Sphere Clumps in LAMMPS

LAMMPS LAMMPS Mailing List Mirror
1

Hi Lammps Users,

Is it possible to create Sphere Clumps in LAMMPS? starting with one spherical particle, 1 becomes 2, 2 becomes 4 , 4 becomes 8 etc or any such pattern? Has anyone done such kind of simulation before?

Thanks!

Jai

2

If you want that to happen dynamically, as a simulation
runs, then you need to use a potential where the spheres are
attracted to each other. LJ can clump particles (e.g.a water
droplet) if you use it at the correct thermodynamic state point.

Steve

3

Hi Steve,
Thanks for the response.

We want to simulate the pour in such a way that 1 spherical particle when it flows down the box should create 1 or more new spherical particles. Say if we start the simulation with 10 particles, when the pour completes it should become 100 particles forming some clumps. In this case easy spherical particle will create 10 new particles on its way down the box. Is this kind of model achievable in LAMMPs?

Thanks!
Jai

4

Hi Steve,
Thanks for the response.
We want to simulate the pour in such a way that 1 spherical particle when it

what "pour" of what? are you referring to the "pour" example for
granular particles?

flows down the box should create 1 or more new spherical particles. Say if

do you mean you want a single particle be replaced by two particles?
do the two particles have to have the same size as the single particle?
would it be a sufficient model to have the one particle just expand?

we start the simulation with 10 particles, when the pour completes it should
become 100 particles forming some clumps. In this case easy spherical
particle will create 10 new particles on its way down the box. Is this kind
of model achievable in LAMMPs?

a lot of things are possible in LAMMPS. in your case, you will most likely
have to program a suitable fix class to make it work.

axel.

5

Hi Alex,
Yes. I was referring to ‘pour’ example. It would be sufficient to have one particle just expand.

Jai

6

Hi Alex,
Yes. I was referring to 'pour' example. It would be sufficient to have one
particle just expand.

then check out: fix adapt using the "atom" attribute.

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

cheers,
     axel.

7

Hi Alex,

Thanks for pointing me to fix adapt command. I tried it and it is worked. The particle size increases as they slide down the slope. The simulation is close to what we expect.

But would like to improve it by trying the other option of single particle being replaced by two particles of same size and properties as the single particle. And also try to create some new particles on the sides or close to the already flowing particle at random and make them all flow together.

Can you please guide me on how about doing these.

Many Thanks!

Jai

8

Hi Alex,
Thanks for pointing me to fix adapt command. I tried it and it is worked.
The particle size increases as they slide down the slope. The simulation is
close to what we expect.
But would like to improve it by trying the other option of single particle
being replaced by two particles of same size and properties as the single
particle. And also try to create some new particles on the sides or close to
the already flowing particle at random and make them all flow together.
Can you please guide me on how about doing these.

for that you will have to do some c++ programming by yourself (or
hire/bribe/persuade/threaten somebody to do it for you). you have
to create a new fix class that will create those "atoms", similar to,
say, fix pour. it is no rocket science, but needs to be done carefully
so it works in parallel and blends in with the rest of lammps. the
basic instructions on how to do it are in the online manual.

cheers,
     axel.