[Pw_forum] total energy in DFT+U

[Matteo Cococcioni]

Dear Farah,

first of all: total energies are always defined up to a constant. if
anything, you should be looking at energy differences (e.g. between
different crystal structures or magnetic orderings if that applies to your
case) rather than total energies per se. Also, your expectation is not
right: the DFT+U total energy (as usually written) is not variational with
respect to U. That is why (or one of the reasons why) typically you need
linear-response calculations to compute it (PRB 71, 035105) rather than a
trivial minimization of the energy.

Best regards,

Matteo



On Thu, Jun 8, 2017 at 9:34 AM, FARAH MARSUSI <marsusi at aut.ac.ir> wrote:

> Dear all,
>
> I have done DFT+U calculations in a semi-empirical way by increasing U
> parameter until the magnetic moment (M) equal to experimental value was
> obtained along with a phase transition from metal to insulator, as seen in
> experiment. All physical findings are reasonable, except total energies. By
> increasing U and reaching to experimental M, total energy is increasing
> too. I expected the lowest total energy for it!
>
>
> Best regards,
>
> F. Marsusi,
>
> Department of Physics,
> Amirkabir University of Technology.​
>
> --
> This email was Anti Virus checked by  Security Gateway.
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20170608/d08dd163/attachment.html>