Why cannot fillet command produce G2 continuity?

(Alex) #1

In the following file you see G1 continuity in filleted areas. Zebra evaluation is also showing G1 especially in corners.
In Alias we can select G1 or G2 continuity in fillet tool. Why does Rhino 6 lack this feature?
Untitled.3dm (319.0 KB)

(Pascal Golay) #2

Hi Alex - it is better than that, but on a flat sided box the analysis mesh is so coarse that it is not that clear in Zebra - one thing to keep in mind when using the false color display analysis modes is that you need a very tight mesh to get a good read - one way to do this, for analysis meshes is to just load up the imitial grid quads with a large number - that is fast and evenly distributed.


(Alex) #3

Thanks for the guidance.
What advantages and disadvantages does Rhino have over Alias Autostudio?
In Alias we can form freely. For example we can turn a simple sphere into a complex shapes like a vase just by playing with CVs while we are in Nurbs mode not polygon or Tspline mode. Can this be done in Rhino 6?

(David Cockey) #4

Previous recent threads with questions and responses about Rhino compared to Alias:


Out of curiosity could you upload a step file of an Alias version of this cube.

(David Cockey) #6

Perhaps you could experiment in Rhino. Create a sphere and try dragging control points.Add control points if needed using InsertKnot or InsertControlPoint. You may need to rotate the sphere so that orientation of the isocurves is as desired.

(Pascal Golay) #7

You will need to Rebuild the sphere if you want to monkey freely with the control points.


(Alex) #8

I provided you what you wanted:
Can you make such a cube (the same quality) in Rhino 6? I’m curious to see.

test.stp (1004.6 KB)

(Alex) #9

Please check out my files in response to Stratosfear…
Can you create such a cube with the same quality in Rhino 6? (G2)


Thank you. That’s pretty impressive. It is annoying having to go in and match surfaces to clean up continuity issues, with Rhino Fillets. Having VSR shape makes it a lot easier.

(Pascal Golay) #11

I know this does not help anyone now, but for what it is worth, the WIP results:


(Alex) #12

WIP is about Rhino 7?


If it’s not in R6 it’s not good enough. (but there’s a lot of free space behind the dot in R6.< __ > )

// Rolf


Could you please post a step file of the WIP fillet?


I was expecting to see the Alias fillets to all be Class A. The corner rounds are not. If the job requires it I’ll go back and manually clean up my fillet/blends where needed.

(Alex) #16

What software does your image belong to?
Is it Rhino?
The corner consists of two G2 and 2 G1 continuity. Those two edges that are in horizantal direction don’t need to be G2. But you can make them G2 too.


It’s all Rhino with the VSR shape plug in for surface matching analysis.

(Tom) #18

you should also crown the cube, in order to better compare results. You will hardly see any flat surface in real world filleting scenarios, where g2 fillets make sense. VSR was great, I would even claim deviation analysis was the most critical feature. Very often I still match manually in situations where any algorithm fails, no matter if Alias, Surf or Rhino.


I guess I don’t understand why fillet conditions this easy, produce less than perfect results. It’s not just Rhino, I’ve tried a bunch of Cad packages.



// Rolf