Hi Guys,
It turns out that this IS a precision problem but not perhaps in the way you might expect.
Explanation:
When Blue Kenue searches for common nodes between the two meshes it uses a proximity tolerance value equal to 1% of the minimum edge length. In this case it turns out to be 0.015662
In Christophes example above the distance between the two nodes is 0.031145 hence we're missing this common node.
I'm guessing that you have saved the submesh to a selafin file at some point and then extracted it again. Whereas the GreaterMesh was not. The Selafin file is single precision so
the node coordinate Y value 4093807.469 is rounded to 4093807.5
And viola, the error.
I suppose the solution is for NRC to implement support for double precision selafin files. In the meantime a work-around is to keep all data in native EnSim t3s/t3m file formats until the end.
Cheers... Martin