[Pw_forum] Succeeded in geometry optimisation but failed in band calc.

[Hongze Xia]

Dear QE users and admins,

I used QE to optimise the structure of an orthorhombic phase HfN. I used "relax" to minimise the force and stress on the structure. After few days of efforts, I was able to get a satisfied result out of the software however when I tried to calculate the band structure for this optimised unit cell, I got c-band not converged error using "bands". I tried to use "nscf" but the result is the same. Someone told me that for "bands" and "nscf" they are basically the same. I am not that convinced. Could someone explain these two terms please? In the attempt to solve the "c-band not converged" problem I tried to raise the "ecut" and "nbnd". Someone suggested that reducing "mixing_beta" may help but I found no luck in that. I am just so curious that for a "scf" calc. it was well converged and no error occurred but for "nscf" it was not converged which is strange since "nscf" used the same kohn-sham potential (if I was right?).

Thank you so much.

Best Regards,
Hongze Xia
PhD candidate in Photovoltaics Engineering
University of New South Wales
Sydney 2052 Australia

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20130611/8c7cf63f/attachment.html>