Measuring a Theoretical Sphere

[[Ev...]]

Morning all,

I've been brought a small part from one of our other plants and asked can I measure it on our Contura. The part is 1/8th inch tall and has a spherical radius callout on it, see attached. When I asked if we had gage balls for this, I was told the last one was dropped and couldn't be found... 6-8 week lead time on a new one apparently.

My question is can this spherical radius be measured on a cmm? If I scan the shown surface, will that be enough to tell me what that radius is? What would be the best practice for a base alignment? The inside is a hex, with a flat bottom.

[[Er...]]

It looks there is is a top plane on the Part? If there is, and it was cut with the same tool that cuts the spherical Radius. I would use the top plane as the origin and then translate your alignment in Z+ .885" which would be your true center at nominal. Next measure a circle taking points on each side of the hex. Evaluate your Radius sized, with the X, Y, and Z Constrained. Also, I would check what your X, Y, and Z values are when the radius is constrained. If they are close to 0.000, then your Radius will be accurate.

Also, I would consider checking the surface as a profile and see how true the spherical radius was machined. This will give you an idea if there is form error as well. The actual size will be found within the average of all of these.

Curious, How tight of tolerance are you dealing with?

[[Er...]]

It can be, but it may take several attempts.

The Base Alignment can be just for finding the part reliably. Really that is it's main purpose.
Maybe take some points and make a 3D Best Fit alignment? Definitely with a tight loop.

For a measurement strategy I would use many circles as I could from the outside edge in to the hex.

I would use CAD evaluation to see how the shape of the radius looks. It may also point out errors in scanning.

[[Ma...]]

If you can obtain how it was made ( i doubt lathe ), then you can figure out what can be best approach.

If OD and inner sphere was made with one tool and one path, then you can be sure with plane. I would preffer touch points instead of scan paths.

[[Ev...]]

Please sign in to view this quote.

Yeah, there is a plane on the top of the part, which is where the origin is. I also like the idea of checking the profile. The tolerance is ±.005" for the spherical radius.

Please sign in to view this quote.

Definitely a few ways to make the BA, I think 3D best fit might be easiest. When it comes to making loops would you go with 0.002 for the break condition? I have a terrible time deciding what type of break condition I should go with.

[[Er...]]

"Definitely a few ways to make the BA, I think 3D best fit might be easiest. When it comes to making loops would you go with 0.002 for the break condition? I have a terrible time deciding what type of break condition I should go with."

I always set the loop for 5 times. If Calypso starts on a third I need to find out what is going on. Damaged part? Inconsistant parts that foul the Base Alignment? What I don't want is to waste measuring a part that the CMM can't find reliably. That may alter everything measured.

I like 0.002 as a good starting point with that range of tolerance. If there are a really tight dimensions I will set it at 0.0001, watch the default printout to see what the delta is coming in at, and set it a little higher than the delta.

[[Ri...]]

Looks like a CAD model, Freeform it.

[[Da...]]

Please sign in to view this quote.

If you use a result element, you can actually include the loop delta in the report - just output "baseSystem().valueA". I habitually include this - that way when I am using pallet fixtures etc, I don't need to look at the default report. You can also have it auto-abort if it fails if you run from characteristics.