The ABINIT Users Mailing List ( CLOSED )

[Manuel Pérez Jigato <Manuel.PerezJigato@fys.kuleuven.be>]

hi

when you follow the sequence of convergence tests for all parameters  
in a GW calculation, i guess it is important to make a plan....

what i mean is that, for fully dynamical GW (without plasmon pole  
approximation)
is it a good idea to do all convergence tests using the plasmon pole model
considering it is cheaper? or else, you need to test convergence using fully
dynamical GW calculations?

something else, when you carry out your GW for a system whose LDA  
ground state is metallic (wrongly) expecting GW to open a gap, is it a  
good idea to use one of the broadening schemes, or probably it is more  
consistent to calculate from start (including LDA and GW) using the  
tetrehedron method without broadening?

any other points to be considered when carrying out GW convergence  
tests would be appreciated (i've got the tutorials GW1 and GW2 as  
reference).

thanks

Manuel Perez Jigato
KU Leuven

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm