Question about how to put interaction with certain groups to 0

Hello everybody,

I am trying to simulate a polymer confined between two plates (carbon).
At the beginning I want the polymer to be attached to the bottom plate only. After the polymer settles on the bottom plate it will interact with both the top and the bottom plates under different conditions.

So, I created a system with the polymer between the top and the bottom plates.
I want to euillibriate the system upto the point when polymer makes a monolayer film on the bottom layer.
So my question is how I make the polymer interact with atoms in bottom plate only.
I know I can put them in two different groups ‘polymer’ ‘top’, but how do I make pair interaction ‘0’ ?

I am using cvff for the interactions, and it has all difefernt types of interactions.

I want to keep both the plate fixed ( setforce 0) during the simulation.

Thanks for any replies.

Mani

Hi Mani,

So my question is how I make the polymer interact with atoms in bottom plate
only.
I know I can put them in two different groups 'polymer' 'top', but how do I
make pair interaction '0' ?

I don't think you want their interactions to be zero, otherwise the
polymer will not "feel" the existence of the upper wall and may just
pass through the upper wall as if there is no wall at all. Instead,
you need a repulsive wall to prevent polymer interact with the upper
wall. Try fix wall/reflect or fix wall/lj93.

Ray

Yes.
One tiny addition:
You don't have to use any of the "fix wall" commands to create a wall,
by the way. Instead you can prepare a data file with atoms which
belong in the wall represented explicitly. To immobilize these atoms
atoms in LAMMPS, exclude them from the group to which you apply fix
nvt or fix nve. If you have any questions how to do this, I can
elaborate.

  This way you can give the atoms in the wall any properties you want
them to have (like charge).

Check out how to define a group using this command:
http://lammps.sandia.gov/doc/group.html

...and how to use that group with fix nve or nvt here:
http://lammps.sandia.gov/doc/fix_nve.html
http://lammps.sandia.gov/doc/fix_nh.html

Cheers
Andrew
(I was looking up the lammps command documentation, and I keep getting
"request exceeds maximum bandwidth limit". Is sandia.gov under
attack today?)

Andrew,

Thanks, but this alternative will not work for Mani's purpose of
preventing the polymer interacting with the upper wall. What do you
immobilize in this case? The polymer or the surface of the upper
wall? Even with proper immobilization, it does not prevent
interaction between them.

Ray

Andrew,

Thanks, but this alternative will not work for Mani's purpose of
preventing the polymer interacting with the upper wall. What do you
immobilize in this case? The polymer or the surface of the upper
wall?

Umm... the wall?

I'm sorry. I was actually not thinking about Mani's question at all.

I'm afraid I steered the conversation off topic. I was just talking
about how Mani immobilizes the wall atoms.
(Mani is currently using "fix setforce 0" to keep his wall atoms from
moving. So to accomplish this, I personally just omit the immobile
atoms from the group that I pass to fix nve or nvt. You (Ray)
suggested using "fix wall" commands as an alternative. That could
work. I don't have strong opinions about this and I did not mean to
get side-tracked.)

Back to the main topic (I hope):

Mani are you just trying to turn off the attractive interactions with
the upper wall?

(If so, check out http://lammps.sandia.gov/threads/msg05763.html, or
perhaps http://www.moltemplate.org/lammps_code/pair_lj_charmm_coul_charmm_inter.html.
Code available.)

Also, are you running your simulations at constant volume or pressure?

..I'm actually a little confused myself
Hope I didn't confuse anyone else.
I'll be quiet on this thread (until emailed).
Cheers
Andrew

Dear Andrew,

I'm afraid I steered the conversation off topic. I was just talking
about how Mani immobilizes the wall atoms.
(Mani is currently using "fix setforce 0" to keep his wall atoms from
moving. So to accomplish this, I personally just omit the immobile
atoms from the group that I pass to fix nve or nvt.

Thanks for sharing - I guess this is a nice alternative to setforce 0
since it does not integrate the forces on atoms then set calculated
forces to zero.

Thanks,
Ray