[Wien] Optical properties for new and old WIEN

[Peter Blaha]

Not very helpful either! Did you start with identical vectors or are you 
comparing 2 scf calculations, one with old and one with new version ?

What happens if you run the "new" kram with the "old epsilons" and vice 
versa.

I want to find out where the difference is.

Start with identical vectors, run optic new and old, compare the matrix 
elements.

Start with one optic file and run joint old/new (compare eps2)

Start with one eps2 and run kram old/new. Where are the differences.

Robertson Burgess schrieb:
> Sorry, I should have mentioned these things perviously.
> Epsilon in the two components are similar between the old and new WIEN
> codes, but again there are noticeable differences. These are most
> noticeable towards the lower end.
> 
> We have run these calculations with bulk aluminium, and found that both
> codes agreed very well on the plasmon response.
> 
> Many thanks
> Bob Burgess
> 
>>>> Peter Blaha <pblaha at theochem.tuwien.ac.at> 06/28/07 5:54 PM >>>
> Just a few questions for clarification:
> 
> Is the difference just in the eloss spectra ?
> 
> Is eps2 and eps1 fine ?
> 
> Do the problems also appear in bulk fcc Al ?
> 
> 
> Robertson Burgess schrieb:
>> Dear WIEN users,
>>     Firstly many thanks for your help in the past, I have been
>> calculating the optical properties for slabs of Aluminium and you have
>> all been very forthcoming in helping me along in the past couple of
>> months.
>>     Unfortunately, after getting some very odd results using the
> latest
>> version of WIEN, I had previously requested that an older version,
>> around WIEN2k_03 be sent to me. After running identical calculations
> for
>> a 5 atom slab of Aluminium on both the latest WIEN2k_07 and older
>> WIEN2k_03, the eloss spectra look distinctly different for the two
>> versions. The most notable difference is the difference in the eloss
>> spectra when calculating oscillations perpendicular to, and
>> perpendicular to the slab plane. In the most recent version the
> spectra
>> were similar, with slight energy differences and intensities for the
>> main peak but similar overall shape. Calculations over a range of
>> thicknesses showed that this trend continued, with no clear trend
>> between thickness and difference between peaks in either component. In
>> the older version, however, there was a very distinct difference
> between
>> the eloss spectra for either direction.
>>       I plan on looking through the two versions of code to try and
> find
>> where this difference could have occured, but any suggestions of where
>> we might look would be very appreciative. I have attached the
> structure
>> file I used, again it is a 5 atom thick slab of aluminium separated by
>> vacuum.
>>
>> Many thanks in advance,
>> Bob Burgess
>> University of Newcastle
>>
>>
>>
> ------------------------------------------------------------------------
>> _______________________________________________
>> Wien mailing list
>> Wien at zeus.theochem.tuwien.ac.at
>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>