table potentials

LAMMPS LAMMPS Mailing List Mirror
1

Hi LAMMPS users,

I am interested to model a bond which has anharmonic distribution.

I want to use bond_style table for the same.

Now, to get the table potential, I am following these steps

  1. Inverting the potential as - KB T ln { P® /r^2 }

  2. force = - dU/ dr

But I am not able to get back the initial P®, Instead atoms are moving very far from each other.

What stupid thing, I am unable to figure out… Can anyone help me how to get proper table potentials from the normalized distributions

Thanks

spandana

2

I suggest you plot the values you are listing in the bond

table file to insure they are the function you expect. You
could start with something simple like a harmonic bond

to insure you are able to use bond_style table correctly.

Steve

3

Thanks steve,

The energy profiles look much like I wanted them.

Using Xmgrace differences, I calculated the force using forward difference method.

With this style, atoms are moving very far apart.

Still can’t figure out the mistake.

Thanks
spandana

4

The author of bond_style table is listed at
the top of the src/bond_table.cpp file with his email.

I suggest sending him an email and seeing if
he has a test problem you can try to verify
it works as you expect.

Steve