On 05/17/2011 02:36 PM, RCP wrote:
Hongyi Zhao wrote:
...
1- As for the total energy, should we use the final energy for E_KS,
i.e., something like the following one in the ouput file of siesta:
siesta: E_KS(eV) = -2456.6000
Yes. Moreover, there should be a further line "total energy" or so, that
should match
the above.
According to your above description, I find the following section in the
output.fdf file:
------------
siesta: Final energy (eV):
siesta: Band Struct. = -996.193016
siesta: Kinetic = 1603.368791
siesta: Hartree = 5297.075816
siesta: Ext. field = 0.000000
siesta: Exch.-corr. = -724.052081
siesta: Ion-electron = -12251.099159
siesta: Ion-ion = 3618.106609
siesta: Ekinion = 0.000000
siesta: Total = -2456.600024
------------
We can note that there is slightly different between the Total energy
(-2456.600024) and the E_KS ( -2456.6000 ). Any hints?
Furthermore, does the final E_KS is the same to the final Total energy?
I mean, should I use some corrections in order to obtain the total
energy for the computations of strain energy?
The above is THE ENERGY for whatever conditions you impose.
Do you mean the one appearing in the Final energy section?
2- In order to obtain the total energy at a specific strain, after I
strained the system, then should I do the full relaxation with
MD.VariableCell = False?
Yes, safest. In any case you can judge by yourself if relaxations beyond
the 1st SCF
cycle matter for your system or not: all needed numbers are there.
Regards.
--
Hongyi Zhao <[email protected]>
Institute of Semiconductors, Chinese Academy of Sciences
GnuPG DSA: 0xD108493
