[Pw_forum] Cutoff and K-grid convergence

[Francesco Macheda]

Hi everybody,
I have a question about parameter convergence in SCF calculation. Before
getting SCF results, I usually perform a total energy convergence on cutoff
(keeping K-grid fixed) and on K-grid (keeping cutoff fixed) [incidentally,
I know that total energy is not the best quantity to be brought to
convergence]. As at the moment I'm working on a pretty big system (40 and
even more atom per cell) this convergence takes really long time to be
done,  I'd like to have a rough estimate on convergence threshold.

The question is : suppose that I perform an SCF putting cutoff=a and
K-grid=bxbxb. Suppose that if I repeat the calculation with cutoff=a+1 and
K-grid=bxbxb I get an energy discrepance of 10^-{c} . Can I say something
(even an upper limit) on the discrepance that I would get between
calculation with cutoff=a and K-grid=bxbxb and  cutoff=a and
K-grid=(b+1)x(b+1)x(b+1)?

This would save me a lot of time.

Thank you in advance,
Francesco Macheda
Graduand at Pisa University, Italy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160418/532d8d18/attachment.html>