Scanned two surfaces

[[Ky...]]

I scanned 2 surfaces on each end of a shaft 1.601/1.6014 tolerance. Now ew are having a disagreement with the floor. on the CMM we are getting 1.60176 on one end and 1.60148 on other end. Question when it scans does it take an average, the highest points or what. The people on the floor are trying to find the .00027" with a mike and want to know why they can't. I am at a loss. just did a re-qual and all probes are within less then 1.5 microns. I need to be able to tell them something because we are trying to prove out a new machine and I want my data to be correct.

[[Ow...]]

How it calculates the size is how you choose to evaluate it. By default, LSQ (avg) is what is reported. You could try outer tangential element and Inner tangential and see how it changes but, if your roundness (form) is good, there shouldn't be too much difference.

It's not unusual for operators to measure a smaller diameter with the average micrometer, it's a feel difference.

One way to prove the CMM is right is if you have a certified or proven XX or XXX Master cylinder around the same size and scan it on the CMM. If it measures good on the CMM, then have them measure it with their micrometer but, if possible, don't let them know the certified size. In my experience, 9 times out of 10, they'll measure the size smaller.

[[Ky...]]

These are 4 flats on a cylinder. 2 on each end not the cylinder itself. and mearuring across the flats is 1.601 to 1.6014.

Thanks

[[Ow...]]

Ooops, that's what quick skim-reading does.
Same concept though if you have a master part.
Make sure your filters are taking out outliers properly

[[Ri...]]

I assume that this is a Plane-to-Plane measurement? How big are these Planes? By default, you are measuring using the LSQ functionality, but that can be changed to show the max or min if you want.

I'm not a fan of correlating CMM measurements to a hand mic because the measurements on the hand mic vary from operator to operator, and only show the local size (two-point), and not the same thing you are measuring/evaluating on the CMM.

[[Ky...]]

Richard
Planes are 1.00 X 1.50. on both sides of the part.
Kyle

[[Ad...]]

If you need correlation to a micrometer the best way to get there is using point to point distances rather than planes. Fitting will cause correlation issues anytime the surfaces aren't parallel and flat. On a surface that size I would do four outer quadrants and one in the middle minimum with point distances. (assuming you're in inches not mm)

[[Ma...]]

Try using Caliper Distance. You can output minimum, maximum, and mean distance between the features.

[[Ja...]]

This attachment is a good example of how you could get different results. You might be able to create profiles and use the CMM graphics to show how they could get what they got with a micrometer.

LNL_Distance.pdf