[Pw_forum] Energy keeps going down

[david furman]

Hi all,

I sucsessfully calculated equation of state for CaF2 by varying celldm(1)
in a loop.
then fitting the EV data to murnaghan EOS to get a parabola like plot.

Now I try to calculate EV data for CaTiO3 (Pnma space group) by the same
method.
But, no matter how large the celldm(1) is (keeping the b/a and c/a fixed),
the energy still goes down instead of going up, with no parabola like
behavior.
I assume this is wrong, because when distances between atoms are large
(expansion of unit cell), this state should be higher
in energy than the equilibrium volume state, just like when compressing the
unitcell the energy goes up because of repulsion.

Can anybody tell me why is this happening ? Could it be a bug ?

P.S
The pseudopotentials i'm using are all PBE and verified with other systems
so they should be OK.
There are 20 atoms in the primitive unit cell. I'm using a 8x8x8 k-point
grid with 180 Ry cutoff for wavefunction.
I've opened the structure in Xcrysden and it looks OK, no weird things.



-- 

*David Furman*
*
*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20121031/e22cdd4e/attachment.html>