No, I did NOT do any constrains.
runsp -orb
save_lapw ...
runsp -eece
absolutely no convergence problems !!!
PS: My current ifort version (2016.3.210) fails to compile lapwdm
properly. It over-optimizes even with -O1. Had to use -O0, otherwise I
got runtime errors.
On 01/23/2017 08:08 AM,
Thank you Peter for your reply. I had no doubt that it was possible to
converge the calculation using such strategy. My point was that such
calculation was easy to converge starting from the scratch, directly
using PBE0 and without constraining the total magnetic moment of the
unit cell. At
Sorry, but I cannot reproduce this.
Starting from a converged GGA+U calculation, -eece converges smoothly,
keeps the :mmt at 8 uB and a reasonable gap of 1.6 eV and :MV goes to 10-4.
(quick test with few k-points and rkmax only 6.5)
Am 20.01.2017 um 22:16 schrieb Laurence Marks:
I can
3 matches
Mail list logo