Sweep2 result can't reach G2 continuity

I’m parcticing 3D modeling with Form vs. Shape 2 and stuck with an issue.
I’m following the steps at page15 and building the transition surface. But the surface which is generated with Sweep2 coulde reach G2( Zebra shows that’s G0) to the surface at the right side. I have tried Matchsrf but there is no tangible progess.

Did I miss something? Thank you!

螢幕擷取畫面 2022-11-23 170628

Hello- please export all of the inputs to the sweep to a new file and post that here.


1 Like

Sorry for my insufficient information. Here’s the screenshot of the sweep2 setting and the 3dm file.

form_vs_shape_2.3dm (336.4 KB)

Ah, OK - so, the cross sections are arcs and matching cannot be done with arcs or other ‘rational’ curves as cross sesctions


You’d need to use the Rebuild setting and use at least 6 control points to get curvature on both rails.
But, I’d do it like this I think, as three surfaces:


1 Like

Just some general tips and tricks from my side:

You cannot match G2 properly if your surrounding shapes do not support that. Best situation for matching is if,
a.) all surrounding surfaces are already matched to G2 or higher.
b.) all surrounding surfaces are (ideally) single-span, non-rational and follow a clear & simple iso-direction=“flow”, ideally rectangular shaped or fan-shaped, with as few points as possible
c.) Your surface to match is single-span at least 6x6
d.) for any issues in a.) b.) you need to subdivide the blending. However, the more you subdivide, the more you decrease the visual quality of your blend, no matter if G2 or not. So further subdivision is always a trade-off.
e.) If G2 is hard to achieve, rather go for G1 but with as few points as possible. This usually looks nicer, than a technical G2 with a bad overall flow. For 4 sided G1 you need at least 4x4.


I didn’t know this restriction! Thank you so much!
But the three surfaces seem more difficult to match G2 to the right blending surface, don’t they?
螢幕擷取畫面 2022-11-24 165334

Thank you!
The obstacle for me now maybe is the right surface I use Blendsrf to connect two surfaces. If I made it single-spawn with Rebuild, it’s gonna be tweaked and deviate from the two surfaces.

I found that the mismatch between two surfaces was because that they weren’t joined.
After I joined them, there was no more G0. Only G1 and G2 coexsisted.
But I learned something new to me from @pascal and @TomTom because of this carelessness, thank you!

Hi Kuo - Joining will align the mesh vertices of the analysis mesh alng the join so yes, if the mesh is accurate/fine enough to begin with, it may remove a ‘false’ misalignment of the stripes. Use EdgeContinuity as well, this will not be de;pendent on the analysis mesh.


1 Like