<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"><HTML DIR=ltr><HEAD><META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1"></HEAD><BODY>
<DIV id=idOWAReplyText47271 dir=ltr>
<DIV dir=ltr><FONT face=Arial color=#000000 size=2>This is simply a warning to
people having trouble with elastic constant calculations in wien2k.</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial color=#000000 size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial color=#000000 size=2>I had problems obtaining
sensible (not even good, just sensible) results for elastic constants of cubic
TiNi.</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2>After some time I discovered that this is
due to the fitting routines used to fit the energy vs strain curves obtained for
eos, rhomb and tetra strains.</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2>For eos there is no problem.</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2>For rhomb and tetra, the routines are set
to use a polynome of maximal degree (ie, number of points - 1). This leads
to absurd results. (Let me add that I used rather small strains (for which
a quadratic fit seems nice enough), which may amplify the
pathologies).</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2>The crazy thing is, the available fitting
routines are quite sophisticated : they contain criteria to choose the fitting
order themselves based on some statistical significance test, but right now this
test is overridden!</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2>After letting the routines choose the
fitting order themselves (with a minimum of order 2), the results got a lot
closer to what they should be ...</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2>Secondly : I remember someone was going to
generalize the SRC_elast module to some lower symmetry groups. If this is
you : for which symmetries are you doing this? I would like to calculate
elastic constants for rhombohedral systems next; any chance you could help me,
or we could help each other?</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial size=2>Thanks,</FONT></DIV>
<DIV dir=ltr><FONT face=Arial size=2></FONT> </DIV>
<DIV dir=ltr><FONT face=Arial color=#000000 size=2></FONT> </DIV></DIV>
<DIV id=idSignature76869 dir=ltr>
<DIV><FONT face=Arial color=#000000 size=2></FONT>Kevin Jorissen</DIV>
<DIV> </DIV>
<DIV>EMAT - Electron Microscopy for Materials Science (<A
href="http://webhost.ua.ac.be/emat/">http://webhost.ua.ac.be/emat/</A>)</DIV>
<DIV>Dept. of Physics</DIV>
<DIV> </DIV>
<DIV>UA - Universiteit Antwerpen</DIV>
<DIV>Groenenborgerlaan 171</DIV>
<DIV>B-2020 Antwerpen</DIV>
<DIV>Belgium</DIV>
<DIV> </DIV>
<DIV>tel +32 3 2653249</DIV>
<DIV>fax + 32 3 2653257</DIV>
<DIV>e-mail <A
href="mailto:kevin.jorissen@ua.ac.be">kevin.jorissen@ua.ac.be</A></DIV>
<DIV> </DIV></DIV></BODY></HTML>