distorted unit cell, crash due to nqc .ne. nq1*nq2*nq3
Posted: Mon Oct 28, 2019 5:17 pm
Hi all,
I'm trying to run EPW for a distorted MoS2 single layer system. Due to the distortion most of the symmetries are gone.
For the undistorted lattice I get 7 irreducible q-points for a 6x6x1 phonon grid. For the distorted lattice the number of irreducible q-points increases to 13. This is what EPW gave me in this case:
Another information in the EPW log which bothers me is:
I guess this has something to do with the problem. Is it possible that the symmetries found by EPW are not the same as those in the ph.x and pw.x runs? Because at the last irreducible q-point nr 13 all 36 q-points of the coarse mesh should be done, but the counter is at 23 (see first log snippet).
I tried to overcome this issue by recalculating the phonons with nosym=.true. and noinv=.true. prior to ph.x. From that I got 36 q-points. Now EPW crashes with the following output (the last lines before the crash):
I guess this time it is an out-of-index issue, because q(37) should not exist. Why EPW tries to calculate more q-points than there are? If the -q points were neglected, I'd end up with exactly 36 q-points. So why does EPW take -q points into account even though all symmetries were disabled? How can I fix this problem?
Thank you in advance!
Maxim
I'm trying to run EPW for a distorted MoS2 single layer system. Due to the distortion most of the symmetries are gone.
For the undistorted lattice I get 7 irreducible q-points for a 6x6x1 phonon grid. For the distorted lattice the number of irreducible q-points increases to 13. This is what EPW gave me in this case:
Code: Select all
===================================================================
irreducible q point # 13
===================================================================
Symmetries of small group of q: 2
in addition sym. q -> -q+G:
Number of q in the star = 1
List of q in the star:
1 -0.494155181 -0.285300626 0.000000000
q( 23 ) = ( -0.4941552 -0.2853006 0.0000000 )
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Error in routine elphon_shuffle_wrap (23):
nqc .ne. nq1*nq2*nq3
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%Another information in the EPW log which bothers me is:
Code: Select all
Progress kgmap: ########################################
kmaps : 1338.62s CPU 1340.16s WALL ( 1 calls)
Symmetries of Bravais lattice: 8
warning: symmetry operation # 2 not compatible with FFT grid.
1 1 0
0 -1 0
0 0 -1
warning: symmetry operation # 4 not compatible with FFT grid.
1 1 0
0 -1 0
0 0 1
Symmetries of crystal: 2
I guess this has something to do with the problem. Is it possible that the symmetries found by EPW are not the same as those in the ph.x and pw.x runs? Because at the last irreducible q-point nr 13 all 36 q-points of the coarse mesh should be done, but the counter is at 23 (see first log snippet).
I tried to overcome this issue by recalculating the phonons with nosym=.true. and noinv=.true. prior to ph.x. From that I got 36 q-points. Now EPW crashes with the following output (the last lines before the crash):
Code: Select all
===================================================================
irreducible q point # 20
===================================================================
Symmetries of small group of q: 2
Number of q in the star = 1
List of q in the star:
1 -0.498273140 -0.087967693 0.000000000
In addition there is the -q list:
1 0.498273140 0.087967693 -0.000000000
q( 36 ) = ( -0.4982731 -0.0879677 0.0000000 )
q( 37 ) = ( 0.4982731 0.0879677 -0.0000000 )I guess this time it is an out-of-index issue, because q(37) should not exist. Why EPW tries to calculate more q-points than there are? If the -q points were neglected, I'd end up with exactly 36 q-points. So why does EPW take -q points into account even though all symmetries were disabled? How can I fix this problem?
Thank you in advance!
Maxim