mirror of https://gitlab.com/QEF/q-e.git
Bug fixing: Integrals of atomic wavefunctions (tab_at) should be rescaled
as it is done with other pseudopotential integrals (tab, qrad) when the cell changes (vc-relax) otherwise the projectors in the LDA+U case are wrong. This was wrong already in 3.0 (and probably earlier versions as well). It affects LDA+U + vc-relax (or vc-dynamics). SdG git-svn-id: http://qeforge.qe-forge.org/svn/q-e/trunk/espresso@3397 c92efa57-630b-4861-b058-cf58834340f0
This commit is contained in:
parent
f2dade1657
commit
fac3acb88e
|
@ -20,7 +20,7 @@ subroutine scale_h
|
|||
USE cellmd, ONLY : at_old, omega_old
|
||||
USE gvect, ONLY : g, gg, ngm
|
||||
USE klist, ONLY : xk, wk, nkstot
|
||||
USE us, ONLY : nqxq, nqx, qrad, tab
|
||||
USE us, ONLY : nqxq, nqx, qrad, tab, tab_at
|
||||
!
|
||||
implicit none
|
||||
!
|
||||
|
@ -50,6 +50,7 @@ subroutine scale_h
|
|||
!
|
||||
tab(:,:,:) = tab(:,:,:) * sqrt (omega_old/omega)
|
||||
qrad(:,:,:,:) = qrad(:,:,:,:) * omega_old/omega
|
||||
tab_at(:,:,:) = tab_at(:,:,:) * sqrt (omega_old/omega)
|
||||
!
|
||||
! recalculate the local part of the pseudopotential
|
||||
!
|
||||
|
|
Loading…
Reference in New Issue