This is an issue I have raised before but have never really gotten a good solution. I run into this all the time. That is, places where I need a radius to go to zero.

In the attached have a three-sided opening. The left corner forms an angle. Mathematically, I should be able to have a surface that is tangent at all three edges. That requires going to a zero radius at the left corner. I get all kinds of bad surfaces or surfaces that have odd shapes at the left corner.

I get the same problem even if I cut the opening to make it four-sided.

Hello- I guess you are shooting for something like this?

But, more than likely there is a better way to build some of this - partiicularly the complex trainagular thing. Hard to say without seeing the surrounding faces.
-Pascal

Yes, something like that to fill the corner. This is a problem I keep coming back to. When you get to the end, the triangles appear. You can cut the opening to get a four-sided opening but the problems with bad surfaces appeal just the same.

Right - I would see if I couldn’t make a rectangle and trim it across the diagonal with the surface there that makes the edge of your triangle

-Pascal

Hi @miano,

You have some interesting curvature around the area, and seeming discrepancies between adjacent panels. I wonder if that is contributing to the difficulty? Is changing the existing surfaces permitted?

Regards
Jeremy

Yeah, I think this all should probably be built differently, using more ‘standard’ primary and transition surfaces.

-Pascal

The problem, as always in nautical stuff, is getting to the corners at the end The large surface is simply a networksrf using lines and arcs as frames. Using sweep2 creates accordion pleats. That surface cannot go any farther aft because that would create a 5-sided surface. The problem is how to finish the final bit of faring.

Interestingly here, the plans do not who had it was done with measurements. The large surface points are taken directly from measurements. There are no instructions for fairing the corner in the blueprints.

One possibility I have considered is to cut that surface to the aft curves but I would sill have to figure out how to bring it to zero radius.