[Wien] Such a vast difference in two values of K’ for the same case.

Peter Blaha pblaha at theochem.tuwien.ac.at
Mon Jul 30 09:36:41 CEST 2012


I would not bother too much about B', but B is changing by 20 %, which is too much.

You have to analyse your data to find out where your error is. Maybe in one set of
calculations you have not got the optimized c/a ratio (and corresponding energy) for
each volume ???

PS: 5 points are probably ok for equillibrium V and c/a, but B will always have quite
some error.

Spin-orbit:   Emax in case.in1 should be increased.
Emax in case.inso can stay at eg. 1.5 Ry
You don't need to worry about "qtl-B" at very high (unoccupied) energies (unless for
some spectroscopy you need these high states).


Am 30.07.2012 06:15, schrieb Vinayak Mishra:
> I have basically two questions to ask. I will like to mention that I
> have gone through the related previous questions and their answers on
> registered users site. I will be highly obliged if you could kindly
> spare some time to go through the following two queries.
>
> 1.) I was doing c/a optimization at five different volumes in bct
> structure for a particular case and was trying to get equilibrium
> volume and tentative value of bulk modulus and its pressure
> derivative. I gave optimization for one set of 5 volumes, say  +5.0,
> 0.0, -5.0, -10.0, -15.0 and another set, say  +5.5, 0.5, -5.5, -10.5,
> -15.5  and then get EOS, then I find that there is a huge difference
> in the value of pressure derivative of bulk modulus K’ for the two
> cases. I have plotted both the cases together in the plot.ps file
> attached herewith. Here in the plot ‘E’ is the same constant value of
> energy for both the curves. Only thing is, for the sake of clarity of
> seeing the two curves together, in the solid curve 0.001 Ryd of energy
> is added so that it is shifted up by that much energy so that one can
> see the difference between the shape of two curves clearly. Both the
> curves as such don’t seem to be much different, then why  such a large
> difference in K’. Even otherwise also, why should it matter  whether
> the run is given for 1st set of volumes or the 2nd set of volumes.
> Could you please comment on this Prof. Blaha?
>
> 2.)While doing a spin-orbit calculation, we need to increase Emax upto
> 10 Ry in case.in1 file even though we are interested in eigenvalues
> near Fermi energy. But as said in users guide, in the
> "second-variation" these high eigenstates  form the basis functions
> for the SO calculation. Without these high eigenvalues there will be a
> poor basis even for states near Fermi energy.. But these high
> eigenvalues are giving large QTL-B value of 980 or 1500 value, at high
> energies say around 4.0 in case.output2 file if emax is set as 5.0
> (instead of the default value 1.5) in case.inso file also apart from a
> Emax=5 (default in SO calculations) in case.in1 file, when we try to
> calculate partial charges by giving the command 'x lapw2 -so -qtl'.
> After this, the command 'x tetra'  is given to get density of states.
> So as said earlier, this is so when case.inso file also has emax = 5.0
> i.e the maximum energy for which output eigenvectors and eigenenergies
> will be printed (p.96 of users guide). But if we put emax =1.5, which
> is the default value in case.inso file, then anyway at high energies
> no computation of eigenvectors and eigen energies is done and no such
> statement appears in the case.output2 file. It is true that fermi
> energy is far below the energy level at which high QTL-B is quoted in
> case.output2 file. My question here is: How and where Emax = 5.0 of
> case.in1 file is used so that getting large qtl-b in output2 while
> calculating partial charges, is meaningless or unharmful, if we happen
> to keep emax equal to 5.0 in case.inso also. I just want to confirm
> that a large qtl-b value at energies 0.5 to 1Ryd higher than Fermi
> energy in case.output2 file is harmless and that I can use Emax upto
> 10 in case.in1 file without any hesitation. I will highly appreciate
> your clarification on this.
>
> Thank you very much in advance for your attention and time.
>
> V. Mishra
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-165300             FAX: +43-1-58801-165982
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------



More information about the Wien mailing list