(i didn’t really investigate what the exact problem is with your model… just modeled the nodes fresh (well, extracted the seam from one of yours then revolved it, aligned it with the larger object, then polar arrayed it)
I suspect it is to do with the isocurve lining up with your blue object (rotating the central element a bit solved the issue). Merge all faces still a problem though.
I am puzzled why yours worked in mine does not. I have been playing with all kinds of variations here and am unable to create one the does a union of some sorts.
Hi Jeff - yours is slightly different in that the center of the little revolves fall on the edge of the big circle - I noticed that Jim’s do not, quite. My guess of the moment is that the original placement makes the singularity of the spherical ends intersect the large revolve right on a knot-line, and in yours it misses. At any rate, bug noted for V6:
I did a test in moving the blue object just a minute distance (0.0001’) away from the center of the big object, not all the way to the edge, and the booleanUnion succeeded.