[gmx-users] Re: What is epsilon_r?

[Andrew DeYoung]

Dr. van der Spoel,

Thanks for your time!  If you have time, I am wondering about your comment
that for  epsilon_surface, "zero means in fact infinity and this is the
default."  Does this mean that there is an error in the manual?  The
manual's entry for epsilon_surface
(http://manual.gromacs.org/current/online/mdp_opt.html#ewald) says, "The
default value of zero means it is turned off", which seems to imply that the
default is "no surrounding dielectric" rather than "surrounding dielectric
constant of infinity."

Also, in looking at the paper on which the 3dc is based, it seems that your
memory is correct:  the 3dc only affects the dipole correction.  Still, the
manual's promise that epsilon_surface does not affect the slab 3DC variant
of the long range corrections is true.  I guess that this means that the
contribution of epsilon_surface is turned off in the code when the 3dc is
used?  

Thanks!

Andrew DeYoung
Carnegie Mellon University

> No, zero means in fact infinity and this is the default. Only for very
special cases should you change this.

> The 3dc only affects the dipole correction if I remember correctly. If you
use it a refernce to a paper appears in your log file.