Re: [ESPResSo-users] multiple potentials on a polymer

[Olaf Lenz]

Hi!

On 01/08/2011 05:02 PM, Edward Kim wrote:
> By the way, Could I use an user-defined potential (which is not
> defined in Espresso)? If so, would anyone let me know how to?

You have two possibilities: either you use a tabulated potential (see in
the User's Guide), or you implement your own potential in C. To do that,
you should just look up the C-implementation of e.g. the Lennard-Jones
potential (lj.{c,h}) and see how it is integrated into ESPResSo. If you
do that, it would be nice if you could send us the new code so that we
can integrate it into ESPResSo.

> I have a newbie question. I'd like to apply more than two potentials
> on a polymer chain. It seems that syntax "polymer" doesn't have an
> option for multiple potentials. So, can I just make a script like
> below ? Which is correct, (1) or (2) ??

I actually wonder why you would want to combine a FENE and a harmonic
potential in the first place? The FENE potential is intended to be very
similar to the harmonic potential close to r0 (e.g. its derivative is
approximately equal). If you add them up, I don't see what you gain by
that. But I assume that you have your reasons.

>From a computational point of view, to be optimal, you should define a
new bonded potential that is a combination of both potentials, instead
of using two bond potentials, but the effects of this are probably
negligible.

I don't see where in (1) you try to activate the second potential? You
only define $bond_id_harmonic, but never use it.  (2) should work. Also,
you could mix both: the polymer command does nothing else than creating
a number of adjacent particles, so you can simply loop over them
afterwards and add the second bond. Otherwise, the script looks OK.

Olaf

-- 
Dr. rer. nat. Olaf Lenz
Institut für Computerphysik, Pfaffenwaldring 27, D-70569 Stuttgart
Phone: +49-711-685-63607

olenz.vcf
Description: Vcard