Re: [ESPResSo] Questions about Thermostat

[Torsten Stuehn]

Hi Lorenzo,

> I have been using Espresso to simulate the process of soot
> nanoparticles agglomeration (so, strictly speaking, NOT for molecular
> dynamics) in a quiescent fluid.
> So far I have used the Langevin thermostat to account for collisions
> between fluid molecules and nanoparticles.
> I really have two separate questions for the list.
> (1) What is exactly the effect of having the rotation feature enabled
> in the configuration file? In what sense does it modify the Langevin
> thermostat?

If ROTATION is switched on, the number of degrees of freedom per particle
is 6 (3 translational + 3 rotational ones) compared to 3 (translational only)
if ROTATION is switched off.
The frictional and the stochastic force term of the langevin thermostat are
applied to each degree of freedom of the particles.
The kinetic energy of the system is calculated by summing up the
translational and the rotational kinetic energies (if ROTATION is switched on)
of all particles. This must be taken into account when calculating the
temperature of the system:

     2 * E_kin
T =  ---------
      dof * N

where dof=3 (ROTATION switched off) or dof=6 (ROTATION switched on) is the
number of degrees of freedom of each particle.

> (2) In the simulations I ran, I start with a certain number of
> monomers (i.e. the initial particles I place in the computational box)
> which stick upon collisions thus giving rise to aggregates.
> The diffusion coefficient of the resulting aggregate does not take
> into account the fact that a monomer deep inside the aggregate is
> "screened" by the surrounding monomers and hence feels a different
> noise with respect to a "free" Brownian monomer.
> I wonder whether the DPD thermostat in Espresso can somehow account
> for that, but I am not sure it is a reasonable choice and would like
> to know the opinion of someone on this list.

I'm not sure, if I understood your question correctly. Do you want to
apply different DPD prefactors (gamma and temperature) to particles
whether they are inside or outside of aggregates ?
In the current ESPResSo version this is not possible. There can only
be one DPD thermostat acting on all the particles of the simulation.
The only exception so far are "fixed" particles (and their close neighbours)
which are not at all effected by the DPD thermostat. But fixed particles
do not move at all and thats probably not what you want.


Greetings,
Torsten.