Re: [ESPResSo-users] Hard Dimers in 2D

[Rudolf Weeber]

Hi,
On Wed, Jan 08, 2014 at 03:38:59PM -0600, Salvador H-V wrote:
> I want to simulate a two dimensional system of  hard tangent spheres
> (dimers).
> 
> I am interested in a rigid dumbbell, so, I need to avoid sliding of one
> particles on the surface of the other particle.
> 
> I was wondering what is the proper way to setup this system with ESPresSo.
> 
> I am considering two options:
> i) rigid bond with length bond equal to the diameter of the particles  +
> exclusion command
> 
> ii) Virtual sites to create a rigid object + ¿exclusion command?
If you only need to keep the distance between the two spheres forming the 
dumbbell constant, both will do the trick.
If you decide to use virtual sites, place the non-virtual particle in the 
center of the dumbbell, don't define any interaction for it, and vs_auto_relate 
the two spheres to it. Define LJ potentials for the two spheres.

The question of "sliding" is only relevant, if you couple something to the 
rotational degree of freedom of the individual spheres. Tis would be the case, 
e.g., if you assign the individual spheres forming the dumbbell a magnetic 
moment.
If that is, what you need, let me know.
Otherwise, you can just ignore the rotational degree of freedom of the 
individual spheres.

> Finally, I would like to know the reliability of the Langevin Dynamics
> thermostat, to calculate the dynamical properties of this system, such as
> the Mean Square Displacement of the center of mass of the dimers. What
> about rotational properties?
The rotational degrees of freedom follow a similar Langevin equation as the 
translational ones.
You cannot set the Langevin gamma for the rotation independently, right now. 
Should you need that, I have a (littlte tested) patch, which I can provide if 
need be.

The Langevin thermostat influences the dynamics of the system.
I think, at least in cases where the Langevin thermostat is used just for 
thermalization and not to model an implicit solvent, people equilibrate with 
the thermostat and then turn it off for the MSD measurement, but I have no 
personal experience with that.


Hope, that helps,
Regards, Rudolf