Weird enthalpy in the last node in Dynamic[One/Two]PhaseFlowPipe

Issue

Description

In both DynamicOnePhaseFlowPipe and DynamicTwoPhaseFlowPipe models it looks like there is one too many node or at least h[end] assume weird values which does not seem to affect the calculations (it looks like h[end] is not connected to `C2.h` for example). This problem was first seen by @damien.faille. This is the case at least in both Fluid and WaterSteam packages.

Steps to Reproduce

This issue appears in the examples:

• SimpleExample.TestDynamicDrum_CC: see TubeEcranBoucleEvaporatoire.h in a discretized plot.
• SimpleExample.TestDynamicOnePhaseFlowPipe: see dynamicOnePhaseFlowPipe.h in a discretized plot (see picture below).

Logs and Screenshots