OffsetSrf fails to create solid

Rhino
,
1

Shape was inspired by shape in How do I fix overlapping surfaces but is not identical.
Shape with planar surfaces. Polysurface was created by extruding a curve to a point.

OffsetSrf fails to create a surface.
Portions of the result are not symmetric where the input is symmetric.
Offset was 10mm inward.

Version 8 SR14 (8.14.24311.6001, 2024-11-06)
OffsetSrfFail.3dm (2.5 MB)

image
image1372×1306 64.8 KB

Interior surfaces are not missing. If the open are was simply capped a non-manifold surface would result.

image
image1920×868 61.9 KB

Input is symmetric but the result is not symmetric.

image
image2406×973 335 KB

1 Like
2

OffsetSrf outward results in a more obvious mess.
OffsetSrfFail-2.3dm (2.7 MB)

image
image1491×1268 125 KB

1 Like
3

this is a self intersection master class when offsetting…

but, you can work around it, duplicate and offset the curve at the base of the object inward 10 units, then extrude curve to point (set the point -10 units below the exterior top tip of the model) and then make a planar surface between the inner and outer object and join.

image
image927×667 173 KB

1 Like
4

@theoutside I reported this as an example of how OffsetSrf needs to be fixed/improved. This example with simple planar surface should offset correctly. Below is a simpler version which also fails. This should be put on the heap of work needed on OffsetSrf.

Here is a very simple example:
OffsetSrfFail Simple.3dm (2.1 MB)

image
image914×1231 125 KB

1 Like
5

we agree there is work to do on offset, I was simply offering a different way to approach this type of model solution that typically works reliably.

1 Like
6

It should go on the heap of YouTrack items.

It is fundamentally flawed advice if the intent is to create a model with a 10 unit wall thickness. It will result in a model with a thinner, variable thickness wall.

The base curve is planar so offsettng it 10 units will result in it being less than 10 units from the outer surface. Moving the point at the apex down 10 units will result in being less than 10 units from the outer surface. Given the angle at the apex the point will be even closer to the outer surface surface than the base curve. The combination of the offset curve and the moved point used to construct the inner surface will result in a surface with variable offset less than 10 mm from the outer surface.

1 Like
7

That’s why we should add dimensional and geometrical parametric constraints to Rhino so we don’t have to calculate all that trigonometry. :face_holding_back_tears: :melting_face:

Would be cool to see this setup in GH :face_holding_back_tears:

8

Or just have OffsetSrf provide good results.

1 Like
9

True, but offsetsrf is just one way of doing something that can be done more than one way. And not necessarily the best way.

But it would be nice if it were automatic. It almost seems like maybe filleting is similar or has similar problems. I’d have to be a nurbs mathematician to know for sure though.

It’s like Rhino needs to learn how to automatically trim, extend, add, subtract edges and faces etc. :face_holding_back_tears:

10

Hi, I posted the original model in How do I fix overlapping surfaces which inspired the shape in this post. I agree the OffsetSurface tool could use some work, but I am also new to Rhino so I didnt know if there was an obvious solution to the problem. Would anyone have any more suggestions, I’m still struggling with it and I need the wall thickness to be at least 1mm for 3D printing. Would really appreciate any input. Thanks!

1 Like
11

if you are 3d printing check this out-

and you’ll want more than 1mm for printing unless you are using a very sla fancy printer.
2.5mm is closer to what you’ll need for a decent print, and that is even thin.

I’d actually aim for more like 3.5mm if you are making a pla part. Sla you can go thinner, but it’ll be very fragile.