[Wien] volume optimization and case.in1 file

[Ling Miao]

Hi,

Thanks for the answer, it helps me solve the problem of plotting E
vs. volume. Now I can get the curve, but it does not look like the one
shown on manual. Even though the Emin and Vmin and bulk modulus are very
close, but the shape are different. my curve has E at -10% closest to the
Emin and the other side goes up very high, while in the manual, it looks
symmetric, with Emin close to 0% volume change. the values are listed
below: 

Murnaghan  Vo     B(GPa)    BP        Eo
manual 138.1132  252.9912  4.2216  -1783.957067
mine   138.0491  248.5960  4.1532  -1783.957172

Is it possible that my package is not compiled corrected?

I also have some questions about the Fe (P27) and Cd (36)examples in the
'step-by-step introduction' written by S. Cottenier, 2002. I kind of
understand the modification of case.in1 file in order to get correct
linearization energies in the Fe example, however when i read the Cd
example,i am confused. Why there is no local orbital by using a step and
'STOP'? As I notice, in the case.in1_st there is always a STOP to search 
a local orbital energy, and i know each specific case need additional
modification of the case.in1 file, but why we see all those in the Fe
example while not in the Cd example? And actaully i am not able to get the
same ENE and EFG data even though i use the same case.in1 file shown on
p36. I get much lower energy and my gradient tensor is negative. I want to
know have anybody done these before, can you tell me if your results are
the same? Can you give me a clear answer about how to chose LO and APW+lo
for each orbital?


Thank you very much for your help
Ling



--------------------------
Message: 27
Date: Tue, 16 Nov 2004 10:14:24 +0100 (CET)
From: Peter Blaha <pblaha at zeus.theochem.tuwien.ac.at>
To: wien at zeus.theochem.tuwien.ac.at
Subject: Re: [Wien] Volume Optimization
Reply-To: wien at zeus.theochem.tuwien.ac.at

> Click 'X Optimize', choose 'vary VOLUME with constant a:b:c ratio' and
> enter -5, -10, 0, 5, 10 in the strcture changes box with each value in a
> seperate line. 
> Click 'edit optimize.job', uncomment 'x dstart'
> Click 'run optimize job', untill the run finishes,
> finally click 'plot' using case_vol_-5.scf file

Why would you specify "case_vol_-5.scf"  ????

You want to plot the energy vs. volume for all cases you have done!
So leave this field with the default:  using  *     .scf files

(This can be used to select particular names of scf files like in a
standard ls command.  Suppose you do 2 runs with 1000k and 5000k points.
You modified optimize.job and    save_lapw ${i}_1000k    (or 5000k).

For plotting, using *1000k   or *5000k would select the desired run.)