Dear all,
I was trying to building a surface which is confined by two indenters from both side in z direction. But whenever the indenter is applied on the surface, the dihedral problem will happen immediately. I have tried to put the indent command at different positions in the input file but it doesn’t work very well. This surface used to run well in the same process without the indenters.
here is my input file which include the indenter part. The force field I was using is “pcff”
units real
atom_style full
pair_style lj/class2/coul/cut 10.0 8.5
bond_style class2
angle_style class2
dihedral_style class2
improper_style class2
dimension 3
boundary p p m
minimize 1.0e-10 1.0e-10 10000 10000
timestep 1.0
fix wall1 all indent 50.0 plane z zlo lo units box # I also tried to set the K value lower like 10, doesn’t work either
fix wall2 all indent 50.0 plane z zhi hi units box
fix pre all nve/limit 0.02 #This is the part that I want to let the atoms move slowly away from each other and get rid of the uncomfortable configuration in the first place. This part works well
run 3000
unfix pre
fix 1 all nvt temp 500.0 900.0 100.0 # This part I want to give the polymer high temperature for a while so they can get enough kinetic energy to entangle with each other. That’s where the dihedral problem happened
run 20000
unfix 1
fix 2 all npt temp 800.0 300.0 75.0 x 1.0 1.0 1000 y 1.0 1.0 1000 couple xy # I also tried to move the “indent” command before this fix, then the simulation can run well until it meets this part
run 1000000
Thank you very much
Jing
If the problem is that dihedrals can't find
atoms and be computed, then it's likely the
indenters are pushing atoms too far away
too fast. Bad input yields bad dynamics.
Steve
Sorry.
The dihedral problem looks like this
“WARNING: Dihedral problem: 18 3077 29455 29454 29470 29471
1st atom: 18 102.223 45.0269 16.1078
2nd atom: 18 101.998 45.098 17.2283
3rd atom: 18 -123.35 172.169 428.58
4th atom: 18 101.82 44.2953 18.7417”
I knew that this is usually caused by the bad configuration. In my case, it’s probably because the indenters are push the atoms too fast and too far away. I tried to increase the distance between indenter and surface and also tried to decease the K value to be as small as 5. Both of them still results in the “dihedral” problem. Could you please provide some suggestions that can improve the input?
Sorry.
The dihedral problem looks like this
"WARNING: Dihedral problem: 18 3077 29455 29454 29470 29471
1st atom: 18 102.223 45.0269 16.1078
2nd atom: 18 101.998 45.098 17.2283
3rd atom: 18 -123.35 172.169 428.58
4th atom: 18 101.82 44.2953 18.7417"
I knew that this is usually caused by the bad configuration. In my case,
it's probably because the indenters are push the atoms too fast and too far
away. I tried to increase the distance between indenter and surface and also
tried to decease the K value to be as small as 5. Both of them still results
in the "dihedral" problem. Could you please provide some suggestions that
can improve the input?
if you are pushing atoms too fast, you may also
need to reduce the velocity of the indenter.
however, whatever happens, should be easy to determine
by simply dumping your coordinates frequently and
visualizing them.
there is plenty of advice for situations of bad dynamics
in the lammps documentation and the mailing list archives.
axel.