[Pw_forum] scan the lattice constant-seems not a closed question

[W. YU]

Dear PW users,

I have read the topic about scanning lattice constant
and found some explanations about the problem met by
XunLei. The answer by Nicola is the most detailed one.
But new question arise from here. In the answer,
Nicola said:

== You see in fact your energy drop going to the
left.==

But it seems to me the energy drop going to the right.


And furthermore, the example figures given by Nicola
about the convergence test of Diamond and Graphite a
bit puzzled me. As is put by Nicola and basically I
believe so, more complet basis set will give lower
energy. But in the example figures, especially the
Diamond case, the energies obtained at the same k
points seem to go higher with the increasing cutoffs. 

Also in my personal experience of convergence test.
There are situations when I fixed a certain cutoff
energy and keep increasing the number of k points, the
energy some times will go down first and go up later.
I couldn't figure out why.

I hope I am not misunderstanding something here and if
I'm right, could anybody give some further comments.

Thanks a lot.

W. YU 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com