[Pw_forum] Bug in ph.x sometimes occur in hexagonal phases

[Uri Argaman]

Thank you Andrea
I think the phonon calculated with 13 dynamical matrices is wrong because
when I calculate the anisotropic coefficients of thermal expansion in
zirconium I do PW calculations as well as phonon calculations in several
lattice parameters. In some of them (not in all of them), QE calculate 13
dynamical matrices and in the other 12. When I use this data to calculate
the anisotropic coefficient of thermal expansion I got very strange results
that do not agree with some publications which also calculate it from DFT.
I do not know what is the problem. Maybe there is some problem in q2r.x.
Using nr1=30, nr2=30, nr3=48 indeed solve the problem. In addition,
different ecut and ecutrho also solve the problem.
I now calculate the coefficient of anisotropic thermal expansion in
zirconium again with ecut=50 and ecutrho=250. These calculations always
produce 12 dynamical matrices. These calculations takes a few days on my
machines. When I will get the results I will report if it makes any
difference.
Uri Argaman
Ben-Gurion University
Israel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20150827/3c78920c/attachment.html>