Small Profile, Large Sphere

[[Ja...]]

I have an I.D. Sphere that measures roughly Ø15.25". The print is asking for .001 Profile of a Surface to no datums. I have about 1" of the sphere in height - not much of the surface at all. Also, the O.D. to I.D. thickness is less than a half inch. Is this reasonable on a CMM?

Preliminary results: 6 Measurements, Largest measurement - Smallest measurement = .0009 : This would lead me to assume the repeatability of the measurement is equal to the tolerance and therefore we shouldn't try.

[[Ro...]]

I use points instead of scanning when dealing with small sections of a sphere.
Ive found 12-15 points to be very good for size and location.

12 points-2.png

[[Er...]]

But if you have a profile with out any datums, location is unimportant, and 12 points is not enough to accuratly determin form errors. (depending on size of the feature)

In your case I would have used ~250 points with the tolerance in mind. If you are scanning, reduce speed and consider a larger probe diameter for better mechanical filter if repetabillity is an issue. Also make sure your paths of scanning dont contain any step change in direction. Thats what I would try first.

Or just take 5 points and report form 🙂