Radius to Zero at a Corner

Here is a problem I encounter frequently but have not found a good answer to. I would like to fill this hole while maintaining tangency with the upper surface. The radius needs to drop to zero at the end. This a mathematical possibility but does not work well in my Rhino attempts.

Let’s say I do NetworkSrf with edge tangency, the right edge comes out tangent but the top edge does not.

If I do Matchsrf I can get the edge tangent but that create a bulb at the end.

I was wondering if anyone has come up with a good solution to this problem. It occurs at the end that has to match up to other surfaces so there is no flexibility in the final red and blue edges.

Problem Corner.3dm.zip (122.1 KB)

I’d use sweep 2 rails command to make it work, the bigger challenge is you may get some stretch marks on your Emap.

You can reduce this warping effect by applying Smooth command with Fix boundaries checked,



Hello - I’d build it like so:



This is a really clean output. Thank you Pascal

The only trick is how to find the dividing curve such that it ends tangent to the left curve.

You can do something like this…basically, project curves from the top, extend them and fillet using pascal’s filletcrv tool so that you can snap to the points. Easier to explain in video, maybe:

Thanks for the video. The FIlletCRV is new to me so I was kind of wondering what was being done.

The best I have been able to do on my own is this:

I filled in the entire area with one surface.
Trimmed to cut a 4-sided hole (both pieces shown).
Sweep-2 to fill the hole (which at this point sill not match at the left edge).
Trimmed the left half of the fill away.
NetworkSRF the rectangular opening at the left.

I get many need-tangent-to-the-corner problems. Needs tangency in three sides (not the upper right).

Here’s one that I got to look OK.

The problem is it is a deadly triangle (NetworkSrf). Sometimes it mysteriously becomes a bad surface.

Pascal has the right idea, I think.