Issues with start values in ThermoSysPro

[Francesco Casella]

During today's Board meeting, I discussed with Daniel Bouskela of an issue that they are frequently encountering with ThermoSysPro models. All the models in the library work with Dymola, but many of them do not work in OpenModelica because of some missing start values.

One way to avoid this issue is to make sure that we expoit the information which is implicitly given by the modeller when adding start values in the model, i.e., to prefer those variables as iteration variables when using tearing. We already did something along these lines in #5458, but apparently this is not enough.

For example, the model ​ThermoSysPro.Examples.Book.SimpleExamples.HeatExchanger.TestDynamicTwoPhaseFlowPipe fails during initialization, while it works with Dymola.

Daniel suggests that we implement a feature to load the missing start values from an existing solution (e.g. a .mat result file obtained from Dymola) and automatically add modifiers to the model to ensure convergence. This can be considered as a last resort measure, which can useful in a context where you have a few Dymola licenses and you want people to reuse those models with open-source tools.

I would personally prefer to understand first if we can avoid these failures by exploiting all the available information in the original Modelica model.

Karim, Andreas, if you are interested you could have a quick look, by comparing how Dymola solves the equation where OMC fails.

[Francesco Casella]

Retargeted to 1.17.0 after 1.16.0 release

[Francesco Casella]

Retargeted to 1.18.0 because of 1.17.0 timed release.

[Francesco Casella]

Ticket retargeted after milestone closed