Move NBOUND parameter from discretization to unit_xxx/adsorption parameters group

[schmoelder]

Moving the parameter NBOUND from the discretization group (where it is needed in the code) to the unit_operation parameters group makes more sense in the user interface.

See also #18

[Immudzen]

I would be okay with reading it from either location but I am against just moving it since you break all existing simulations and gain nothing in the process.

[schmoelder]

I understand your point. And for now being able to read it from both places seems reasonable to me.

For consistency's sake I still think it should be moved but maybe this can be done at some point in the future when other breaking changes are also introduced.

Just for some background, in CADET-Process the configuration of discretization parameters is separate from the model parameters. Hence it leads to some inconveniences when finally setting up the h5 file. Nothing too dramatic, just has always bothered me. ;-)

[schmoelder]

Actually, it should probably be moved to the adsorption parameters group since it sets the expected length of the isotherm parameters.

[lieres]

I agree with Bill that we should be very careful with breaking changes unless absolutely neccessary.

[schmoelder]

I also found that the number of particle types NPARTYPE is set in the discretization branch.

While I agree that we should definitely not introduce breaking changes (either by implementing a file version converter or by simply looking at the old location if parameter is not found), the location of these parameters is somewhat unintuitive.
I would rather keep the physicochemical model separate from the numerical discretization. This is especially relevant since now the configuration of adsorption models and multiple particle types is different for the CSTR and the column models.