Fluid connectors between MSL and TFS

[hubertblervaque]

I am trying to build a plug between the MSL fluid ports and the TFS fluid connectors (inlet, outlet and undirected). I developped two approaches: by assembly of "boundary" components or by a system of equations. In both cases, I have difficulty to have a robustess plug working for all the unit tests I defined.
Is there a fundamental conflict between MSL and TFS approaches that makes it impossible to create a plug between the two fluid connectors?
Has anyone ever built such a plug (for instance for other library)?

[dzimmer]

The fundamental approach between TFS and MSL is indeed different. It is best to read the suggested literature to fully understand this. TFS interface is created such that no large non-linear equation systems shall spread across components if used properly.
Developing an interface is not impossible though but needs to be done with care. In any case the interface makes only sense to connect different sub-systems not on the component level.
We are currently developing and testing such an interface. When it is ready we will make it part of the TFS Library.

[hubertblervaque]

Dear Dr Zimmer, thank you for your answer. It was a pleasure to read the literature about ThermoFluidStream, especially because the concept of inertial pressure is far from common.
I spent a couple of days on developing and testing my "working in some cases" plug. It would a pleasure to take part in the development of your solution. Feel free, you or a member of your team, to contact me at hbe@modeliconseil.com to see whether my contribution could be useful.

[dzimmer]

Everyone is welcome to contribute. Feel free to offer a link to a fork or raise a merge request. We can then review and decide. Also test cases can be of interest to us. Simply keep in mind that we need to have our own internal schedule.

[FrancNep]

Hi,

I think that the best way is to develop a model of Volume (based on MSL or TFS Volume ) integrating both bi-directional interfaces from the MSL and TFS library. It's simple, robust and performance.

Best regards,

[nieweber]

Hey,

thank you for your suggestion!

Indeed that is also the way we are internally developing an adapter between TFS and MSL. We are still conducting some performance tests and are working on a guideline for the modeler as the initialization of the volume and and the mass-flows on MSL and TFS side need to be appropriate in edge cases for very small volumes.

If you did some tests on your own, it would be great to have a look!

Best regards
Niels