[Wien] E-parameters for surfaces

[Javad Hashemifar]

> 
> This brings me to a question about -in1new that has struck me 
> sometimes. If a saved case (with in1new) is restored, or even simply if 
> an scf-cycle with in1new finished and you start it again by another 
> run_lapw (with in1new), then the :DIS in the first iterations is very 
> high. It takes 5-10 iterations to come back to the original low :DIS. 
> This does not happen if in1new is not used. Is this normal? Has 
> somebody a solution?

Dear Stefaan,
According to my experience this problem occur in spin polarized cases and its
due to differrent linearization parameters for up and dn spins. by using in1new
switch the program (after proper iteration) linearize the case.in1 files
separately for up and down spins in each cycle by the following method.
1. lapw0
2. write_in1_lapw -up
3. cp case.in1new case.in1
4. lapw1 -up
5. write_in1_lapw -dn
6. cp case.in1new case.in1
7. lapw1 -dn

therefor after finishing, case.in1 files contains only the linearization
parameter for down spins and hence when you continue a calculation by another
runsp_lapw the DIS jumps ( because lapw1 -up uses the linearization parameter
of down spins at the first iteration) . To prevent this jump the linearizatin
proccess after continiung a calculation should be start at first iteration.
 
I had the same problem before but I could solve it using the following method.
I always save the case.scf2up(dn) and case.in1 files in addition to the other
files that save_lapw saves.
when I want to continiue a previous calculations always I use in1new -1 that
means linearization should be start from the first iterations.

I hope to it can help.

Best regards,
Javad Hashemifar



=====
-------------------------------------------------------------------------------
Javad Hashemifar                                      phone : ++49-30-8413 4804
Permanent : Physics Department, Isfahan University of Technology, Isfahan, Iran
Current : Fritz-Haber-Institut der MPG, Faradayweg 4-6, D-14195 Berlin, Germany
e-mails:  hashemi at fhi-berlin.mpg.de          &          hashemifar at ph.iut.ac.ir
-------------------------------------------------------------------------------

__________________________________
Do you Yahoo!?
Yahoo! Finance Tax Center - File online. File on time.
http://taxes.yahoo.com/filing.html