It take almost 10 seconds to generate and a few more to display on my system.
The result looks a lot different (but no better) when the Sweep curve moves either way in the X direction. But everything looks OK if I substitute a simple circle for the Sweep curve.
I tried Rebuilding and Reparameterizing the Rail curve - but these changes don’t help. So I thought maybe a rail curve can’t have kinks in it. But smoothing them out didn’t help either. Could there be some other tweak that will produce the expected result? SweepDemo.gh (16.4 KB)
Only with what I call crenelated sides instead of a flat/straight one.
I’ve seen things like you showed, but don’t understand how that can be the result from a Sweep1 with a horizontal Rail and vertical Sweep curve.
Michaelkreft: that’s close to what I’m trying to make, so I’ll try Sweep2 with a smaller diameter version of the Sweep1 curve on the top.
This all seems quite strange and unexpected - I’ve done lots of shapes with Sweep1 using different Interpolated curves as the Sweep and never had a problem like this.
I’m getting some very strange results when I use Sweep2 with the Sweep1 (bottom) curve moved vertically as the Sweep2 curve. It seems like there is some very strange effect caused by the corners in the rail curves. What I’m trying to do is find a simpler way to create something like this:
I try not to use meshes because they seem to behave in unexpected ways - which I realize is a brain problem and not a mesh problem. I’ve got a method that works, but it’s somewhat complex and it just seemed to me the Sweep1 approach would be simpler if I could get it to work.
Based on the odd things I’ve seen just now it looks like the sweep routine is taking each straight line section of the rail curve and doing something like an extrusion with it.
But when I take out the corners by filleting the Sweep1 simply fails. Sweep2, on the other hand, does this:
Exactly correct - the photo I showed above is what I was trying to recreate with a simpler approach. When I look at the surface edges of this method I can see the desired curves in there, but there’s also loads of extraneous stuff that would be difficult to eliminate. And doing that would not achieve my objective of finding a simpler solution for an approach I already have.
while I think I understood (maybe ) the geometrical behavior you want to achieve for the green faces, I didn’t get how you want the blue ones to behave
Thanks inno - it never would have occurred to me to locate sweep curves the way you did. I tweaked the curve a bit and added some stuff to make an inside surface. It ends up as a nice Closed Brep:
Yes, RailRevolution works just like I thought Sweep1 would work. (I’ll leave it to the McNeel gurus to figure out why it doesn’t) But clearly it does not. By using RailRev the whole process becomes much simpler. It also offers the benefit of not requiring a circular cross section, which means I’ll be able to use Rail shapes like this:
I don’t think this is up to the McNeel gurus to figure out. Sweep is working exactly as intended. You may need to review the basics of how the freeform surface generation tools work.
This is, for instance, how you would create the surface with Sweep2; by either moving the profile to the gear or scaling the gear to the profile:
) the geometrical behavior you want to achieve for the green faces, I didn’t get how you want the blue ones to behave
