Re: [ESPResSo-users] Basic Question about using LB

[Stefan Kesselheim]

Hi Salvador, hi Espresso-users,

Am 23.04.2012 um 20:26 schrieb Salvador H-V:

> Hi,
> 
> I am trying to simulate a quasi-two dimensional mixture suspension composed 
> by rigid dumbbells and spheres.

This brings up immediately the question how you want to model the dumbbells and 
spheres. You can make the dumbbells out of several particles connected by bonds 
that are stiffened with a bond angle potential and spheres just by single 
particle. This if course limits the accuracy as the "point coupling" scheme 
only coarsely represents the corresponding particles. Other possiblities would 
be the raspberry model (Lobaskin et al) or the "wall" model as used in most of 
the LB community (first by Ladd ...). 

> 
> I am planning to use LB fluid and LB as a thermostat. However just to be on 
> the safe side, I would like to ask you the following:
> 
> a) It is possible to use LB fluid for a no-cubic box. I want to use a box 
> with different length in z-direction. I just want/need a layer of fluid in 
> z-direction to cover the particles.

Yes. Noncubic is fine.

> b) Because of the periodicity of the system, I was wondering if LB fluid can 
> handle partial periodicities (1 1 0) 

What would be the boundary condition in z direction, if not periodic? A wall 
can be created with the lbboundary command. Other BCs are not implemented, as 
pressure BCs for example are nontrivial in LB.

> c) Finally, there are experimental results  where a uniaxial anisotropic 
> particle is characterized by two translational hydrodynamic friction 
> coefficients, Da and Db, respectively, for motion parallel and perpendicular 
> to its long axis. Could you share some thought about it will be possible to 
> observe such behavior using LB fluid.

One possibility would be to determine the Diffusion tensor (where Da and Db are 
the diagonal elements) from the mobility tensor by applying a force in the 
corresponding direction and constraining the system  (e.g. dumbbell axis 
parallel to z-axis) and measuring the mobility. The mobility then is related to 
the Diffusion tensor by the fluctuation dissipation theorem, which just states 
that D = k_B T mu. Please make sure that you apply a body force on the fluid so 
that the net force on the system is zero then.

It should also be possible to use a Green-Kubo relation measure it from 
fluctuation of the system in equilibrium.

I hope this helps a bit. If you have more questions don't hesitate to ask, I'm 
just a bit busy and my answer might be a bit briefer than necessary :-). 

Cheers and good luck,
Stefan

> 
> Thanks a lot in advance,
> 
> Salvador H-V
>    
> 
> 
> 

Stefan Kesselheim
Institute for Computational Physics
University of Stuttgart
address@hidden