[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ESPResSo-users] Espresso modifications and more....

From: Michael Winokur
Subject: Re: [ESPResSo-users] Espresso modifications and more....
Date: Mon, 29 Aug 2011 15:16:10 -0500

Hi Olaf,

1. According to the J. Chem. Phys. paper (125, 064103 (2006)) by Golubkov and Ren (a former student(?) of Jay Ponder), Tinker uses the PME algorithm but I can't verify it.  However the source code has the terms well laid out.  (I note that there seems to be quite a few algebraic errors and typos in the equations of this paper.) There is also paper by T.R. Walsh (Molecular Physics, Vol. 100, 2867 (2002)) which has a nice _expression_ for a central point quadrupole.

Tinker is nice in many ways but it doesn't use the quaternion formalism employed by Espresso.  To me, using the Eulerian matrix approach is much more cumbersome than the quaternion once one gets used to it.  I also like Espressos transformation from the body center-of-mass frame to the laboratory center-of-mass (although it took me quite a while to figure out what was going on when I first looked at the coding).

2. I'm currently using an older 2.1.2j version because it was the one I started with.  I was a little concerned that with each new version I  would be forced to reenter and revise the parameter number lists for interprocess communications and I didn't wish to redo this for each release.   It is way too easy to get out of sync and chase one's shadow as it were.  If you wish I could get the development version and see if I could integrate my revisions.  It will take me a few days as the semester is starting and I have other matters that will quickly take priority.

3. And about converting time steps to seconds? 



On Mon, Aug 29, 2011 at 2:29 PM, Olaf Lenz <address@hidden> wrote:
Hi Michael!

On 08/29/2011 09:23 PM, Michael Winokur wrote:
> Thanks for the quick reply.  Although it wouldn't be trivial it could
> be added.  Tinker has this coded in Fortran (which means I can
> almost cut'n'paste) and I could revise the Espresso dipole module to
> deal with both dipole and quadrupoles.  I'm not sure I want to deal
> with bound dipoles.

What algorithm does the quadrupole code use? Does it use some variant of

> I also note that I've revised all my Gay-Berne modifications this
> summer to better deal with anisotropic particle rotational inertias.
> I found and fixed some subtle mistakes but now the code appears to
> give the right physics using reported benzene  parameters with the
> thermostat on so that rotational and translational motion partition
> the kinetic energy evenly.  Would anyone be interested in these
> revisions?

In the last months, we have indeed looked into the issue of inertia. We
would be very interested in your modifications! Are you using the
development version of ESPResSo from git? If so, we can easily cooperate
via github.

> Simulation isn't as easy as it looks....

True words!


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

reply via email to

[Prev in Thread] Current Thread [Next in Thread]