Respected sir/madam,
I’m trying to calculate stress/strain curve for single layer graphene. I have following doubts about my simulation.
I gently request you to go through and help me.
- Initially, I have to relax the system for 30ps and time step mentioned in the paper is 5 fs. when i run the graphene sheet with that time step atoms in the sheet are missing. so I changed the time step to 0.0005 fs now my atoms are not missing. Do this have any effect on potential energy of the sheet?
- I’m using airebo potential, I’m getting potential energy of -153 where I’m expecting it to be around -1700( from the paper)
why is this happened and does this have any effect on stress/strain curves or ultimate strength value.
- how can we find gradient of potential energy curve using lammps?
Thanks & Regards,
Venkata Pavan Kumar. Miriyala
pavan.kumar9111@…24…,
contact number: +91-9032307746
Input file is attached here.
input.graphene (871 Bytes)
Respected sir/madam,
I'm trying to calculate stress/strain curve for single layer
graphene. I have following doubts about my simulation.
I gently request you to go through and help me.
1) Initially, I have to relax the system for 30ps and time step
mentioned in the paper is 5 fs. when i run the graphene sheet with that time
step atoms in the sheet are missing. so I changed the time step to 0.0005 fs
now my atoms are not missing. Do this have any effect on potential energy
of the sheet?
0.0005fs is ridiculously small and will waste a lot of computer time
to get the (rather short) 30ps total time. are you sure you have the
correct units setting and are using the proper timestep unit?
the size of the time step should not affect the potential energy, but
it should be reasonable. if it is not, there is something wrong.
impossible to say, what, though.
2) I'm using airebo potential, I'm getting potential energy of -153 where
I'm expecting it to be around -1700( from the paper)
why is this happened and does this have any effect on stress/strain
curves or ultimate strength value.
absolute energies are completely meaningless unless you compare two
identical implementations. numbers without units are even more
useless, if that is at all possible.
3) how can we find gradient of potential energy curve using lammps?
extract the energies from the log file and then fit the numbers to a
suitable equation of state (or if you absolutely want to ignore
entropy contributions, just fit to a suitable polynomial and extract
the gradient from those coefficient. a trivial exercise in calculus.
there should be lots of tools for both.