if anyone gives it a try to build a similar thing in grasshopper.
go for it, would love to see it. First version of grasshopper was called “explicit history”. happy spaghetti.
Onshape
@Rhino_Bulgaria
here you go - just a fast version - still room to improve, but it shows the power: onshape-link
i have to say i am not a onshape expert - i mainly use it to show the power of it in class. .. And i was lazy with the complex transition, just used a single Fill Surface.
there is 3d tangent / curvature constraints and they update.
rhinos _matchSrf with history is far from that…
looking forward for this functionality in rhino.
And yes trimmed patch like xnurbs, new patch in rhino Wip, Boundary or Fill in Solidworks / Onshape / Plasticity…
i would claim, their is no need for high end single-span class-A modelling.
Casting a signet ring … there is a lot of sanding and polishing after casting:
What about an even simpler and far less curved NURBS geometry (good for two-piece stamping), such like a car fender that’s not sanded like the rings? An epic fail for every parametric technique.
I worked quite a bit with casting small products, both plastics and metal, and what @Rhino_Bulgaria says is true; like with 3D printed parts, you spend quite some time with various surface finishing processes, so with casting, you have leeway. Just see how jewellery is made, or zinc alloy/magnesium alloy cast toy cars, or even larger expensive items like cutlery or table culture products, or expensive hansgrohe faucets cast from brass and then electroplated, see below (see the raw brass cast on the left, that’s really how things look).
That does not give you an accurate definition of the hole’s profile, hence it looks a bit wavy (a slight S-shaped profile). The same weakness is seen on NURBS models made in Rhino using the same approach with trimmed hole.
I’ve previous experience with Creo and it’s a beast: hard to learn but unbeatable on modeling whatever shape you want. I switched to Onshape some years ago for multiple reasons and now it’s almost capable but much simpler.
I don’t think to get time to apply for this challenge soon but I’ll check @Tom_P link and maybe fork it with a more precise version.
@Rhino_Bulgaria I see your point but not all the applications need Class A modeling. As other said most of the time the process would hide small modelling mistakes. Even cars are “hand polished” after the stamping because the process leave some marks.
I’m aware of the machine polishing (rarely done by hand nowadays) on car stamping dies either from steel or ceramic, but we speak about an extremely minor polishing that still requires an excellent NURBS surface quality with a maximum deviation of about 0,01 mm and 0,05 degrees.
Obviously, the potential constraints technology in Rhino will be mostly suited for simple mechanical parts rather than free-form shapes.
I’ve been building my own tools since years, back in time matrix would cost about 7 or 9 k, I don’t remember, way too much for a goldsmith apprentice. Anyway, the signet ring was one of the first things I scripted, so please don’t be too hard on me for bad coding. Indeed the side curve is a tricky thing, mostly because there are many different signet rings and forms.
Asymmetric forms like shields tend to be very tricky, but it is doable with a hole. Even if it’s not implemented in this version, the key to real universal use would be to adjust this curve accordingly:
At the moment I don’t have enough requests for signet rings, so I don’t put resources in developing it further.
So, here you go with my approach…I admit I have the curves as output as well, just because a don’t want too many parameters and depending on the curve and the ring size, it might look a bit odd at times. That’s why I also have the curves as output in case I need to adjust it in rhino. I find it really hard to find the balance between usability and adjust everything with a parameter.
It’s done with networkSrf.
Based on my experience with a very large auto maker prior to 2009, automotive show surfaces go through multiple iterations. First there is the studio process which eventually results in the studio releasing the surface. Then another group takes the surface as released by the studio and reworks it to add/adjust/refine the split lines, seams, flanges and other details needed for manufacturing, and to ensure the surface meets the “Class A” requirements and tolerances. The surface math data is released to manufacturing for making dies, molds, etc. The “Class A” requirements are ultimately about the quality of the surface, not the details of patch arrangements, degrees and similar.
The Class-A requirements are primarily due to the need to provide a clean geometry, both, for aesthetics and negative offsetting with a minimum amount of errors.
There is a major drawback with patch surfaces used for this particular ring shape. The body must look the same thickness from most angles, despite that the upper portion is wider. It’s a pure visual aspect of the geometry. In your example, the upper portion appears thinner due to the lack of a direct control over the edge of the hole and the shape of the upper surfaces. Perhaps this negative could be mitigated a bit by adding an extra diagonal curve profile where my upper NURBS surfaces meet together. Or a horizontal curve profile at the 70% height of the ring.
I wonder how Rhino 9 WIP’s new “Patch” tool will handle such cases.
@Rhino_Bulgaria: I have masses of respect for what you do with cars. But are you making up these criteria for jewelry?
A ring shank having all along the same thickness looks like an old German tank, we don’t want to wear that, it’s not pretty. I am sure that there is no need at all for Class-A surfaces in jewelry. That’s an industrial product designer and car designer thing. goldsmiths in general do by nature of the casting production what car designers do (or did) with clay, but for every single piece.
It’s all about having consistency. The sample ring has a noticeably thicker upper half when you look at it from front view (towards the hole). However, as the camera angle changes, the upper portion gets significantly thinner than the lower half at about 62-65% of the total height of the ring, making the latter “heavier” by appearance. The upper half must retain its thicker nature as much as possible to look visually more pleasing.
Also, the upper and lower halves of the ring are connected by 0,1 degree G1 tangency instead of smoother G2 curvature. This is another aspect of product design (especially jewelry) that make the shapes more beautiful.
The issue is caused by the irregular angle between the parallel hole and the curved body of the ring. Notice that the edge between both surfaces is sharper is at approximately 62-65% of the height, where the tangent curves are added to measure the angle.
An epic fail for every parametric technique.
