Restructuring of Media

[mwetter]

The current implementation of Annex60.Media divides the media in how they are modeled (e.g., IdealGases) and then in the type of the media. This leads to the following abbreviated structure:

|-- ConstantPropertyLiquidWater
|-- ConstantPropertyLiquidBrine (not yet implemented)
|-- GasesConstantDensity
|   |-- DryAir
|   |-- MoistAir
|   |-- MoistAirUnsaturated
|-- GasesPTDecoupled
|   |-- DryAir
|   |-- MoistAir
|   |-- MoistAirUnsaturated
|-- IdealGases
|   |-- DryAir
|   |-- MoistAir
|   |-- MoistAirUnsaturated

I suggest to reimplement this as

UserGuide
|-- Water
|   |-- ConstantProperty
|-- Brine (to be implemented, with different amounts of solution)
|   |-- ConstantProperty
|-- DryAir
|   |-- ConstantDensity
|   |-- IdealGas
|   |-- PTDecoupled
|-- MoistAir
|   |-- ConstantDensity
|   |-- IdealGas
|   |-- PTDecoupled
|-- MoistAirWithFogRegion
|   |-- ConstantDensity
|   |-- IdealGas
|   |-- PTDecoupled

The rational is that users are looking for a medium, such as moist air, and then decide on what physics it need to capture. This renaming would also avoid the clumsy name MoistAirUnsaturated, which should be the default medium for reasons of numerical efficiency.

To avoid a repetitive formulation of what physics is implemented in the medium, a user guide would describe the physic of the respective media.

If a library of HVAC components would have specialized models for air, water, and brine, they could mirror the structure of the medium and assign a default media. For example, there could be

|-- Fluid (with unspecified medium)
    |-- Sensors
    |-- FlowMachines
|-- MoistAir
    |-- Sensors
    |-- Fans
|-- DryAir
    |-- Sensors
    |-- Fans
|-- Water
    |-- Sensors
    |-- Pumps

Such a hierarchy may be autogenerated, except for the info section.

[mwetter]

Based on the phone call of Nov. 21, 2013, we attempt the following structure

UserGuide
|-- Water
|   |-- ConstantProperty
|-- Brine (to be implemented. Per to send models)
|   |-- ConstantProperty [may be variable density, based on what IDA is using]
|-- MoistAir
|   |-- Incompressible [with density as a function of T. 
                                  Per to send model. 
                                  This used to be ConstantDensity.]
|   |-- PTDecoupled   [keep unless we can show that 
                                 Incompressible computes as well]

Ruben reported that density as a linear function for air is sufficiently accurate from -10 to 80 degC.

DryAir will be removed as there is no compelling use case.

[mwetter]

The Water and Air media have been implemented, reviewed and merged to the master.
What remains to be done is implementation of Brine solutions with user-selectable concentration.

[mwetter]

Closed because we have separate ticket for brine.