards
E. Shakerzadseh
E-mail: [email protected]
 

Responder a