angle lj/sdk

LAMMPS LAMMPS Mailing List Mirror
1

Hi Lammps Users,

with pair_style lj/sdk ( also with pair_style cg/cmm ) for Coarse-Grained MD

angle_style sdk uses harmonic form,

K * ( thetha - thetha o )^2 ,

Now in documentation, K a prefactor, with the repulsive part of the non-bonded lj/sdk pair style between the atoms 1 and 3.

Now, how come repulsive part is included in force constant ‘K’.

  1. Non-bonded interactions are calculated, if exceeded three atoms. ( i.e. from 1-4)

now , why for angle terms, non-bonded paramters are referred…

Thanks in advance,

B karteek kumar
Molecular Simulations Lab,
JNCASR.

2

Hi Lammps Users,

with pair_style lj/sdk ( also with pair_style cg/cmm ) for Coarse-Grained
MD

angle_style sdk uses harmonic form,

      K * ( thetha - thetha o )^2 ,

Now in documentation, K a prefactor, with the repulsive part of the
non-bonded lj/sdk pair style between the atoms 1 and 3.

Now, how come repulsive part is included in force constant 'K'.

no, it isn't. the sentence could be worded better.
please note the comma.
K only applies to the harmonic angle potential.
a quick look at the sources will confirm it.

2. Non-bonded interactions are calculated, if exceeded three atoms. ( i.e.
from 1-4)

   now , why for angle terms, non-bonded paramters are referred..

because in the coarse grain model, the angle forces are rather weak
and due to the use of exclusions (1-4 should *not* be excluded, btw),
there is no repulsion that would prohibit atom 1 and 3 to stay right
on top of each other. to give an example, take a nonane molecule,
which would be represented by 3 beads. without the repulsion in the
sdk angle, you may end up with molecules that act as if you had
only two beads and one had double the LJ interactions, i.e.
instead of: A - B - A you would have: B - 2*A
this is clearly not physical and the sdk angle style avoids this.

you can see this as doing only a half-exclusion, i.e. in the SDK
model you only remove the attractive 1-3 interaction.

HTH,
     axel.

3

Actually, I do see why someone might be confused.
The documentation for this angle_style says:

"The sdk angle style is a combination of the harmonic angle potential,
   E = K*(theta-theta0)^2
where theta0 is the equilibrium value of the angle and K a prefactor,
with the repulsive part of the non-bonded lj/sdk pair style between
the atoms 1 and 3."

http://lammps.sandia.gov/doc/angle_sdk.html

Without reading the source code (or the paper links on the pair style
sdk docs), it's not that clear exactly what this angle style does.
Anyone want to augment the documentation with a sentence or two?

Cheers
Andrew

4

If Axel (or anyone else) wants to provide the
full formula in a LaTeX file, like those in doc/Eqs,
we'll add it to the sdk doc page.

Steve

5

i'll add this to my todo list.

the text is an obvious victim of copy-n-modify.
not sure, if there is an easy way to represent
the interaction in a formula.

axel