[Wien] about the spin-orbit coupling

[Peter Blaha]

when you run:     x dstart -c -up
There cannot be an error about reading case.in2  (because it should be 
case.in2c

So you ran x dstart -up   ???

You have to understand the effect of spin-orbit for magnetic systems.

It can happen (ONLY for MAGNETIC SYSTEMS) that spin-orbit reduces 
symmetry and you may loose inversion or even get more inequivalent atoms 
than before.
initso tries to handle this and even rewrites the input files (if SO 
made 2 atoms out of one position, all in1,in2,inc,... files have to be 
updated.
In addition the charge density ("stars" of rho_k) is different and the 
LM-list,...

There might be cases, where initso is not completely successful in 
modifying all files and one needs further adaptations by hand, because 
the symmetry changes are too difficult.
But nobody can help you without the specific example in hand.

xiaojuan liu schrieb:
> Thanks for your kind help!
> 
>  (1) i got the help files, thank you!
>  (2) if there is a comparatively large difference between them, how can 
> we avoid it.
>       Change which parameter in the input files?
>  (3) when I try to commit  x dstart -c -up/dn, there is always some 
> errors which is show in the follow:
>  
>   Input/Output Error 140: Invalid integer
>  
>    In Procedure: init
>         At Line: 72
>  
>       Statement: Formatted READ
>            Unit: 15
>    Connected To: TbCuMnO.in2
>            Form: Formatted
>          Access: Sequential
> Records Read   : 8
> Records Written: 0
>  
> Current I/O Buffer:
>  
>  
> ** Address Error **
> 
>  
> End of diagnostics
>  
> 0.000u 0.000s 0:00.00 0.0%      0+0k 0+0io 0pf+0w
> error: command   /home/lxjuan/WIEN2k/dstart dstart.def   failed
> 
> Can you help me in these questions?
> 
> Thank you!
> 
> --- *08年12月3日，周三, Peter Blaha /<pblaha at theochem.tuwien.ac.at>/* 写道：
> 
>     发件人: Peter Blaha <pblaha at theochem.tuwien.ac.at>
>     主题: Re: there is no case.help03x files
>     收件人: karenliuciac at yahoo.com.cn
>     日期: 2008,123,周三,12:13上午
> 
>     Please use the mailing list next time!
> 
>     a)  x lapw2 -h      tells you how to run lapw2 with help-files.
> 
>     b) most likely the error is
>      already in the previous iteration
>     (the first SO iteration). Check the scf file and compare it with the non-SO scf
>     file.
>     SO should be a "small" difference in most quantities.
> 
>     c) when you have a difficult symmetry, it could be that the
>     extrapolation of the old density fails. In this case you need to
>     restart with   x dstart [-c] -up/-dn to generate new densities.
> 
>     xiaojuan liu schrieb:
>     > Dear Peter Blaha:
>     >    I'm carrying out some calculations on TbCu3Mn4O12 systems, and I
>     want to add spin-orbit coupling effect in the [111] direction. According to the
>     userguide, first, the spin-polarized calculation were carried out, then initso,
>     and then runsp -so. But SCF always stop at the second cycle, showing SECECT
>     error. I want to refer the TbCuMnO.help03X files, but I can not find these files
>     in the new version of WIEN2K_08. Now it's difficult for me to adjust the
>     linear energy of El without these kind of
>      files. Can you give me some advices
>     and guidance.
>     >    Thank you.
>     > 
>     > 
>     > ------------------------------------------------------------------------
>     > 好玩贺卡等你发，邮箱贺卡全新上线！
>     <http://cn.rd.yahoo.com/mail_cn/tagline/card/*http://card.mail.cn.yahoo.com/>
> 
> 
>     -- 
>                                           P.Blaha
>     --------------------------------------------------------------------------
>     Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
>     Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
>     Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
>     --------------------------------------------------------------------------
> 
> 
> ------------------------------------------------------------------------
> 好玩贺卡等你发，邮箱贺卡全新上线！ 
> <http://cn.rd.yahoo.com/mail_cn/tagline/card/*http://card.mail.cn.yahoo.com/> 
>