check_watertight needs default value of tolerance to check model for water tightness

[makeclean]

check_watertight, currently when no tolerance is provided defaults to zero, there should be either

• a default value of tolerance to check against
• determined by the h5m file input

[pshriwise]

Wouldn't we just use the faceting tolerance as taken from the .h5m file. Or perhaps an order larger? (10xfacet_tol?)

[makeclean]

Yes, or possibly down to dags facet intersection tolerance 1.e-6

[pshriwise]

I would be ok with that. I'd like it working as robustly as possible.

[pshriwise]

Issue handled. There are two ways in which check_watertight works. One in which is uses the proximity facet vertices to ensure that they are within some tolerance of each other (the tolerance being specified by the user after the -f flag in the command), and the other is a topological check, making sure that each vertex handle on one surface skin has a matching vertex handle on the corresponding skin of an adjacent surface.

[makeclean]

Ok, is this reflected in the documention?

[pshriwise]

Not yet. I'll add it into the doxygen documentation for check_mesh_for_watertightness.