Re: [ESPResSo] Basic Questions

[Axel Arnold]

Am Mittwoch 18 März 2009 00:46 schrieb Salvador Herrera Velarde:
> Dear All,
> I have some basic questions about ESPResSO that I couldn't figure out the
> answer.
>
> I am writing some basic scritps for 3D colloids (beads) that interact via
> the repulsive part of Yukawa potential (Debye-Huckel in ESPResSO). Righ
> now, I can calculate the radial distribution function, the structure factor
> and the MSD.
>
> a) First, I want to study structural and dynamical properties of these
> systems in 2D. What I did to simulate a two-dimensional was:
> #   --- The simulation Box is periodic in both  directions
> setmd periodic 1 1 0
> #   --- Simulation Box
> setmd box_l $bl $bl $bl
> & during the cycle to place the particles inside the simulation box:
> part $i pos $posx $posy $posz type 0 q $zeff fix 0 0 1
> so, my question is: Is this the best way to model a 2Dim system in ESPResSO
> ? Should I do something similar for 1D?

Basically, there is no other way to do that; you cannot recompile the code 
with less dimensions for positions or so.

> b) Also, I am interested in binary mixtures of Yukawa colloids. However, I
> don't know how to specify the interactions for mixtures in ESPResSO. For
> potentials like Lennard-Jones you can specify the type of the particles: 
> inter type1 type2 lennard-jones... but... how I can do the same for the
> Debye-Huckel potential ?
> It is possible to specify different Bjerrum lengths for the species?

No, the Bjerrum length is a global constant. You can only assign different 
charges to the different colloid types. Typically, you also do not have 
different Bjerrum lengths, but different (effective) colloid charges. If you 
cannot formulate your Yukawa-prefactor as lb*q_i*q_j, then you can think of 
implementing a new short ranged yukawa-interaction. Just grep in the source 
tree for e.g. SMOOTH_STEP (which is what you define to get the smooth_step 
potential), and you can quite easily get how to do that.

> c) Finally, I am using the example provided in the user's guide to
> calculate the average of the gr and Sq. But (sometimes) when I stored many
> files configurations (to improve the statistics), I have some problems when
> I tried to calculate the averages of the gr & Sq. The result during the
> average is the next:
>
> missing operand at address@hidden
> in expression "0 + address@hidden"
>     (parsing expression "0 + ")
>     invoked from within
> "expr $ai + $bi"
>     (procedure "vecadd" line 3)
>     invoked from within
> "vecadd $avg_rdf $rdflist"
>     ("for" body line 34)
>     invoked from within
> "for {set i 1 } { $i <= $nfiles } { incr i }  {
>    set filename Pos$i
>    set f [open "$filename" "r" ]
>    while { [blockfile $f read auto] != "eof" } {..."
>     (file "Debye-H.tcl" line 463)

Hmm, no idea. It seems that rdflist is somehow mixed up. Perhaps, you print 
out the rdflist before adding it up? Maybe you can see what is wrong with it.

Axel

-- 
Dr. Axel Arnold 
Fraunhofer SCAI
Schloss Birlinghoven, 53754 Sankt Augustin, Germany
Tel: +49 2241 14 2575