Further development of pump and compressor models

[dlohmeier]

It appears that there is quite a variety of how pressure increasing units are modeled in pandapipes. Technically, there are pumps for liquids and compressors for gases, but the underlying physical and control model is very important as well. Therefore, we should decide on how to model these units and how we would like to distinguish aspects such as "Is it a pump or a compressor?", "Does it have a curve or a fixed pressure ratio?" and "Would it matter whether the pressure is increased or decreased?" All these aspects can be important in reality, but most components in pandapipes don't clearly distinguish between all of these aspects. Some notes:

• pressure increasing units could have very different behavior, e.g. according to a pump curve, a fixed pressure ratio, a fixed outlet pressure, a fixed pressure lift. These kinds of behavior are spread among different components in pandapipes
• a pump is a pressure increasing unit that implements a pump curve (dp in dependence of vdot), it doesn't distinguish between liquids and gases, only the implemented pump curves in std_types are not suited for gases --> can it cover pressure decrease?
• a compressor is a pressure increasing unit with a fixed pressure ratio between inlet and outlet, maybe using it with liquids could be problematic, but not generally forbidden --> can it cover pressure decrease?
• a pressure control sets the outlet pressure, so it could increase or decrease the pressure and is not related to the phase of the fluid. Usually it should model a pressure controlling valve, but it could be used as a pressure controlling compressor / pump as well.
• circulation pumps can implement either a fixed pressure lift or mass flow and set the outlet pressure as well, but we haven't found a perfect way to ensure mass flow balance with circulation pumps yet.

How do we deal with these inconsistencies? What would be a good way to ensure that all components are used correctly? How can we combine physical models with control models and ensure that on the one hand we don't need a new component for each possible control strategy and on the other hand the pump / compressor tables don't get too large to deal with? Should circulation pumps be treated as normal pumps or does their specialty of being used in closed systems make them less suited for an integrated model?