Can anyone help me with a patch layout? The sample file has two different sets of Reference geometry: one concave (reference images) and one convex (reference images) profile:
Part of the problem may be the reference geometry (Reference layer). If you think better input geometry would make the problem easier to solve, go for it.
Whatever the solution is, I need to make it into a History-enabled template file like this. We need to quickly scale to the right finger size and ring height (Base layer). The ringâs head shape and shankâs profile, width, thickness, taper, etc., can be edited separately (RefParents layer). Then there are a couple of editable blend crvs to connect the shank to the head.
Solution #1 (paid): Xnurbs could help with filling that kind of 5-sided holes.
Solution #2: Use two 4-sided surfaces and match their border. Similar to the approach seen in this video:
Solution #3: Extend the vertical profiles, meet them to the same point on the top, then create one 4-sided surface (âSweep 2 railsâ, or âEdge surfaceâ, or âNetwork surfaceâ), rebuild the surface with degree and control points to your liking, edit the control points to achieve the desired shape, then go to side view and split its top portion to open a hole for the main diamond. Like this:
With regards to your V7 file, my Rhino 7 says that it was saved as Rhino 8 file format, so I still canât open it there.
As for the geometry inside your model, I figured out that Rhino produces some weird bug with the ! _EdgeSrf command. In most cases with 3 input edges it fails to build a proper surface depending on the order of selecting the input surface edges. To make it work, you have to pick the middle edge first, then the remaining side edges. Hopefully the developers will fix this with a future update.
Cool, I didnât knew about that. I may give it a try, even though it would be best if the file uploaded above was in V7 format, so that other people with older Rhino could also help.
Another bug (maybe?) is that Rhino 7âs âPatchâ tool produces much better quality patch compared to Rhino 8. At least in this case.
Iâm playing around with this one. It has the added benefit of being a single (although not ideal) target for FlowSrf. The base shape is just step one of many. All of the stones and cutouts have to be added as well.
This is a common trick in NURBS modeling where a simple 4-sided surface is split in a clever way to achieve a complex-looking shape with 5 or more edges. The only downside of this approach is that the split area will have some minor deviation and will not match your top reference surface with 100% precision. However, I donât think that itâs a big issue for a small object such like a ring.
If your goal is to be able to orient plenty of stones and other elements all over the surface, I recommend you to use a single closed surface built with the âTorusâ tool.
Now you can edit the control points to achieve the desired shape. Add more control points (either via ! _InsertControlPoint or ! _InsertKnot with Midpoints=Yes or Symmetrical=Yes option) if need more local control over the shape. This is by far the most perfect way to achieve smooth continuity across the entire surface.
I wonder if the Rhino development team intends to fix the bug with âEdgeSrfâ, which I showed in the video in post #4 above? That tool works properly on 3-sided cases only if the user picks the middle surface edge first. It fails when the order of picking the edges is different.
None of these give me the control Iâm after, which is my fault. I think Iâve been approaching this incorrectly, and itâs hard to convey this ringâs specific âfeel.â Iâm spending the afternoon trying to match the original ringâs patch/sheet metal layout.
This style was stamped from a flat sheet using an âO-Hubâ pattern.
The green section was sheet metal cut to fit the center stone size, and the blue wedge was cut to fit the gap that resulted from a fixed-sized O-hub mating with different-size/shape top plates.
Iâm missing the control that leads to the feel of that seam with the blue spine. After soldering, the initial sharp corners are filed into round. Itâs very distinctive. I can tell if something is early 1900s just by running my thumb over the spine.
Itâs hard to edit the shape and match the target center plate along its entire diameter using the sheet metal layout with radial isocurves. You can get it near the quads or at the midpoint between the quads, but not both at the same time:
Itâs 5 editable curves (2 of the 7 inputs are projections for tangency), but I think I can get it down to 3 editable plus 2 projections. Iâll have to try that another day.
I may give it a try, even though it would be best if the file uploaded above was in V7 format, so that other people with older Rhino could also help.
