Re: [ESPResSo] analyze pressure and energy problems

[Olaf Lenz]

Hello!

İbrahim İnanç wrote:
> PID 24522 failed on node n0 (127.0.0.1) due to signal 7.

Signal 7 is a bus error - very strange. How did you compile ESPResSo, 
and on what platform?

> I want to simulate polybutadiene with 32 chains of 128 carbon atoms (32 
> rep. unit x 4). I changed the units in order to have the real atomic 
> units. But somehow when I put analyze energy command it gives the below 
> error message specially for higher integration steps or higher 
> temperatures..
> /
> domain error: argument not in valid range
>     while executing
> "expr [analyze energy kinetic]"
>     ("while" body line 8)
>     invoked from within
> "while { $i<$n_cycle} {

Can you please send the ESPResSo-script that you have been using? It is 
very hard to debug the problem as long as we don't know any more facts.
From the description, I would guess that the problem is that you try to 
use atomic units.
ESPResSo works best when all numbers are in the order of 1. When you use 
parameters that are several orders of magnitude below or above 1, it 
might happen that the algorithms become numerically unstable, or even 
that some heuristics used by ESPResSo to check the validity of the 
parameters don't work.
This means, that you should *not* use SI units when doing a molecular 
simulation, instead use the more natural units, like Angström for 
lengths, and fs for times.

Cheers
        Olaf