The ABINIT Users Mailing List ( CLOSED )

[Masayoshi Mikami <masayoshi.mikami@gmail.com>]

Dear Anil,

I do not understand what is going on there... but let me put another thoughts:
From the log, I was still wondering whether you could mix the different type 
of
pseudos of Ti (pspcod 4) and Nb (pspcod 6) :
------------
 iofn2 : Please give name of formatted atomic psp file
 iofn2 : for atom type   4 , psp file is ../pseudo/nb.fhi
  read the values zionpsp= 13.0 , pspcod=   6 , lmax=   2

 iofn2 : Please give name of formatted atomic psp file
 iofn2 : for atom type   5 , psp file is ../pseudo/22ti.psp_mod
  read the values zionpsp= 12.0 , pspcod=   4 , lmax=   2
(snip)
- pspatm: opening atomic psp file    ../pseudo/nb.fhi
 OPIUM generated Nb potential
  41.00000  13.00000     60306                znucl, zion, pspdat
    6    2    2    0      1140   0.00000     
pspcod,pspxc,lmax,lloc,mmax,r2well

- pspatm: opening atomic psp file    ../pseudo/22ti.psp_mod
 Titane.ion 3s2.3p6.4s0.3d2 rcd=1.65, rcs=rc0=1.25, ecut 32/54
  22.00000  12.00000    940000                znucl, zion, pspdat
    4    3    2    2      2001   0.00000     
pspcod,pspxc,lmax,lloc,mmax,r2well
------------
How about preparing Ti (FHI-type, pspcod = 6, with zionpsp = 12.0) ? 
I was also wondering the local part of the pseudopotential should be of the
similar character... (i.e. in the above case, l=0 (s) of Nb PP and l=2 (d) of
Ti PP will be mixed ... is it OK ??) 

In passing, I should say that we should use the latest stable version of
v.5.4.4 (or v.5.5.4) rather than v.4.4.4 (it is too old).  Still, I do not 
know
why the v.4.4.4 accepted the original input... (I do not have the old source.
The error message is issued by "pspini.F90", but the routines in v.4.6.5 and 
in
v.5.4.4 do not seem different so lot...)

Regards, 
Masayoshi