Inner/Outer Tangent Diameter on Torus

[[Me...]]

Just wondering how some of you would come up with the 2.263 basic diameter as shown below.

I am scanning 4x 2d curves, recalling them into circles, using result element and adding the radius to the location of the center of said created circles from curves. I feel like there is an absolute easier way of doing this 👩‍🏫

[[Cl...]]

If you have two curves 180 deg apart in the Y axis (or what ever axis yours are in), you can use the minimum/Maximum
coordinate to get the furthest points of that axis. Insert the min point Y + the max point Y of the curves in the result element.

[[Ch...]]

Well,

Please forgive me, but I'd wonder why that would be a basic and who wants "Actuals" for a basic on an inspection report.
The basic should be tied to possibly a profile or maybe a position, but in this case, it is unusual for this diameter.

Can you tell us if there is a FCF or tolerance tied to this basic Ø?

Anyways, to answer the question if I had to measure this , I would use either 2D curve, or radius from the cad model, and probably pick out MaxCoordinate in 3 or 4 places around and construct the circle out of that.

Or possibly if this diameter also has a basic linear attached to it, I might just measure points or scan at the basic linear.

Good Luck !

[[Me...]]

This particular company requires all basics to be included with FAIs, that diameter is just a reference for a .002 Profile tolerance.

It's extremely annoying to give all the basics though, that's for sure.

[[Am...]]

One of our customers requires all Basic Dimensions to be measured as well it is ridiculous especially with some of their prints being so convoluted, the dates on the prints are from the late 60s early 70s and I know no one was giving them any sort accurate results for some of these.

[[Me...]]

Struggle = Real

[[Da...]]

2D curve, scallop patterned ( 6-8 times) create a theo plane at location , create intersections of each 2d and the plane, recall into a circle. Don't ask me how to get the actual location for the plane cause there isn't enough radius to get a good feeling about that...

[[Me...]]

I constrain the radii to get the location, constrain the location to the radii.

That part is easy.

[[Da...]]

There is a subtle difference between using Multiple Radii as opposed to using multiple curve intersections to produce a constructed circle. I don't know how much it will matter in your case.

When you use circles and then intersect those circles with a theo plane what you get is a series of PERFECT circles based on the NON perfect sets of data. So, if the small sections of circle have a high form error you might get some skewed intersection points.
( due to their small section, they will probably be a bit off)

Using curves produces an actual intersection at the actual plane. This is the closest you can get to contacting the part at the exact height. ( of course your height will be a projected PERFECT plane so there's that !)

In my opinion , since we are going to recall points into a circle, I want actual points. We are going to Create a new PERFECT circle from the data. Why would you want to start with already , in a sense, "filtered" data ?

I work with ultra tight tolerances so all this makes perfect sense to me , if you have more open tolerances then this is probably overkill !

[[Me...]]

I've worked with tight tolerances for 2D & 3D profile scans and have never used a projected plane for measurements like the ones I need. I'm not having any issues with the location or radius. I just needed that 2.263 tangent diameter and after a few different methods, the most repeatable and logical method is using Polar Distance with the Maximum setting.

I am definitely interested in seeing the results of projecting my recalled radii into a projected plane.

[[Da...]]

Shawn, just an idea: what if you use some kind of small enough star stylus with big enough ruby balls to do a self centering circle?
https://qualityforum.zeiss.com/topic/17464-building-first-star-probe/ of star stylus borrowed from Richards

Edited Tuesday at 02:18 PM
replaced link from messtechnik zu qualityforum

[[Ri...]]

That would be my first go-to. I'd probably use Curve to help find the location that I'm looking for first (especially if I'm using an XXT).

Theoretical measurements are great, but I don't think anything beats actually putting a probe on something.