[Wien] NEC01 charge leakage too large

[Peter Blaha]

 From your case.scfm file I can see:

zhang714 at perlmutter:login21:/pscratch/sd/z/zhang714/wien2k_jobs/SrVO3> 
more SrVO3.scfm
:CINT001 Core Integral Atom   1    9.99999991
:CINT002 Core Integral Atom   2   27.99986575
:CINT003 Core Integral Atom   3    1.99999995
------------------------------------------------
So the core charges for V, Sr and O look perfectly ok. There is NO core 
leakage.
------------------------------------------------
        CHARGES OF NEW CHARGE DENSITY
:NTO   : INTERSTITIAL CHARGE  =     0.000005
:NPC   : INTERSTITIAL CHARGE  =     0.000069
:NTO001: CHARGE SPHERE   1    =    19.285753
:NTO002: CHARGE SPHERE   2    =    35.475474
:NTO003: CHARGE SPHERE   3    =     7.894361

:NEC01: NUCLEAR AND ELECTRONIC CHARGE     85.00000    78.44432
  :WAR   :   NEC01: average large charge leakage:   1.3111368889508412 
    e/atom
  :ERROR :   NEC01: average charge leakage too large: 
1.3111368889508412      e/atom
------------------------------------------------
Again, the charges inside spheres (:NTO001-3) look ok. However,
:NTO (charge in interstital) is definitely wrong. It must be much higher 
and accounts for the missing electrons.

First of all, check in your case.clmval file the Fourier coefficients of 
the density. (bottom of the file). Is the number of PW ok ? does the 
list of K vectors look ok ? Are the coefficients reasonable (or are they 
all nearly zero ?). Compare this list with the list in case.clmsum 
(coming from dstart).

Possible reasons for the problems:

case.struct (wrong symmetry operations). The case.struct you sent 
previously is ok, are you really using this one ?

lapw2: It uses FFTW for the Fourier part. Maybe recompile/reinstall 
(make all   in SRC_lapw2; cp lapw2 lapw2c ..; check carefully for any 
error message. (If FFTW is the problem, also dstart and lapw0 should 
give a problem ?)

mixer: Also here, try to recompile.

Try other examples, eg. the simple TiC from the usersguide. Does it work ?

-- 
-----------------------------------------------------------------------
Peter Blaha,  Inst. f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-158801165300
Email: peter.blaha at tuwien.ac.at
WWW:   http://www.imc.tuwien.ac.at      WIEN2k: http://www.wien2k.at
-------------------------------------------------------------------------