[Pw_forum] wrong total_weight error with matdyn
Thomas Gruber
thomas.gruber at physik.tu-freiberg.de
Tue Sep 11 15:22:16 CEST 2012
Dear all,
I have done a phonon calculation for a huge system and got finally the
dynamical matrix.
&inputph
tr2_ph=1.0d-13,
prefix='Li12Si7.0.0.1',
recover=.true.
ldisp=.true.,
nq1=1,nq2=1,nq3=1,
amass(1)=28.086,
amass(2)=6.914,
outdir='./'
fildyn='Li12Si7.0.0.1.full.dyn',
/
To exploit the acoustic sum rule I run q2r.x
&input
fildyn='Li12Si7.0.0.1.full.dyn',
zasr='crystal',
flfrc='q2r.fc'
/
and finally matdyn.x
&input
asr='crystal',
amass(1)=28.086,
amass(2)=6.914,
flfrc='q2r.fc',
flfrq='matdyn.freq.phdos'
nk1=1,nk2=1,nk3=1
/
But I get the error message with matdyn.x:
Norm of the difference between old and new force-constants:
7.87014401166549149025
0.833333333333333
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
from frc_blk : error # 1
wrong total_weight
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
I looked in the PW_forum and found two threads, but they had no solutions:
http://www.democritos.it/pipermail/pw_forum/2008-November/010604.html
http://www.democritos.it/pipermail/pw_forum/2011-March/019851.html
My system is orthorhombic and the beginning of my *.dyn file is:
2 152 0 1.8897261 0.0000000 0.0000000 0.0000000 0.0000000
0.0000000
cubic
8.342943957 0.000000000 0.000000000
0.000000000 19.224088437 0.000000000
0.000000000 0.000000000 14.029483693
1 'Si ' 25598.8229845270
2 'Li ' 6301.72549010253
and the only strange thing I found is that it is identified as cubic.
Another orthorhombic system from me (which worked) is identified as
hexagonal. But when I change cubic to hexagonal in q2r.fc, then matdyn.x
encounter the same problem. Can someone give me a hint how to calculate
my phonon frequencies with acoustic sum rule?
Best Regards.
--
Thomas Gruber
Institut für Theoretische Physik
TU Bergakademie Freiberg
Leipziger Str. 23, 09599 Freiberg, Germany
Tel: +49 (0)3731 392006
Email: thomas.gruber at physik.tu-freiberg.de
More information about the users
mailing list