Add laminar flow to manning

[verheem]

No description provided.

[visr]

This seems like a nice improvement, replacing the current hardcoded threshold for the relaxed root with a dynamic one based on the Reynolds number. We can keep #2471 open in case we also want to add e.g. a DarcyWeisbachResistance node.

@Huite what do you think? Based on testing real models this does not affect performance much, but should be easier to explain.

This still need an update of the ManningResistance docs at https://ribasim.org/reference/node/manning-resistance.html, after "Therefore we use the modified expression"

[Huite]

@Huite what do you think? Based on testing real models this does not affect performance much, but should be easier to explain.

Looks good to me! As a sanity check, it might be worthwhile to see for e.g 10 sets of connection parameters (slope, width, etc.) how the new threshold compares to the old.