<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix"><br>
OK, I see, it is the acoustic phonon that gets imaginary. Then you
are right, it is an accuracy issue, but it can be fixed by
imposing the Acoustic Sum Rule in the dynmat.x or matdyn.x run
(see PHonon/example02, for instance). Otherwise you would have to
raise cutoffs and number of k-points to very high values in order
to get a decent frequency.<br>
<br>
Gabriele<br>
<br>
<br>
<br>
<br>
On 08/20/2014 11:13 AM, Kvasov Alexander wrote:<br>
</div>
<blockquote cite="mid:CB2FAFD1-9CA2-44C2-85E5-FD4895458852@epfl.ch"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=windows-1252">
<div>Dear Gabriele,</div>
<div>yes, I changed the volume, it's less then the relaxed value.
this way TO phonon is stable at gamma point. </div>
<div>the right values for TA branch should go to 0.</div>
<div>Alexander</div>
</blockquote>
<br>
<br>
<pre class="moz-signature" cols="72">--
Dr. Gabriele Sclauzero
ETH Zurich
Materials Theory
HIT G 43.2
Wolfgang-Pauli-Str. 27
8093 Zürich, Switzerland
Phone +41 44 633 94 10
Fax +41 44 633 14 59
<a class="moz-txt-link-abbreviated" href="mailto:gabriele.sclauzero@mat.ethz.ch">gabriele.sclauzero@mat.ethz.ch</a>
<a class="moz-txt-link-freetext" href="http://www.theory.mat.ethz.ch/people/postdocs/gsclauze">http://www.theory.mat.ethz.ch/people/postdocs/gsclauze</a></pre>
</body>
</html>