Re: [h5md-user] Volume

[Pierre de Buyl]

Hi,

On Wed, Nov 06, 2013 at 04:41:32PM +0100, Felix Höfling wrote:
> Am 05.11.2013, 16:25 Uhr, schrieb Konrad Hinsen <address@hidden>:
> 
> >Looking at the thermodynamics module, I notice that the variable
> >"volume" is obligatory. That raises the question of what to do for
> >systems that have no volume, because they are unbounded in at least
> >one dimension.
> >
> >One possible answer is "don't claim to use the thermodynamics module",
> >but I wonder if that's really the intention.
> 
> When I put this in the module draft, I briefly thought about this.
> My conclusion was that thermodynamics in an unbounded volume is not
> well defined. Although there are certainly extensions around to cope
> with this case, e.g., to describe supernovae explosions ...
> 
> The motivation for making the volume obligatory was to guarantee the
> conversion between extensive, per-particle, and per-volume
> quantities. I have no strong opinion about that, we may move it back
> to the optional elements (and replace it by "density" being an
> intensive variable?)

The problem is that, when we are pushing the boundaries of science [1], we find
ourselves in situations where we cannot convert everything (semi-open systems,
subsystems, etc.).

I am in favor of not having 'volume' mandatory. And neither 'density'.

The alternative is, as Konrad suggests, that the 'thermodynamics' module would
be less used.

Pierre

[1] Which we are doing on a regular basis, indeed :-)