Is it necessary for svZeroDSolver to require explicit direction for the flow problem?

[zasexton]

Currently svZeroDSolver requires that junctions provide the inlet vessels and outlet vessels for setting up a 0D simulation in the input file; however, direction of the flow field should be part of the solution? In the case where we are building 0D models from patient-specific 3D analogs the connectivity of the 0D model (i.e. explicit ordering of inlet vessels and outlet vessels) may already be known and trivial, but for cases with internal loops this may be unclear to a user and not known a priori. Are we being too restrictive in the problems we can handle by preserving this convention or are there other reasons why we require this?

[mrp089]

It's not necessary to differentiate between inlets and outlets, as you can see in the code for the Junction. In the equations, we only enforce conservation of mass and that all pressures are equal.

For the BloodVesselJunction, we actually differentiate between one inlet and $n$ outlets. In the schematic, every outlet is connected to the inlet via a BloodVessel. To make that invariant to inlets/outlets, we could redefine it as every connection having a BloodVessel element that all connect at one central junction node.

[mrp089]

Opened #167 to address this