|Subject:||Re: [ESPResSo-users] DPD|
|Date:||Thu, 11 Sep 2014 11:42:53 +0200|
On 09/10/2014 07:02 PM, Dudo wrote:
> On Wed, Sep 10, 2014 at 7:09 PM, Christoph Junghans <address@hidden
> <mailto:address@hidden>> wrote:
> 2014-09-10 9:08 GMT-06:00 Dudo <address@hidden
> > Please,
> > should I generate additional particles for using the DPD thermostat?
> I don't really understand why you would do that. Can you give some
> details about the problem you are studying?
> Hi Chris,
> well, what I would like to do is to drag a polymer molecule through media,
> with a constant velocity with respect to the media as possible.
> Hence, even if I fill the periodic box with a huge number of "solvent"
> bodies it would start moving as one phase after some time.
> So this is not the way.
But that is physical. Assuming you intend to drag the molecule by
applying a force, this will 'pump' momentum into the system which will
diffuse according to the viscosity of the medium. Unless you dissipate
momentum by some means (e.g. walls), the whole system will accelerate.
If you stop applying the force at some point, velocity gradients will
decay while total momentum of the system is conserved.
> What I would like to see coming from my simulation, is at first the
> molecule orienting with its lowest energy/lowest friction or
> hydrodynamic radius profile with respect to the direction of the
> movement through the media, with the molecule started re-shaping after a
This sounds like you intend to look at the response of the molecule to
an external flow. In that case, you need to decide first what kind of
flow you would like to impose. Simple examples might be uniform, linear
or parabolic flow profiles. However, if you are merely interested in the
mobility tensor of the molecule, you may be able extract it from
equilibrium simulations by means of linear response theorems (keep
finite size effects in mind for a periodic system).
Note that such properties have been studied extensively for various
kinds of polymers (linear chains, stars, rings), and you may want to
take a second look at the available literature.
Ulf D. Schiller
Centre for Computational Science
University College London
20 Gordon Street
London WC1H 0AJ
|[Prev in Thread]||Current Thread||[Next in Thread]|