Pump kinematic viscosity/dynamic viscosity

[RaphaelGebhart]

Corrected error dynamic viscosity vs. kinematic viscosity and put an example in there that might help you for further discussion
closes #139

[RaphaelGebhart]

The error (mixed kinematic/dynamic viscosity) caused the major difference at design point, but at the moment im not yet really sure if it will remove all issues (at design point there is still a difference of about 3% in volume flow rate for a give pressure difference and of about 5% in pressure difference for a given volume flow rate).

First of all, the approach used in Thermofluidstream.Processes.Internal.ThermofluidStream.Processes.Internal.TurboComponent.dp_tau_centrifugal is designed to scale pump maps (messured with water) to other fluids (like oil).
The scaling factors for volume flow, head and efficiency according to Gülich, Kreiselpumpen: Handbuch für Entwicklung, Anlageplanung und Betrieb, 3. Auflage, Chap. 13.1, Tafel 13.1 Berechnung der Kennlinien für viskose Medien do not tend to 1 for the reference fluid itself! Therefor this approach does not seem to be most useful in the simple case, i. e. if you would like to use the pump for water.
Besides, there are various scaling approaches in the literature, where ISO/TR 17766:2005(E) (also mentioned by Gülich), might, in my opinion, be the first choice.

For the simple case (using e.g. a water pump map and a water medium model), i suggest to implement a version similar to MSL.Fluid, which im already working on.

[tobolar]

It seems the original definition of mu_in (i.e. = Medium.dynamicViscosity(state_in)) is now done for newly added eta_in.

For this, a conversion like convertElement("ThermofluidStream.Processes.Internal.TurboComponent.dp_tau_centrifugal", "mu_in", "eta_in"); shall be enabled.
This in turn shall be only done for the next major library version. So I added a corresponding milestone.