Hi everyone,
I am wondering if anyone has encountered the following message in the output file?
Warning: Lagrange Multiplier tolerance too high: 1.776356839400250E-015
+Try using a smaller time step
+Gram-Schmidt being performed, spin: 1
This problem has been bugging me for weeks now. Once the warning message appears, the job is 'stuck' at that point and does not progress. I have no idea what is going wrong here. I have tried changing the fake_mass, tolerance (i.e. by adding another line 'tolerances 1e-15 1e-15' in the nwpw module), and time_step, but with no success. Does anyone have a solution/suggestion to this problem?
Here is my original input file (168 atoms in total).
................................
scratch_dir /scr/pmand
start pmand
memory 1400 mb
echo
print low
geometry
N -0.34600 37.95800 6.38200
.
.
.
.
.
.
H -9.63500 32.49000 11.41700
H -2.17520 30.40610 15.99030
symmetry c1
end
nwpw
simulation_cell units angstroms
boundary_conditions periodic
lattice_vectors
29.5 0.0 0.0
0.0 24.0 0.0
0.0 0.0 27.5
ngrid 118 96 110
end
wavefunction_cutoff 70.0
energy_cutoff 280.0
xc pbe0
car-parrinello
Nose-Hoover 100.0 298.15 100.0 298.15
time_step 4.0
fake_mass 400.0
loop 100 1000
xyz_filename pmand.xyz
emotion_filename pmand.emotion
ion_motion_filename pmand.ion_motion
end
end
set geometry:actlist 1:156
task pspw energy
task pspw car-parrinello
...................................
Thank you so much for reading this and any advice is much appreciated.
|