[Wien] ROTDEF error when adding spin-orbit coupling

[Roberto Iglesias]

Thank you, Peter, that fixed the problem!

Kind regards

Roberto



Peter Blaha wrote:
> Recently I encountered a similar problem. It is due to some rounding problems
> in symmetso (mainly for non-orthogonal lattices like R or monoclinic).
> 
> As you can see, in your case the z-coordinate of the second atom should be
>   0.7507380
> 
> You need to fix the struct file by hand.
> 
>> I successfully ran a spin-polarized calculation on a rhombohedral 16 atom supercell. Then I tried to add 
>> spin-orbit coupling to the system, with the magnetization along the 1 1 1 direction. When I try to runsp_lpaw 
>> -so I get an immediate error, see the lapw0.error file:
>>
>>   'ROTDEF' - no symmetry operation found.
>>   'ROTDEF' - for jatom, index 612
>>   'ROTDEF' - atomposition of jatom   0.2492620   0.7503657   0.2492620
>>   'ROTDEF' - atomposition of index   0.7507380   0.2496343   0.7507379
>>
>> I know this is related to the atomic positions given in the last two lines. Even if a successful calculation 
>> has already finished, perhaps I should run again nn, sgroup, symmetry and a full initialization before 
>> proceeding.
>> I'm using WIEN2k_07.3 version. I've senn similar problems in the mailing list, but they appear mainly in the 
>> first stages of initialization or with dstart.
>> Any suggestions are welcome.
>>
>> Kind regards
>>
>> Roberto
>>
>