[gmx-users] 2-D liquid

[Mirco Wahab]

On 15.10.2014 06:06, Cai wrote:
> I am trying to simulate 2-d liquids interacting with simple potential like
> Lennard-Jones potential.
> Can it be done by specifying "pcoupletype = semiisotropic" in the .mdp
> file? I mean enforcing a normal pressure in x-y direction while very high
> pressure in z direction to constrain the motion in z.

You have a lot of problems to solve when emulating 2D from 3D, eg.
- periodicity
- self-interaction of the pair-force
- particle/wall-interaction (if using walls)
and more.

> Or do I need to define walls in the z direction?
> Does anyone have similar experience? Your help are highly appreciated!

I'd think it's not possible or you have a pseudo-2D which looks
like 2D but isn't.

Here,
http://www.ph.biu.ac.il/~rapaport/websim/moldyn2d.html
is probably all you'll need for that. Click on the colored dots!

Regards

M.