Some tunable parameters unstable

[TomasKulhanek]

Some tunable parameters unstable in BloodyMary example.
e.g. system.T_ambient,system.p_ambient

changing them may cause simulation to go to unstable weird condition.

reproducible in https://bodylight.physiome.cz/Bodylight-Scenarios/bloodgases/#komplexnisimulator.md

Example moving pressure from 102 kPa to 91 kPa cause simulation to fail

.

[filip-jezek]

In fact, the simulation does not fail, but takes ages to recover - observer the console. Perhaps better handling of the reinitialization? In dymola the initialization takes about 1s, so this might be a lead.

Can you perhaps identify another troublesome parameters? These two are everywhere and thus hard to pinpoint the exact cause.