Surface Curvature Direction

Hi all,

I am working on a series of single curved panels. They are derived from filleting two sides of a tapered truncated pyramid. As you can see in the screenshot the fillet radius (V Isocurve) is 350mm .

I have to identify the curvature direction and I sketched two methods, gh file attached below.(Method 2 is actually a piece of code I took from an old post of David Rutten, I (well…Claude did) just translated to C#.

In my naivety I would have expected the radius of the identified curve to be 350mm, as per fillet radius, but it’s not the case as the reference surface is a tapered pyramid.

The main problem here is that I would have expected all the panel surfaces to show at least the same curve radius, as they all come from the same original surface. I understand that both methods are approximations of curvature, but some of the discrepancies are in the range of several millimeters. It doesn’t sound like much, but it was enough to give me the itch that something might be off.

Any idea or suggestion would be much appreciated.

P.S.: all the panels are created splitting and trimming the original surface, so the curvature should be consistent.

Thank you!

Paolo

Curvature Check.3dm (584.0 KB)

2026002_Panel Curvature Direction Check.gh (25.0 KB)

Seems fine to me…


2026002_Panel Curvature Direction Check VR 1.gh (44.6 KB)

Underlying untrimmed surface reports 350 at all points across the arc for all surfaces. You are using a Circle Fit to get radius. That’s not going to end well. Curvature gives you a circle anyhow.


Oscillating Circles will show you the radius in the main surface direction:


2026002_Panel Curvature Direction Check VR 2.gh (57.0 KB)

which is consistent between the surfaces, but not consistent around the fillet, because, well, your surface is essentially an Extrude Along so no surprise there.


If you want uniform curvature in the direction you will have to formulate the surface differently. Right now, it is like a cylinder with a slanted axis. Horizontal sections will have arc. Sections made normal to the suface will be elliptic.

3 Likes

Some surgery to create consistent radius:


Surface Curvature Direction VR 1.gh (22.7 KB)

Maybe could have been cleaner with the base surface creation which I then fillet.

3 Likes

Thanks so much Volker.

The mistake was indeed in the construction of the base surface!

1 Like