Add CSTR with variable porosity

[jbreue16]

As per discussion in #211, we need to make porosity variable for the CSTR.

TODO

• Create new unit operation CSTRVarPor based on the CSTR unit operation. This is described in the dev guide; please also leave a comment or make changes to the guide.
• Adapt the interface: Change POROSITY to CONST_SOLID_VOLUME, which will be the constant solid volume of the tank. Porosity will be computed from that and not be exposed to the user anymore.
• Recompute the current porosity from the liquid and solid volume in the residual and Jacobian implementation of the unit

[AntoniaBerger]

Do we want to consider that the inlet of a cstr also has somethinkg like "porosity".

For example, if we put sand and liquid in a cstr? Or is the porosity of cstr only depending on the change of the liquid volume?

[jbreue16]

We can think of that in a second step, at first i would assume solid to be fixed and given by the user.

Iam not sure however if adding solid is actually done in practice, I dont think so.
Theres usually a fixed bed of particles thats usually expensive and is not pumped through the system but fixed in individual units.
Seems like a very special case to me and I'd wait and see if anybody actually wants to use that in cadet.

[AntoniaBerger]

In order to complite TODO 2 we distinguish between the state $V^{\ell}(t)$ and the parameter $V^{s} = const \in \mathbb{R}_{\geq 0}$.

Thus the equations for the CSTR change a little bit:
$$\frac{\mathrm{d}}{\mathrm{d}t} (V^{\ell} c^\ell_i) + V^{s} \sum_j d_j \sum_{m_{j,i}} \frac{\mathrm{d}}{\mathrm{d}t} (c^s_{j,i,m_{j,i}}) = F_{\text{in}} c^\ell_{\text{in},i} - F_{\text{out}} c^\ell_i + V^{\ell} f_{\text{react},i}^l\left( c^\ell \right)+V^{s}\sum_j d_j f_{\text{react},j,i}^s\left( c^\ell, c_j^s \right)$$

$$\frac{\mathrm{d}V^{\ell}}{\mathrm{d}t}= F_{\text{in}} - F_{\text{out}} - F_{\text{filter}}$$
Where the flowrates $F_{\text{in}} , F_{\text{out}} , F_{\text{filter}}$ are associated to state $V^{\ell}$.

I think, we would complicate the equation if we would write the first equation as a function of porosity: $$\varepsilon(t) = \frac{V^{l}(t)}{V^{l}(t) + V^{s}}.$$ As we would get an additional coupling of states through a product.

[jbreue16]

you can already push a branch with your current changes. After youve done that, we can open a PR and colaborate better.

[jbreue16]

Closed as per #245