[Pw_forum] using ev.x

[e kb]

Just a small note to be fair to Keane EOS after recently posted examples:As far as I know, Keane EOS is based on an assumption at high pressure.(B' is a monotonicly decreasing function of pressure, and reaches to a limiting value as pressure goes to infinity) Therefore it works best when both low pressure and high pressure input is supplied and intermediate pressure values are found (like when we interpolate between acoustic lab data and shock wave data) So the deviations are not weird when only 5% compression is considered.  cheers, emine kucubenli, SISSA, Italy
--- On Wed, 9/9/09, Tone Kokalj <tone.kokalj at ijs.si> wrote:

From: Tone Kokalj <tone.kokalj at ijs.si>
Subject: Re: [Pw_forum] using ev.x
To: "PWSCF Forum" <pw_forum at pwscf.org>
Date: Wednesday, September 9, 2009, 2:02 PM


Let me just add one more item about bulk modulus.

>From time to time I see the bulk modulus written with too many
significant digits, e.g. 179.3 GPa. Due to the fact that it is a second
derivative calculated numerically and due to a signal/noise that Stefano
mentioned, it is a sensitive quantity. E.g. adding a few additional data
point(s) (or expanding/shrinking the range) may easily change the last
two digits above. 

Here is an example (silver-bulk) of how bulk modulus changes with the
number of data points (the range of scan is from -4% to +4%, and the
fits are performed for several subsets of so-obtained data points):


Bulk Modulus, B0 (in kbar units)::

 ------------+------------+------------+------------+-----------
  # points   | MURNAGHAN  | BIRCH O(1) | BIRCH O(2) |    KEANE  
 ------------+------------+------------+------------+-----------
  15:  1--15 |     896    |     905    |     905    |     610 
  14:  1--14 |     898    |     905    |     904    |     755 
  13:  2--14 |     898    |     905    |     903    |     638 
  12:  2--13 |     900    |     904    |     903    |     748 
  11:  3--13 |     900    |     904    |     904    |     645 
  10:  3--12 |     901    |     904    |     905    |     599 
   9:  4--12 |     901    |     904    |     910    |     665 
   8:  4--11 |     906    |     907    |     909    |     602 
   7:  5--11 |     905    |     907    |     906    |     672 
 ------------+------------+------------+------------+-----------

(the meaning of the first column is "number-of-points: first-point -- last-point")

To add to the Murnaghan vs Keane discussion, it is evident that the
Keane column above displays the largest fluctuations (and also 
different result compared to others). 

Regards, Tone


P.S.: for those interested: the above table was produced by the PWTK 
(http://qe-forge.org/projects/pwtk/) which can calculate lattice
parameter and bulk modulus automatically. For more info see:

Tutorial:
http://pwtk.qe-forge.org/tutorial.html

Examples:
http://pwtk.qe-forge.org/documentation_tclPWTK2F32E20Examples.html

-- 
Anton Kokalj
J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia 
(tel: +386-1-477-3523 // fax:+386-1-477-3822)

Please, if possible, avoid sending me Word or PowerPoint attachments.
See:  http://www.gnu.org/philosophy/no-word-attachments.html


_______________________________________________
Pw_forum mailing list
Pw_forum at pwscf.org
http://www.democritos.it/mailman/listinfo/pw_forum



      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20090909/d95e8f1c/attachment.html>