Please keep the list in the loop.
Your angle energy is suspiciously large.
This suggests a bad initial config, or some
issue with the fix SHAKE and a 180 angle.
I would see if you can run with a flexible
angle first before trying shake.
Steve
Hi Steve,
I'll keep that in mind!
Yes SHAKE should be the problem, because the simulation also runs if I
replace the fix SHAKE with a rigid/nvt fix.
I'll go ahead an try the flexible angle run next week. Thank you for
the information!!
Jan
Hi Steve,
I'll keep that in mind!
Yes SHAKE should be the problem, because the simulation also runs if I
replace the fix SHAKE with a rigid/nvt fix.
LAMMPS just does what you ask it to do.
your bond energy is zero, but your angle
energy isn't. that means that your angle
potential is still active and not properly
specified. if SHAKE would be constraining
your angles, that energy would be zero, too.
that being said, in its current form, the SHAKE
fix in LAMMPS cannot handle 180 degree angles,
since there are no precautions for the resulting
divergence. we've been through this on the mailing
list several times and fix rigid was the only viable
solution without somebody spending some time
on hacking the code.
I'll go ahead an try the flexible angle run next week. Thank you for
the information!!
axel.