Re: [ESPResSo] wall constraint use

[Axel Arnold]

> > 3.) Why do two normal [0 0 +/-1] wall contstraints at dist z=0 and
> > z=10 in a 10x10x10 box give a constraint violation error for one
> > particle which starts at the center [5 5 5]?
>
> Because the distance to the origin (z) is measured in multiples of
> the normal, hence you need to put the second wall (top) at z=-10. In
> fact, I don't know, if it is okay to put the wall directly at the
> simulation box boundaries at all.

It should be ok, provided particles do not manage to tunnel through the 
repulsive potential. In this case, their coordinates will be folded, i.e. the 
come out at the opposite wall instead of crashing the simulation with a 
constraint violation.

> > 4.) Can the wall constraint be placed (in pairs if necessary) at [5
> > 5 5] in a 10x10x10 box with particles above or below the pair?
> No, see 1.

However, a double sided wall might be desirable. It would be nice if you can 
take a look at the current wall constraint and implement an extension for a 
double-sided wall.

> > 5.) Does the wall constraint work with analyze energy?
>
> I don't know.

The interaction energy of the wall with the particles is simply added to the 
nonbonded energy of the particle and wall types. If the wall has a distinct 
type different from the particle types, you can isolate its interaction 
energy via "analyze energy nonbonded <particletype> <walltype>".

Axel

-- 
Dr. Axel Arnold 
Fraunhofer Institute for Algorithms and Scientific Computing (SCAI)
Schloss Birlinghoven
53754 Sankt Augustin
Germany