ZEBRA question

Vid shows a couple of corner surfaces that union three different radius curves. What seems to be a simple task doesn’t turn out as clean as I’d expect. There’s some fundamental problem with the surfaces but I don’t know what or how to remedy it. Pascal thinks I should work from the simple to the complex but trying to union three ‘simple’ surface sweeps has not produced a clean corner like the one shown.

Test.3dm (162.4 KB)
http://www.screencast.com/t/wSNZKBPKwhLt

For what it’s worth, I would suggest you start fresh, and match curves first to each other, still keeping the amount of points to minimum. Since you chose curvature continuity (which is an overkill in case of simple boxes, IMHO, and is suited more for automotive design), the side surfaces can be merged, which gives you less edges to worry about. Thus you can easily blend the corner all at once. After blending you can run RemoveMultiKnot to discard some extra points.

Thanks. I’ve tried different curvatures (higher levels) only to try to get matches. I’ve not used the multknot tool - I’ll learn about it today. I’ve a habit of throwing everything away and starting over as I benefit learning wise. I want to get this area of modeling down before proceeding so I don’t mind the repetition. That said, I’ve been through many many permutations of the advice I’ve been given. I appreciate it all.

One thing that does mystify me is the apparent lack of continuity that results after simply pulling an edge away from an apparently cleanly formed surface as shown in the vid. (I cmd clicked the surface edges and then command dragged the edges) and still on analysis is shows a discontinuity.

If you look at the vid around 20 seconds you’ll see the strange pull wherein a surface is created inside out. I think this is the problem but I don’t know what/how to change the geometry to match up.