I am encountering an issue when I try to create somewhat rigid bonds using a cosine/squared potential with a variety of energy values (theta = 180, k ranging from 1.0 to 100.0) under shear using fix deform. The system equilibrates normally, but when shear is applied the simulations immediately stop with a number of bond atoms missing messages. Under inspection of these atoms, I’ve noted that they’re all cases where the molecule is stretched across the periodic boundary in the shear gradient direction. Making the time step multiple orders of magnitude smaller does not change the execution at all. I’ve only been able to resolve the problem by setting “remap v” to “remap x”, which is not a valid option for my studies.
The atoms are bonded with a FENE bond with an energy value of 30.0. The molecule lengths in this study are short, only 9 particles in length. A 12-6 Lennard-Jones interaction (epsilon 1.0) has been included between all bonded atoms as well, with special bonds set to 1,1,1 to ensure a repulsive interaction between beads of the same molecule.
Simulations are being run on a parallel system, and I get one bond missing error for each processor.
Any help is greatly appreciated.
I am encountering an issue when I try to create somewhat rigid bonds using a
cosine/squared potential with a variety of energy values (theta = 180, k
ranging from 1.0 to 100.0) under shear using fix deform. The system
equilibrates normally, but when shear is applied the simulations immediately
stop with a number of bond atoms missing messages. Under inspection of these
atoms, I've noted that they're all cases where the molecule is stretched
across the periodic boundary in the shear gradient direction. Making the
time step multiple orders of magnitude smaller does not change the execution
at all. I've only been able to resolve the problem by setting "remap v" to
"remap x", which is not a valid option for my studies.
The atoms are bonded with a FENE bond with an energy value of 30.0. The
molecule lengths in this study are short, only 9 particles in length. A 12-6
Lennard-Jones interaction (epsilon 1.0) has been included between all bonded
atoms as well, with special bonds set to 1,1,1 to ensure a repulsive
interaction between beads of the same molecule.
Simulations are being run on a parallel system, and I get one bond missing
error for each processor.
Any help is greatly appreciated.
what are your neighbor list generation settings?
perhaps you need enforcing neighborlist updates more often.
also, please always mention the version of LAMMPS
that you are using. perhaps you are running into a problem
that has been solved already.
axel.
I am using a version from July 19th of 2011.
My neighbor list settings are:
neighbor 1 bin
neigh_modify delay 0
communicate multi vel yes # required for dpd/tstat
You bring up a good point. Since the angle potential is a 3-particle potential, does this mean the neighbor skin should be extended to reach twice the cut-off of any two bonded atoms?