#6 Start values

[Michael M. Tiller]

The "Modelica.Mechanics.Rotational.Interfaces.Compliant" (partial) model defines a start value for the
internal variable phi_rel. This is bad because this effects the state selection in Dymola. Specifying a
start value of "0" is redundant (since that is the default) unless your intention is to affect the state
selection. The problem is that the current heuristics make this an irreversible process (i.e. I cannot go
back and say "wait, no, I don't want that to be a state"). Clearly, some of this issue lies with the state
selection algorithm in Dymola. However, I still think that providing a start value of zero is redundant and
potentially confusing and that it should be removed.