InternalWellboreGenerator issue with theta=360°

Describe the bug
The full wellbore mesh (theta=360°) generated by InternalWellboreGenerator works well with single processor. However, it does not work when running the simulation in parallel.

To Reproduce
Steps to reproduce the behavior:

1. Rerun one of the wellbore smoke example after changing theta to 360°: consider two runs, the first one uses a single processor, the second one consider a parallel simulation (mpirun with -x 4 -y 4 -z 1).
2. Observe that the displacement field obtained by the first run is continuous at theta=360° (that is correct). But, the results obtained by the second run is discontinuous at theta=360° (because the faces at theta=0 and theta=360° are not connected when considering mpirun)

Expected behavior
Results obtained by single processor and multi-processors must be identical.

Screenshots
This figure show the discontinuity of uy at theta=360° when running a wellbore problem using mpirun with -x 4 -y 4 -z 1. This error is because the faces at theta=360° and theta=0 are not connected.

Platform (please complete the following information):

• Machine [e.g. XRAI TotalEnergie]

Additional context
We will need to verify the connexion of the mesh partitions when using InternalWellboreGenerator to ensure that the faces at theta=360° and theta=0 are connexted. This issue does not affect any of the examples because most of these examples are set with theta=90° or 180°. The case of cased elastic wellbore has theta=360° but it is a smoke case so we do not need to use multi-processor.

Start with InternalWellboreGenerator and rerun a wellbore smoke example with theta=360° in single-processor and mpirun -x 4 -y 4 -z 1 configurations. Inspect how mesh partitions connect the theta=0 and theta=360° faces. Done means the parallel displacement field is continuous at theta=360° and matches the single-processor result.