[lammps-users] Possible bug in angle style

Hi,
I am trying to get my worm like chain model to match theory, and I think I may have found a bug in angle_style harmonic, not sure. I have a bead spring polymer, and my bending energy is simply not matching theory, and all the parameters I have found in literature that produce a certain persistence length are not working.

I tried a desperate measure. First I compiled the most recent lammps, no change. Then I thought, CHARMM looks like harmonic if I set Kub=0. I did this and I get a different simulation. In principle I should get the same thing right? It is not just a little different, I was getting an average bend angle of 135 degrees with harmonic, then 145 degrees with CHARMM, which is a massive difference in persistence length. Is there a caveat with doing what I did?

Thanks,
Josh

Hi,

hi josh,

I am trying to get my worm like chain model to match theory, and I think I
may have found a bug in angle_style harmonic, not sure. I have a bead
spring polymer, and my bending energy is simply not matching theory, and all
the parameters I have found in literature that produce a certain persistence
length are not working.
I tried a desperate measure. First I compiled the most recent lammps, no
change. Then I thought, CHARMM looks like harmonic if I set Kub=0. I did
this and I get a different simulation. In principle I should get the same
thing right? It is not just a little different, I was getting an average
bend angle of 135 degrees with harmonic, then 145 degrees with CHARMM, which
is a massive difference in persistence length. Is there a caveat with doing
what I did?

you are right, that the charmm angle style with UB disabled is the
same as harmonic. but it doesn't make any sense, that you get
different results. the source code of the two classes is identical,
except for the UB terms, and i just made a test with an input on my
machine and i get identical trajectories, if i replace angle style harmonic
with charmm while setting K_UB to 0.0.

please post a (minimal) example that illustrates what you are seeing.
there has to be something else that is at fault.

cheers,
    axel.