Sangeeta Sharma

Mr. AK These are the works i mentioned to in my initial mesg. (specially Krieger et al.) where we tried it. But it does nothing. So we removed it. Regards, S

Dear Mr. AK. Can you kindly point me to paper that you have in mind (where it is mentioned in the Hamiltonian)? Regards S

Hi Mr. Kefayati, There is no particular explicit rule for questions being posted on the Elk forum. However, my past experience shows that there are 3 types of questions: 1. very well posed and informed question containing a bug reports. This type of questions are taken up very quickly by Kay. 2. questions regarding complex Physics described or potential problem with running of the code. This type of questions are usually next. These two types of questions usually require direct developers input....

Hi Mr. Kefayati, There is no particular explicit rule for questions being posted on the Elk forum. However, my past experience shows that there are 3 types of questions: 1. very well posed and informed question containing a bug reports. This type of questions are taken up very quickly by Kay. 2. questions regarding complex Physics described or potential problem with running of the code. This type of questions are usually next. These two types of questions usually require direct developers input....

Dear Ronald, I took your elk.in and tried to run it. I am unable to reproduce your data. I get exactly the same band structure with M in x or y direction. Can you try the input file I am attaching? I also attach the BAND.OUT file. Best s

Dear Andrew, BSE uses two variable: nvbse and ncbse which you can specify in elk.in. First one is number of valence bands and second is the number of conduction bands used for BSE calculations. If nvbse was equal to all occupied and ncbse all empty states you will get the full spectrum. However, since BSE is very computationally demanding, active space is reduced and you get absorption from only restricted set of states. This also implies that you cannot just add the BSE and RPA spectra. Best Sa...

Two types of files obtained by task 320. Results obtained without LFE are written in files EPSILON-TDDFT-33.OUT. Results obtained with LFE are written to EPSM_TDDFT_33.out. In order to generate results with LFE you must set gmaxrf>1. With task 320 you can use various kernels fxctype =0 is RPA, 210 is bootstrap, 211 is single iteration bootstrap etc. Task 121 also produces optical spectra, but this is always using RPA and without LFE. These results are written to file EPSILON_33.OUT.

Dear wkq, All units in the ELK code are atomic units. 1 unit of field is = 1715.255541 Tesla. Please check page 12 of the Elk manual for this. Setting bfcm or bfieldfc breaks the symmetry and either is sufficient for ferro-magnetic materials. For AFM or non-collinear materials you would need to use bfieldcmt, where you need to point it in the direction of the moment for the corresponding atom. Best Sangeeta