Sphere on Torus

[[To...]]

I need to measure the diameter of an intersection of a sphere on the surface of a torus. Not sure what I mean? Imagine a dimple on a golf ball but the ball is really a portion of a torus. Any ideas?

[[Er...]]

I might have done something like this a long time ago. But I had a torus cutout on the spehere.

What I ended up doing was measuring curves that "fell" over the intersection edge. Placed a CS in the center of the torus thing.
Collect the highest points of the curve. And recall to an ellipse. (Maybe I did some projection? Cant remember.) That was the most repetable of everything I tried.
(Thin lines are curves)


But you might be trying to explain something else?

[[Mi...]]

The shape you're describing doesn't have a diameter, does it? Isn't it saddle-shaped, like the edge of a pringle?

Anyway, the only thing I can think of to do is:
- Measure the sphere (as a sphere)
- Measure the torus (as a torus)
- Measure a number of arcs across the region of interest. For the ones in the dimple, constrain them to the radius or position of the sphere (I'm not sure which would work best). Do the same for the torus immediately around the dimple.
- Create intersections from the arcs, recall into a circle, and report the diameter.

This kinda fudges it, though, since the points won't actually be coplanar, unless I'm picturing it wrong.

[[SH...]]

Please sign in to view this quote.

Something like a ball of bearing on cycle tyre tube?? I have not worked with both sphere and torus simultaneously. Direct intersection is not possible??

[[To...]]

Please sign in to view this quote.

Technically, you are correct but the torus is so close to being a sphere, the customer has specified a diameter and are currently checking it as a diameter on a MicroVu. Thank you for the ideas. I'll give them a try.

[[To...]]

Please sign in to view this quote.

No, intersections don't work with a torus. I have a phrase that I use every time someone mentions they are working a torus and it goes like this. "A torus doesn't play well with others."

[[Ri...]]

Curve is the way.

[[To...]]

Please sign in to view this quote.

Can you elaborate?

[[Ri...]]

Show me an example, and I'll show it to you. Lol. Shoot me an email if you'd like, and I'll take a crack at it.

[[Cl...]]

I'd create several curves through the sphere, scan those and recall them
into a sphere feature, and because its such a small segment you'd have
to constrain location, then intersect with torus and shell for diameter.

[[Cl...]]

I just thought, if the surface at which the sphere intersects the tours not flat, my idea won't work?

[[Cl...]]

Oops, that intersection is not allowed, just tried it.

[[Ch...]]

If the customer is considering the form as a sphere (locally) why not consider it a sphere yourself?

[[To...]]

Please sign in to view this quote.

dimple on a golf ball doesn't work? How about this?

8vxGgBvXrwEDFysRL9CyEP-1200-80.jpg

[[Cl...]]

Six to eight 2d curves slicing through the sphere (and into the torus) like a pie. Get the intersections using kink points.

[[Ri...]]

Two options:

- Vision

- Measure curves that are 90 apart from one another (like in the attached picture), and create the intersections.

There could be a third where if you can actually measure it as a sphere, and just use a theoretical plane (it would need to be tangent to the torus) to create the intersection.

[[To...]]

Intersections of the 2 curves? Isn't that at the bottom of the dimple? How do I get a diameter of the sphere at the surface of the torus?

Clarke, I've never been able to to get kink point to work. How do I populate all of the features?
.

Screenshot 2021-01-14 144140.jpg

[[Cl...]]

Please sign in to view this quote.

2d curve, recall the points into a line and a circle. Pretty sure the construction should be curve - line- circle.
You will have to use two circles, not sure if that will work. If the torus diameter is big enough, you could use
a short line. I've only used the line an circle method.

[[To...]]

Please sign in to view this quote.

Because that is too f'n simple.....LOL At least I'm able to get a diameter at the intersection of 2 spheres.

[[Ri...]]

You could also try step-point and just measure it as a line. Cuts curve out of the equation. Kink point is iffy sometimes.

[[Ri...]]

Please sign in to view this quote.

You can use Curve - Circle - Circle. In this case it should work OK.

[[Mi...]]

Question: What's the advantage (in general, not this part specifically) of using a kink point vs. an intersection construction?

[[To...]]

Please sign in to view this quote.

For a minute, I thought this was a reply from Andreas 😃 Please elaborate on the 2 circles. I think I understand the curve.

[[Mi...]]

Please sign in to view this quote.

He's talking about the three inputs to the kink point. The first should be a curve that goes over the kink, the second should be an arc recalled from the curve just before the kink, and the third should be an arc recalled from just after the kink.

I can't vouch for it, though, as I've never tried it myself. And I'm not sure if the order of the last two features matters.

[[Ri...]]

Please sign in to view this quote.

Nevermind. You cannot select which side of the intersection you want. Intersection is the only way - it may not work because Calypso hates trying to intersect 2d features.