[lammps-users] Compute angles in an amorphous structure

LAMMPS LAMMPS Mailing List Mirror
1

Hi,

I have an amorphous structure with about 3000 atoms.
I have the bond connectivity information and the xyz coordinates for my system.

I needed to compute some angle related properties for the structure.

I use a tersoff type potential, where I have defined my theta parameter.

Do I have to go about defining each of the angles in my data file? Is there a program/tool to compute the angle information table from the xyz and connectivity information?

Thank you.

2

Hi,

I have an amorphous structure with about 3000 atoms.
I have the bond connectivity information and the xyz coordinates for my
system.

I needed to compute some angle related properties for the structure.

during the MD or after the fact?
in which format do you need the angle information?

you are awfully vague. that makes it difficult to give any meaningful advice.

I use a tersoff type potential, where I have defined my theta parameter.

Do I have to go about defining each of the angles in my data file? Is there
a program/tool to compute the angle information table from the xyz and
connectivity information?

yes. the topotools v1.1 plugin for VMD has a guessangles function
that will build a list of angle definitions based on the bond topology.
there may be others.

axel.

3

If you use a Tersoff potential, then you don't define any bonds
or angles in your data file. Tersoff is a bond-order potential
which makes and breaks bonds dynamically depending on
the current atom coords. "Bonds" in LAMMPS, as defined
in the data file, are static, not dynamic.

Steve

4

Hello Steve,

Thanks for your reply.
I use a tersoff style potential parameterized by Munetoh.
When I don’t define angles, bonds, etc. and attempt to compute quantities such as compute angle/local , compute bond/local or use fixes like fix bond/break, LAMMPS returns an error saying Invalid Bond type or Invalid compute command.

Would be great if you can suggest how I can tackle this.

Thanks,
Harish

5

harish,

Hello Steve,

Thanks for your reply.
I use a tersoff style potential parameterized by Munetoh.
When I don't define angles, bonds, etc. and attempt to compute quantities
such as compute angle/local , compute bond/local or use fixes like fix
bond/break, LAMMPS returns an error saying Invalid Bond type or Invalid
compute command.

harish,

these computes and fixes are all meant to be used with
explicitly defined bond and angle potential. for example,
computing the energy contributions is impossible, as those
require the explicit bonded interaction. in a tersoff-style potential,
however, those are all implicit and not depending on the bond/angle
alone, but its environment, too(hence many-body potential). so
using fix bond/break for example makes no sense at all.
...and you will have to do any analysis in post processing.

cheers,
    axel.