<analoq>
i'm looking for a way to create an o-ring groove, but can't really find anything helpful.
<analoq>
my thought process was using extruding a cross section along a spline
<peepsalot>
analoq: is the groove not circular?
<InPhase>
analoq: If circular, rotate_extrude. If non-circular in an unusually curvy shape, I recommend this approach: https://github.com/rcolyer/closepoints
<analoq>
peepsalot: not too familiar with oring groove design, reading up on that currently. My initial idea was to do a c shaped circular groove
<peepsalot>
what i meant was whether the overall groove path was circular
<InPhase>
peepsalot: Interesting link.
<peepsalot>
it was like the first result for o-ring squeeze (couldnt' remember if that was the exact terminology)
<analoq>
peepsalot: yeah i figured that out while looking for a solution too :). No the path won't be circular
<InPhase>
analoq: A demo that follows those guidelines from the link. https://bpa.st/F46Q
mmu_man has quit [Ping timeout: 256 seconds]
<analoq>
that's a nice looking profile
<InPhase>
Is your actual groove composed of circular and straight stretches, or an artistic curvy type path?
califax_ has joined #openscad
califax has quit [Ping timeout: 260 seconds]
califax_ is now known as califax
<analoq>
InPhase: the former
califax has quit [Remote host closed the connection]
<guso78>
Anybody Has an actual Idea for algorithm about my 2d bezier curve
califax has joined #openscad
<InPhase>
guso78: That's bezier? Looks like a stretched circle.
mmu_man has joined #openscad
<InPhase>
So let me start by pointing out it's trivial as a hull between scale deformed spheres at the corners. :)
califax has quit [Remote host closed the connection]
califax has joined #openscad
<guso78>
ITS actual a simple bezier with 3 Control Points
<InPhase>
Can you show the control points? I have like 70% of a idea.
califax has quit [Remote host closed the connection]
<InPhase>
I think you should be able to make a bezier corner coordinate map with beziers of beziers.
califax has joined #openscad
<guso78>
Take 3 of 4 Points of a square aß Control Points and linear Extrude it to Form the Edges
J24k68 has quit [Quit: Client closed]
J24k68 has joined #openscad
<guso78>
Issue with Corner IS that the bezier should bei symmmetrical around the 3 axes, so No hierarchy possible
teepee_ has joined #openscad
teepee has quit [Ping timeout: 260 seconds]
teepee_ is now known as teepee
SamantazFox has quit [Remote host closed the connection]
<InPhase>
True symmetry is quite challenging. Even spheres are hard to make truly 3D symmetric.
J24k68 has quit [Quit: Client closed]
J24k68 has joined #openscad
<InPhase>
guso78: peepsalot made some nice 3-axis oxtahedral symmetric spheres once by interpolating and pushing the points out to the radius.
<InPhase>
guso78: But you lack the starting symmetries that let this work.
SamantazFox has joined #openscad
<guso78>
Sounds Like construction of geodesoc spheres ...??
<peepsalot>
yeah
<guso78>
In Phase ITS.not necessarily a sphere or Radius.its Just needs to bei round and balanced
<InPhase>
Heading into a meeting for a little bit. I'll ponder it more after.
<peepsalot>
you could do something like a sweep between each three pairs of edges, and intersect between them all. this will likely leave a Y-shaped crease
<guso78>
Peeps I cant afford a Gap for Manifold reasons but your proposal Sounds promising. Would you have time to create a Sketch ?
<peepsalot>
maybe later this evening. would help to have your source to start from
califax has quit [Remote host closed the connection]
califax has joined #openscad
mmu_man has quit [Ping timeout: 268 seconds]
mmu_man has joined #openscad
fred_ has joined #openscad
<fred_>
hello
<J24k68>
hi!
<othx>
Welcome J24k68!
<fred_>
anybody around here?
<InPhase>
fred_: Yes, we exist. :)
<fred_>
aah good to know :-)
* J24k68
muffles in the dark
<fred_>
not much action here..
<J24k68>
depends on the topic
<fred_>
yeah I thought just to listen first to see whats happening here
<InPhase>
Reasonable choice, but think more on the 24-hour to 48-hour lurking scale for that. :)
<InPhase>
Conversations will have time gaps between them on a channel of this size scale.
<InPhase>
guso78: So I feel like it might be doable special-casing for corners bezier modeled out of rectangles.
<InPhase>
guso78: I'm thinking about how one would arrange that at the moment.
<InPhase>
guso78: Triangulation is left as an exercise for the reader? :)
rawgreaze has joined #openscad
<InPhase>
guso78: There's a slight axial bias left, but it's very hard to not have one when you're left with the problem of needing to quantize points into a 2D map...
<InPhase>
guso78: But it's not very spatially strong. Certainly less biased than with a sphere.
<InPhase>
guso78: It at least has the highly desireable property of the points aligning exactly with the bezier results of all 3 axes.
<InPhase>
guso78: Start the triangles for the facet generation leaning in the forward direction as you iterate through the lists, to address the declining number of points as the rows rise in z.
kintel has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<InPhase>
And then you might have to special case the top. The top can go from 2 to 1, or 3 to 1, but there's only ever one row, the top one, that has 1 point on it.
<InPhase>
N is always forced upward to the nearest odd number if even.
fred_ has quit [Ping timeout: 255 seconds]
<JordanBrown1>
guso78, InPhase: I would think in terms of BOSL2's Bézier patches.
<JordanBrown1>
Though for some reason they dropped support for triangular patches. I don't know why. Maybe they just expect you to use rectangular patches with one side degenerate.
<JordanBrown1>
Back to pretending to work.
<InPhase>
I heavily preference small libraries (making and using) for OpenSCAD work, but am fond of looking at examples in other libraries if they're clear in isolation.