[QE-developers] Constants in reciprocal space Hartree calculation

[Aaron Mahler]

Dear all,
I'm working on implementing this scheme:
https://arxiv.org/pdf/1707.00856.pdf in QE and need to evaluate Hartree
energy integrals for individual orbital densities. As such I have a
question regarding evaluating the Hartree energy in reciprocal space. At
the of the calculation in v_h in the PW/src/v_of_rho.f90 file, ehart is
multiplied by e2 * fpi / tpiba2.
https://gitlab.com/QEF/q-e/blob/develop/PW/src/v_of_rho.f90#L623
I assume the 1/tpiba2 term is to scale the G^2 terms in the calculation.
I'm unsure what the e2 * fpi is for though. I'm also unsure about right
after that if it's gamma_only there's no factor of 0.5. Lastly I don't
understand why does e2 = 2 if it's the elementary charge squared in a.u.?
https://gitlab.com/QEF/q-e/blob/develop/Modules/constants.f90#L105
Any insights would be very helpful.

Cheers,
Aaron Mahler
Duke University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20180627/6f62edab/attachment.html>